Lifestyle information and access to a commercial weight management group to promote maternal postnatal weight management and positive lifestyle behaviour: the SWAN feasibility RCT

Inclusion criteria

Women eligible to participate were those aged ≥ 18 years who could speak and read English, were expecting a single baby and had not accessed weight management groups in the index pregnancy.

Women who booked for their pregnancy care at the trial site were eligible to be randomised to the feasibility trial if they met the following criteria:

• were aged ≥ 18 years

• had sufficient understanding of spoken and written English

• had no current diagnosis of major psychiatric disorder documented

• had no known abnormality detected in the fetus

• were not involved in another postnatal study (to reduce ‘burden’ of research participation)

• had no identified medical complications (e.g. cardiac disease, type 1 diabetes)

• had no identified eating disorders

• had not undergone previous surgery for weight management.

Exclusion criteria

Women who did not fulfil all of the inclusion criteria were not included in the trial.

Joining the commercial weight management group

As the intervention was to commence 8 to 16 weeks postnatally, it was important that women were reminded about the trial offer. Processes for commencing the commercial weight management intervention included a ‘welcome’ leaflet from Slimming World about joining sessions, which was given to all women allocated to the intervention arm, a ‘congratulations’ text from the research midwives on notification of the birth of the baby and reminder texts sent by the research midwives from 6 to 8 weeks postnatally about joining Slimming World. From 8 to 16 weeks postnatally the women were asked to call Slimming World member services to speak to a consultant who could register the woman, advise her where her nearest groups were located, the venues and dates/times of groups and the name of the local consultant running the group. Following this first call, if a woman was not yet ready to join (e.g. if she felt unwell), she was asked to call Slimming World member services again when she felt ready. At this call, women were asked to provide their height and current weight.

All women were given a Slimming World trial identifier (ID), which was used to ‘track’ the woman through each group she attended, including her weight as taken at each group session. This also enabled women to be tracked who attended groups beyond the trial offer and the extent to which women moved between different groups. In line with all members of Slimming World, women had a ‘swipe card’ that included their free access to up to 12 group sessions and their recorded weekly weight information and date, time and group attended. This approach also meant that women were treated as a ‘regular’ Slimming World attendee, with no potential to be identified as receiving the offer as part of a research trial.

Content of the commercial weight management group

The content of Slimming World group programmes is underpinned by behaviour change models and groups that are homogeneous with respect to content and delivery. 30 Behaviour change techniques are supported by social cognitive theory, with a focus on motivation and self-efficacy for weight management and reducing relapse from the programme. Key techniques include goal-setting, self-monitoring, social support and positive reinforcement. 31,32 Consultants receive standardised training overseen by Slimming World dietitians and nutritionists, which includes motivation to support positive lifestyle changes to manage weight, nutrition and food facts and information on the role of exercise and activity in health and weight management. Consultants repeat training every 2 years to remain up to date with the latest evidence and attend a local programme of safeguarding training approved by the NHS. Groups follow a standard format, starting with a weigh-in for members and an introduction to the programme for new members, followed by a whole-group discussion that includes discussion of group member’s experiences of weight management to help change habits and share healthy swaps (i.e. healthy alternatives to common food choices) and recipe ideas. Sessions can include basic cooking skills, taking cost, cultural preferences and time constraints into account. A food optimising system encourages adherence to healthy eating, and physical activity encouragement includes facilitation of behaviour change, redefining what ‘activity’ can include. The diet includes a combination of different food types: ≈80% combined from fruit, vegetables, carbohydrates and protein; a smaller proportion of calcium and fibre-rich foods; and an allowance for foods high in fat or sugar. Other than limiting the intake of high-fat or high-sugar foods, Slimming World does not promote a ‘restrictive’ diet.

Women allocated to the intervention were offered (with fees waived) the standard membership, namely attendance for 12 sessions run over 14 consecutive weeks to allow 2 ‘holiday’ weeks within the 12-session offer. Slimming World guidance is that, to achieve a 5% loss in weight from baseline (a difference considered to improve health outcomes), individuals need to attend at least 10 out of the 12 group sessions. Attendance was considered in the process evaluation in terms of whether women ‘attended’ for a whole group session or only attended a group session to be weighed, and if women attended ≥ 10 sessions, six to nine sessions, one to five sessions or no sessions at all. On completion of the 12-session offer, women could continue to attend weekly sessions, but Slimming World standard fees would apply.

Lifestyle information leaflet

As postnatal health planning should start in pregnancy,33 an evidence-based positive lifestyle leaflet reflecting current NICE public health guidance for women on breastfeeding, diet, smoking cessation/prevention of relapse, reducing alcohol and managing sleep11,33 was offered following recruitment and allocation to the intervention at 36 weeks’ gestation (see Report Supplementary Material 5). The leaflet was developed with the trial Expert PPI Group and written to comply with Plain English Campaign guidance,34 using pictures and tick/cross messages to enable women of all reading abilities to understand the content. Local women we consulted when developing the trial highlighted that this information was not routinely offered to women, despite recommendations that it should be. 6,33 The research midwives were asked to go through the leaflet with the woman and discuss the content with an emphasis on why the information could support women to adopt practices likely to be of benefit to their own health and the health of their infants.

Proposed primary outcome

The proposed primary outcome for subsequent study was difference between trial groups in weight 12 months postnatally, expressed as percentage weight change and weight loss from the woman’s antenatal booking weight. This outcome was supported by the trial Expert PPI Group, who considered that this would be viewed positively by women, as it did not suggest that they should reach their pre-pregnancy weight.

Secondary outcomes

Data using the following measures were collected at baseline (36 weeks’ gestation) and 6 and 12 months following birth (see Report Supplementary Material 4, 7 and 8). Selection of measures enabled hypothesised shorter- and longer-term outcomes that could result from a postnatal weight management intervention to be assessed. Further information on the measures and why measures were selected for use in the trial are included in Chapter 5.

• Dietary intake: the Dietary Instrument for Nutritional Education (DINE) (University of Oxford, Oxford, UK);35 questions on soft drinks intake were developed for the trial.

• Physical activity: the International Physical Activity Questionnaire – Short Form (IPAQ-SF). 36

• Mental health: Edinburgh Postnatal Depression Scale (EPDS). 37

• Breastfeeding intent, uptake and duration: questions developed for the trial.

• Sleep patterns: questions developed for the trial.

• Smoking: smoking status/cigarette dependence. 38

• Alcohol consumption: Alcohol Use Disorders Identification Test. 39

• Self-esteem: Rosenberg Self-Esteem Scale. 40

• Infant health: questions developed for the trial.

• Impact on body image. 41

• Resource utilisation and costs outcome measures: the EuroQol-5 Dimensions, five-level version (EQ-5D-5L), and the Adult Service Use Schedule (AD-SUS). 42

Trial process evaluation objectives

The process evaluation met objective 4 of the feasibility trial. The aims of the process evaluation were to assess:

1. the acceptability of the intervention and how the intervention was experienced by women, including views on timing of commencement

2. the likely variation in groups attended by women (day/time; whether or not women changed groups/consultant)

3. timing and sources of additional weight management support, including risk of contamination

4. the acceptability of trial processes and procedures.

Trial feasibility objectives reflected MRC guidelines for process evaluation of complex interventions43 with some important exceptions owing to the nature of the trial and intervention proposed. The purpose was not to evaluate the intervention itself as Slimming World weight management groups are a ‘standardised’ intervention, with robust mechanisms to ensure intervention fidelity. Owing to the robust in-built quality assurance and evidence base for the intervention, process evaluation was not designed to answer some standard questions seen in complex evaluations regarding the generalisability of the intervention to other contexts/settings, to provide assurance that implementation/delivery of the intervention has been consistent across trial sites, or to determine mechanisms of impact. This trial reflected a pragmatic trial approach, evaluating the impact of the intervention in the hands of many, where women can choose which group to attend and can switch groups if they like, exactly as they could if they were a ‘standard’ self-referred member of the commercial weight management group.

The acceptability of the intervention and trial procedures was evaluated through brief questions included in the 6- and 12-month questionnaires and through semistructured interviews with women allocated to the intervention arm at 6 months (on completion of the commercial weight management offer) and at 12 months with all women. The probable variation in groups attended by women was assessed using data provided by Slimming World on the groups attended by women (each session attended had a ‘group’ and ‘consultant’ code, and group codes had information regarding the day of the week and timing of the group), and the timing and sources of additional weight management support was assessed through questions in the 6- and 12-month questionnaires.

Ethics arrangements

Favourable ethics approval for the trial was granted by the Health Research Authority (HRA) London – Camberwell St Giles Research Ethics Committee (REC) on 2 September 2016 (reference number 16/LO/1422) and HRA approval was received on 11 October 2016. Approval was obtained from the research and development department of the participating hospital. Table 2 provides details of the substantial amendments to the protocol approved by the REC. The research and development office was notified of all amendments after REC approval was received.

Trial Steering Committee

The Trial Steering Committee (TSC) included an independent chairperson, three independent professional members (a professor of midwifery, a statistician and a researcher with expertise of studies in obesity and weight management) and a patient representative. Non-independent members included the chief investigator. Membership of the committee was approved by the NIHR PHR Committee. The TSC agreed a charter at the first meeting, which was informed by MRC guidance for Clinical Trials Units. 45 The TSC met four times, with the final meeting held to discuss trial findings and TSC views and recommendations with respect to progressing to a definitive RCT. As this was not a Clinical Trial of an Investigational Medicinal Product, an independent data monitoring committee was not required to oversee the safety of participants in the trial and the TSC took overall responsibility for the conduct of the trial.

Questionnaire analysis

Quantitative responses from the questionnaires were analysed descriptively. Free-text questions were read in full by two members of the research team, who independently noted key themes. The thematic framework was agreed and text was coded and labelled according to the three dimensions of the COM-B framework. Data were then examined using frameworks to compare patterns between different groups of women and their responses in relation to capability, motivation and opportunity (e.g. comparing women who attended different ‘doses’ of sessions).

Interview analysis

All interviews (intervention and control) were transcribed verbatim. First, transcripts were read and re-read to ensure familiarisation with the data. Second, interview data were organised and coded using the COM-B framework, initially by two researchers independently (CT and VB). Coding was compared and discussed before all data were coded. As the elements of the COM-B framework are inter-related (not mutually exclusive), data often fit several of the dimensions and, therefore, multiple codes were applied where appropriate. Coding of the remainder of the data (by VB) continued independently and a summary of all coded data was presented in a framework matrix.

Two members of the research team (CT and VB) reviewed and discussed the coded summary data in the frameworks and compared data with and between women to identify themes and patterns in and between different women, including comparisons according to the number of sessions (fewer than six, six to nine or ≥ 10 sessions) they attended. Comparisons were also made on the basis of weight change between antenatal booking weight and weight at 6 and 12 months. This enabled the identification of key factors influencing engagement with the intervention and behaviour change. Similarly, researchers compared the coded summary data and identified themes and patterns from interviews with control arm women in relation to weight loss or gain at 6 and 12 months to allow for comparison with women allocated to the intervention arm. An example framework is provided in Appendix 1.

Integration of process data from questionnaires and interviews

Questionnaire data were coded to indicate if the women also participated in interviews so that ‘double counting’ could be avoided. Themes were compared across both samples. The final thematic framework applied to both sets of data. Interview and questionnaire data are presented according to theme, with attention paid to instances where one of the data sets expanded on or contradicted the other data set (or if one data set was ‘silent’). This was the case sometimes for the questionnaire thematic data set owing to the specific nature of the questions asked.

Service contacts

Service contact data were collected using an adapted version of the AD-SUS42 used to evaluate antenatal psychological interventions for women with mental health problems. 48 The SWAN version of the AD-SUS asked trial participants to self-report the number of contacts made with a prespecified list of community-based health and social care services, hospital outpatient services, accident and emergency (A&E) departments and number of admissions to hospital. An open-ended question also allowed for service contacts not covered by the prespecified services list to be recorded. The AD-SUS was initially administered as part of the baseline questionnaire in which women were asked to report service contacts over the period of their pregnancy to date of recruitment at 36 weeks’ gestation. This covered use of community- and hospital-based antenatal services and non-pregnancy-related service contacts. The AD-SUS was then administered as part of the 6- and 12-month follow-up questionnaires in which participants were asked to report service contacts over the previous 6 months, including contact with services for both mother and infant (see Report Supplementary Material 4, 7 and 8). To measure and cost medical resources allocated to birth, data on the mode of birth for each trial participant were extracted from women’s maternity records provided by the participating NHS trust.

Unit costs

Unit cost data required for costing community- and hospital-based service contacts (including admissions to hospital) and mode of birth were extracted from the publications NHS Reference Costs 2016/1749 and Unit Costs of Health and Social Care 2017. 50 To cost contact with the weight management programme for women in the intervention arm of the feasibility trial, an enrolment cost of £49.50 quoted by the service provider (Slimming World) was applied, covering registration to the programme plus 12 weeks of programme involvement irrespective of number of sessions attended. This was assumed to be charge per participant, which would be levied on a service commissioning body if a weight management programme of the type evaluated were to be subsidised by the NHS or through a local authority public health budget. All unit costs were reported and applied to service contact data at 2017 price levels.

Quality of life measurement

Quality of life data pertinent to estimating within-trial intervention impact on quality-adjusted life-years (QALYs) and cost-effectiveness at 12-month follow-up were collected using the EQ-5D-5L instrument51 administered as part of the baseline, 6- and 12-month questionnaires (see Appendix 4, Table 29).

Modelling out-of-trial programme impacts

An overview of the available evidence required to support economic modelling of the out-of-trial cost and QALY impacts of improved postnatal weight management as part of a larger definitive trial was undertaken to inform general recommendations as to how this work might proceed. Capturing longer-term impacts of postnatal weight management is likely to be complex, requiring economic modelling of one form or another. A broad assessment into the probable feasibility of modelling longer-term impacts as part of a definitive trial was completed, including the plausible time scales over which these impacts might be assessed. A rapid evidence review of the type of evidence required to develop and parameterise an economic model of this type was completed, full details of which are presented in Chapter 7.

Proposed primary outcome: impact of lifestyle information and commercial weight management sessions on maternal weight at 12 months postnatally

The proposed primary outcome was maternal weight change as calculated from a woman’s BMI score recorded at her first antenatal booking visit, which usually takes place before 12 weeks’ gestation, and at 12 months postnatally. Antenatal weight was estimated as weight at the antenatal booking appointment minus 1.25 kg to accommodate early pregnancy weight gain. Linear regression was undertaken to adjust for the most powerful predictors measured pre randomisation and remove any biases caused by chance imbalance at baseline.

An important feasibility question was whether to use the woman’s trial eligibility weight, which was recorded at her antenatal booking appointment, or her weight as recorded at trial entry at 36 weeks’ gestation. Using the eligibility weight, the standard error (SE) of the estimated treatment effect was 1.29 kg; using the trial entry weight, the SE was slightly smaller at 1.27 kg, in line with the expectation that weight used to define trial entry could have been subject to selection bias.

Weight loss as assessed at 12 months postnatally was greater than at 6 months (Table 4), supporting the decision to use 12 months as the primary end point for the feasibility trial.

Secondary outcomes at baseline and 6 and 12 months: influence of lifestyle information and weight management sessions on women’s health and health behaviour

Previous studies have shown that a number of health behaviours during pregnancy are associated with poorer outcomes for a woman and her child. These include tobacco smoking,57 poor nutrition,58 heavy alcohol consumption59,60 and use of illicit drugs. 61–63 These are compounded by poor health behaviours before and after pregnancy, including lack of physical exercise, having a high BMI score and not initiating or persisting with breastfeeding. 56 Secondary outcomes included a range of measures to assess the potential for the intervention, which includes a positive lifestyle information leaflet, to have an impact on women’s lifestyle and positive health behaviours after giving birth as assessed at 6 and 12 months postnatally. We also wanted to consider if the measures of health behaviour used were most appropriate to inform outcomes in a future RCT, and if women would be willing to complete these measures.

Women’s weight reduction by > 5% and > 10% of weight at trial entry

A pre-planned subgroup analysis of the primary outcome was undertaken for women who were overweight (BMI 25.0–29.9 kg/m²) and obese (BMI ≥ 30 kg/m²) at antenatal booking. Although differences were not significant, there were more women with a > 10% weight reduction at 12 months in the intervention than in control arm (Table 5). There was no evidence of an interaction effect (which would suggest that treatment may be more clinically effective in particular groups), although this test lacks power.

Self-reported and research-midwife-recorded maternal weight at 6 and 12 months

There were no differences in weight outcomes among women who self-reported their weight in 6- and 12-month questionnaires posted to the research team (and who did not receive follow-up appointments) and women who were weighed by the research midwives. At 6 months, the effects were –1.62 kg (95% CI –4.18 to 1.65 kg; p = 0.393) in women who had an appointment to be weighed by the research midwife and –5.76 kg (–12.7 to 1.23 kg; p = 0.101) in the self-report group (likelihood ratio interaction test: p = 0.390). At 12 months the effects were –2.22 kg (95% CI –4.97 to 0.53 kg; p = 0.122) in women who were weighed at an appointment and –4.8 kg (95% CI –10.9 to 1.3 kg) in the self-report group (likelihood ratio interaction test: p = 0.390).

Number of weight management sessions attended

At 6 months postnatally, although some women allocated to the intervention arm were still attending sessions as part of the intervention offer, there were no detectable differences in adjusted weight loss [i.e. a difference of 0 kg (95% CI –7.28 to 5.37 kg; p = 0.765)] between women who attended 10–12 sessions, women who had attended nine or fewer sessions and women in the control arm. By contrast, at 12 months postnatally there was a difference in adjusted weight loss of ≈5 kg (95% CI 1.05 to 8.93 kg; p = 0.013) between women who attended 10–12 sessions and all other women, indicating a ‘dose effect’ response in terms of ≥ 10 sessions as recommended by Slimming World.

Attendance

Of the 98 women assigned to the intervention arm, and according to Slimming World records, 46 women (47%) attended at least one session and 52 (53%) did not did not attend any sessions. Of the women who attended at least one session, 19 (41%) attended 10–12 sessions and 19 (41%) attended only one to five sessions (Figure 4). Among those attending, the mean number of sessions attended was 6.74 (SD 3.94). Two women recorded as not attending any weight management sessions were found by the research team to be anomalies. Data could not be found by Slimming World for one woman who reported attending 12 sessions, and a second woman who reported going to a first session stated that she was not able to join the group owing to requiring information about her weight that she did not have (see Non-attendance for further details). The denominator is therefore 46 when based on data provided by Slimming World and 48 when based on questionnaire data.

FIGURE 4.

Slimming World sessions attended by women allocated to the intervention group who attended ≥ 1 session (n = 46).

In the intervention arm, the 12-month questionnaire was completed by 62% of women (32/52 women) who did not attend any sessions and by 80% of women (37/46 women) who did attend Slimming World: 17 women who attended ≥ 10 sessions, seven women who attended six to nine sessions and 10 women who attended one to five sessions. Both of the women who attended Slimming World but did not have their data recorded by the organisation responded to both the 6- and 12-month questionnaires.

Non-attendance

Of the 52 women who did not did not attend any weight management sessions, 39 (75%) provided a reason for this in the questionnaire. Women offered several key reasons, most of which related to ‘opportunity’ issues within the COM-B framework (see Chapter 2). The most common barrier reported was finding time to attend; more than half of women who responded indicated that their circumstances did not allow them time to attend groups. Several explained that looking after their baby took up all their time or energy:

Difficulty with child care and generally getting to grips with being a new mum was more important.

Snapdragon, lost 12.8 kg, 12 months

Some specified that breastfeeding on demand made it difficult to make time, others related that they were looking after older children and several mentioned that they had no child-care support to enable them to come:

I chose to breastfeed on demand so I would not have been able to attend any scheduled sessions easily.

Rose, gained 12.0 kg, 12 months

The timing of sessions in the evening, when children were going to bed, was an additional limiting factor for a few women. Two participants commented that the groups were offered too soon after having their baby, and both indicated that 6 months post partum would have been more suitable for them [see Accessibility of the intervention (physical opportunity), Timing of commencement of the intervention].

Women also identified circumstantial and social factors that meant that they did not have the opportunity to attend. Several women said that their babies were unwell or that they themselves had physical or mental health problems after the birth that prevented them from attending. Furthermore, there were two cases where the women were not enrolled in the programme on time, reporting that there were administrative errors by Slimming World. Other reasons included, as reported by one woman, that as a non-native English speaker she was not confident enough to attend, and, as reported by a few women, that they were either out of the country for a prolonged period of time or had moved away from the city. Only a small number of women in this group found that the groups were difficult to access in terms of location.

Together with these ‘opportunity’ factors that presented barriers to participation, there were also several ‘motivational’ factors highlighted. Some women felt that the weight management group was not suitable for them: they did not feel motivated to lose weight or they lacked confidence in the effectiveness of the intervention. Some were happy with their weight or felt they had already managed to lose enough; one joined an online weight management group that she found helpful and very convenient to access and another decided not to participate because she did not agree with Slimming World’s approach (i.e. minimising fat consumption), misunderstanding that this could mean that the diet allowed high-sugar foods:

Slimming World is based around minimising fat consumption within the diet. I don’t think it is healthy to completely remove a food group from the diet. I also strongly disagree with 0%-fat products as their sugar content is very high and this has its own problems.

Freesia, lost 2.6 kg, 12 months

Finally, one participant appeared to interpret the trial as being targeted at losing pregnancy weight gain, perhaps not recognising that she had an overweight BMI score (27 kg/m²) at the start of her pregnancy. She felt very strongly that the weight management offer was not targeted at new mums and therefore was not relevant to her:

When I called to make the first appointment it became clear that Slimming World was targeted at overweight people rather than new mothers. I was asked whether I would be comfortable with people cheering if I’d lost weight, which made me think I had done something wrong rather than have a child! And also that the programme was targeted to teach me how to eat better, for example by using olive oil and not eating ready meals, which I’d never done. So it sounded more like something for unhealthy people with bad habits, not new mums.

Lily, lost 6.5 kg, 12 months

Views on the weight management intervention

Analysis of the questionnaire and interview data resulted in the identification of five key themes describing women’s experiences of the groups. Each of these is described and evidence mapped against the COM-B domains (Table 17). Themes demonstrate how women perceived and understood the intervention and highlight the salience of factors that may facilitate or act as barriers to uptake and engagement. Each is considered in turn in this section.

Weight loss aspirations

Most women interviewed expressed aspirations for weight reduction and positive lifestyle changes as reasons for deciding to participate in the trial. However, the strength of these aspirations varied across the sample. Aspirations were often related to whether or not they saw weight loss as a priority after birth and to women’s personal evaluations of their weight.

Women who completed the programme by attending ≥ 10 sessions, and most women who attended between six and nine sessions, acknowledged that they had an overweight BMI score and attributed their weight gain to ‘comfort eating’, ‘eating too much’, or ‘eating rubbish’. Some talked about gaining weight during pregnancy whereas others talked about having struggled with weight for a long time during their lives. All expressed a desire to lose weight, some talked about wanting to break ‘unhealthy’ dietary habits to gain a ‘sense of control’ and others talked about gaining weight during pregnancy or wanting to get back to their pre-pregnancy weight. All were overweight at the start of their pregnancies and had gained 3–18 kg during pregnancy. They were keen to take part and felt that the programme would be beneficial:

I guess I got into the routine of comfort eating and putting on more weight. I thought that if I didn’t take this opportunity I would have really been at a loss, so it would benefit me.

Violet, 11 sessions, lost 17.5 kg, 12 months

I’ve tried SlimFast^® [SlimFast, Palm Beach Gardens, FL, USA], I’ve tried Juice Plus+^® [Natural Alternatives International, San Macros, CA, USA] and I’ve tried Weight Watchers [WW International, Inc., New York, NY, USA]. I’ve previously tried Slimming World. There was a lot . . . I’ve always been overweight. I thought it might have been a bit easier to lose weight after having a baby. I was told by quite a lot of people and I was quite excited to try it to see if it would work.

Hyacinth, seven sessions, gained 13.7 kg, 12 months

By contrast, for women who attended six to nine sessions but had gained weight at 6 months postnatally, or those who attended fewer than six sessions, adapting to life with a new baby in those early months was seen as more of a priority than weight management. One woman who attended six sessions acknowledged that she had an overweight BMI score but did not consider weight management to be a priority; her priority was to acclimatise to life with a new baby and her concern was her postnatal mental health and overall fitness and well-being, but not weight loss specifically. Therefore, she felt that the intervention did not address her particular needs:

There wasn’t a holistic approach in terms of the intervention, it was just, ‘right, you’ve got a baby and 3 months to get rid of the weight’. It felt a little disconnected. For me, 3 months into having had a baby, I wasn’t standing there thinking I’ve got to lose the weight. I was standing there thinking, ‘wow, how do I reacclimatise and recalibrate my life to accommodate this little person?’ So I think it was too early and again I don’t think it was the right intervention.

Aster, six sessions, gained 2.8 kg, 12 months

Women seemed less aware of having overweight BMI scores at antenatal booking and focused more on evaluating their weight on the basis of pregnancy weight gain. In this context they did not consider themselves to have an overweight BMI score:

I didn’t gain too much weight during pregnancy. Compared with the people who are at [name of Slimming World group], I really was the slimmest there. The people who are at the group really needed help, you could tell, it’s quite a big difference between me and them.

Anemone, one session, gained 2.7 kg, 12 months; BMI score at antenatal booking 28.3 kg/m²

I’m not too much on the big side. When I was pregnant the body changed and I didn’t add too much weight . . . but I think I just wanted to be able to get back to my shape, post pregnancy . . . I’ve never been overweight. I’ve just always tried to maintain a healthy [lifestyle], tried to eat healthy and tried to keep fit as much as I can.

Iris, one session, gained 6.8 kg, 12 months; BMI score at antenatal booking 29.3 kg/m²

Beliefs and expectations (motivation: psychological capability)

In interviews, most women described long histories of weight issues and prior experience of successful and unsuccessful attempts at weight loss (including weight loss to conceive or for occasions such as weddings or holidays). Previous attempts involved adopting a variety of strategies, most notably some form of exercise, and for some women adopting short-term ‘diets’ such as the SlimFast Plan^® (SlimFast) or the ‘cabbage soup diet’. Three women also had prior experience of commercial weight loss programmes: Slimming World, Weight Watchers and the LighterLife^® programme (LighterLife UK Ltd, Harlow, UK).

Interviews revealed that women’s previous experiences of weight loss and their beliefs about postnatal weight management influenced how they perceived the intervention and in particular their perceptions of its efficacy.

Women who completed the programme, and the one woman who attended seven sessions and reached her target weight, acknowledged that previous attempts had provided them with only short-term benefits in relation to weight loss or that their attempts either through diet or exercise had been unsuccessful. They were often surprised to learn about the role of diet and exercise in weight management, and attending sessions and losing weight facilitated a change in previously held beliefs:

I’ve gone on various diets but half-heartedly, and I’ve always done it though exercise and worked extra hard on the exercise side. So it’s a kind of revelation that it’s 90% food and 10% exercise, which is what I learnt at Slimming World. I knew that it was about being aware of what you are eating and how that can help you lose weight and also how you add exercise into the mix . . . Seeing five and a half pounds’ loss in the first week confirmed it would work. I am proud of myself.

Heather, 12 sessions, lost 2.6 kg, 12 months

At 12 months these women reflected on and described once again about learning about the role of diet and exercise in weight management as key, even if they had gained weight since completing the programme:

One thing I learnt at Slimming World is that diet is more important than exercise. Exercise alone is not going to lose the weight and the impact of cutting out calories in your diet is much bigger than going to the gym. When I was at university I was very active and did lots of sports and I couldn’t understand why I was still overweight. When I look back now to what I ate and drank at university, it’s no wonder.

Hibiscus, 10 sessions, gained 0.3 kg, 12 months

Their beliefs in the intervention were strengthened as they felt that it was safe and relevant to their needs as breastfeeding mothers:

They’ve got a handout for breastfeeding and pregnancy, so you know that the diet is safe to be on and they’ve obviously done their research.

Azalea, 11 sessions, lost 1.8 kg, 12 months

In the questionnaire, two further women who had attended ≥ 10 sessions indicated that the intervention had increased their knowledge and skills: ‘I still go off track but then I just go over the programme myself at home’ (Lotus, 10 sessions, lost 8.1 kg, 12 months); ‘Learnt a lot of substitutes and better choices of cooking foods’ (Gardenia, 11 sessions, lost 1.7 kg, 12 months). Two women mentioned the psychological benefits of attending a weight management group: ‘It made me feel like I had control back over my body’ (Azalea, 11 sessions, lost 1.8 kg, 12 months); ‘It’s been amazing for my self-confidence returning to work’ (Cyclamen, 12 sessions, lost 9.8 kg, 12 months). Both of these women lost weight over the course of the trial and said that the intervention was effective; one still attends Slimming World, and the other stopped for financial reasons.

By contrast, women who attended fewer than six sessions and gained weight tended to be more mistrustful of the intervention. One area of concern was their perception that the programme failed to prioritise exercise. One woman who had previously managed her weight by attending Zumba^® (Zumba Fitness, LLC, Hallandale Beach, FL, USA) classes commented on the lack of exercise as affecting her motivation to continue:

I just felt there should also be an activity, there wasn’t any exercise classes or anything like that and maybe that side of things didn’t attract me . . . I would have also been quite motivated for me to try and go back if I knew that maybe there was some form of exercise class that would be involved in the session.

Iris, one session, gained 4.4 kg, 12 months

Two women commented on this aspect in their questionnaires too. One stated that she ‘did not agree with some of the information regarding exercise’ (Poinsettia, one session, gained 1.8 kg, 6 months) and another that the diet ‘was not geared towards people who train . . . was convoluted – prefer to focus on exercise’ (Poppy, two sessions, lost 7 kg, 12 months).

Another area of concern was the perception that Slimming World was about marketing their own products and promoting the consumption of artificial sweetener, which did not fit in with the beliefs of one woman who attended one session. She was sceptical about the focus of the plan and about the group’s intentions:

I didn’t know what to expect and I was hoping for activity and not just sitting on a chair and encouraging ourselves. I’m well educated in how to lose weight. I didn’t like the idea of having allowable ‘syns’ [points]. I saw they had some boxes [products] with allowable syns which were basically sugar . . . Everything was a Slimming World product; it was mass advertising for their products. So I was sceptical about that and the intentions as well.

Anemone, one session, gained 2.7 kg, 12 months

Several women who attended between six and nine sessions were mistrustful for a number of other reasons, most notably their belief that the plan was unsuitable for women who were breastfeeding. There was a general misconception about the need to eat fat in order to breastfeed. A woman who had previously attended a LighterLife weight management programme commented that she had been sceptical about the ethos of the plan and its promotion of sugar-free food, which she felt was not good when breastfeeding:

I was quite disappointed and quite shocked . . . I was expecting it to be quite advanced ideas, which was the LighterLife stuff, which was you look at your relationship with food. I found I was eating a lot of sugar-free stuff, which I didn’t like very much breastfeeding. Knowing that . . . it’s quite a low-sugar, low-fat diet. I was also really reluctant to cut my fat levels down too much so I still tried to keep in some full-fat milk and I would still eat more fatty avocados and stuff like that . . .

Lavender, nine sessions, lost 0.4 kg, 12 months

In questionnaire responses from those who had attended < 10 sessions, although two women thought that the advice they received in the weight management group was ‘reasonable’ and ‘common sense’ (Sunflower, two sessions, lost 2.4 kg, 6 months), two noted that it was targeted more at people with ‘no healthy eating knowledge’ (Bluebell, three sessions, lost 5.4 kg, 12 months) and that consequently they had not benefited as much: the sessions had ‘not improved my health all that much as I already had some knowledge of healthy eating and exercise’ (Geranium, three sessions, gained 9.9 kg, 12 months). One woman who was also interviewed, and who described the role of exercise in the programme as a barrier, indicated in the questionnaire that she had only partially followed the advice given at the sessions as she found it challenging: ‘some interesting advice, which I took on board, but struggle to plan diet and follow through’ (Allium, three sessions, gained 5.0 kg, 12 months).

Understanding and implementing the intervention (capability and motivation)

Capability in relation to understanding and being able to apply the Slimming World programme appeared to have a dose response effect: the more sessions women attended, the more they appeared to understand the plan and were able to incorporate it into their daily lives.

Women who completed the programme, and one woman who attended seven sessions and reached her target weight, described changes they had made to their eating habits and physical activity. These included walking more or participating in BuggyFit^® (Shabbington, UK) classes with babies as well as dietary changes such as reducing cereal or bread consumption and replacing oil with low-fat sprays. They commented that adopting the plan was ‘easy’ and ‘uncomplicated’ (Begonia, 10 sessions, gained 6.3 kg, 12 months):

Slimming World was a lot easier in terms of a lot more practical for life. I liked the fact that you could have the syns, so you don’t feel guilty having them if you’ve had a terrible day, or whatever, so I liked that. Not having to weigh food makes a big difference.

Amaryllis, 11 sessions, gained 1.4 kg, 12 months

Some felt that that the plan reflected what they were already eating, but they had made some adaptations including changes to cultural foods. Others talked about specific Slimming World recipes that they found useful, e.g. Diet Coke^® (The Coca-Cola Company) chicken, or using their syns when making adaptations to recipes:

It was very easy actually. I was actually quite surprised that a lot of the foods that I was eating I was able to continue eating. I’d have to make adaptions to the sizes or fat-free versions but all in all a lot of the meals that I could prepare are meals that I was used to making at home. It’s just making those little tweaks. I could still cook the barbecue chicken if I wanted to but obviously I’d have to syn the barbecue sauce or adapt to a Diet Coke chicken. I tried to stay away from the fried food. I can just bake the chicken instead of frying it.

Violet, 11 sessions, lost 17.5 kg, 12 months

Women also talked about planning meals and using online resources and the Slimming World software application (app) to help them make better decisions and choices. They attributed gains in weight to straying from the plan. However, they saw the plan as something that they could return to when this happened. Women also described the plan as unrestrictive and something that promoted empowerment and a sense of belief and control and fitted well into postnatal life:

I also knew why I’d gained; it was always the weeks when I’d gone out or eaten something that I knew I shouldn’t. I was aware of that. There were some weeks where it was like a birthday or there was something going on, but it just gave me that element of control. It was just very empowering, it was very much like, ‘you can do this. There is loads of food that is available for you to eat, you are not restricted in any way’. It honestly didn’t feel like a diet . . . If I was breastfeeding, I could just stand up and get a Babybel^® [Bel Group, Paris, France] and there were things I could pre-make and just have ready for myself so I could still do the diet but still be looking after a newborn when you are tied to the sofa. It was nice to think, ‘yes, but I am not eating biscuits, I’m not out of control’.

Azalea, 11 sessions, lost 1.8 kg, 12 months

Women in this group also saw the plan as sustainable and a lifestyle change:

It felt to me to be sustainable, not a harsh diet that I was only going to do for a few weeks. I was still enjoying myself and enjoying my food and I was still losing weight, so that was very motivating.

Hibiscus, 10 sessions, gained 0.3 kg, 12 months

However, at her 12-month interview, this same woman described some difficulties with implementing the plan beyond the intervention period without the support of attending the group and continued use of the Slimming World app:

It’s easier to let yourself off the hook when nobody is checking up on you. At the time it was easy enough and I’ve been trying to keep going with some of the basic principles. One thing I used to use a lot is the app and it had a syn calculator but once your membership lapses you don’t get to access it anymore. Just to continue to have access to the app resources would have been helpful.

Hibiscus, 10 sessions, gained 0.3 kg, 12 months

Women were motivated by what they perceived as the wider benefits of adopting the plan, including benefits that they could pass on to their baby in relation to nutrition and that helped them to stay healthy to look after their baby. These women continued with Slimming World beyond the trial intervention period:

For me it was better because I knew that I was having more healthy foods, so having spinach and stuff like that incorporated into my baby even more. I know that my baby was benefitting from that.

Violet, 11 sessions, lost 17.5 kg, 12 months

Adopting the plan was also seen as beneficial to addressing other health issues, for example hypertension:

I think with this where I am is the high blood pressure and I’m an older mum, it took me 15 years to have her, so I want to enjoy this. I don’t want to be ill and having to go to the doctors and worry about my blood pressure. I’m off the pills completely now.

Orchid, 10 sessions, lost 3.7 kg, 12 months

Women in this group were motivated by seeing a positive difference in their weight and being weighed every week to keep them on track. In this context, women talked about their sense of accountability throughout the active intervention period; this included accountability to the group and to themselves, which had a motivating effect:

Also for me knowing that I have to go in and get weighed and knowing that it would get told to everyone whether I’d gained or lost kind of gave me that little check to stay on top of it.

Daffodil, seven sessions, lost 13.0 kg, 6 months

The weighing keeps you on your toes like ‘oh I’ve got to weigh in’ and you want to do well and you get in a little competition with yourself.

Orchid, 10 sessions, lost 3.7 kg, 12 months

Four women from this group were followed up at 12 months, two of whom had continued with Slimming World beyond the intervention period. They all talked about continuing to implement the principles they had learnt through attending the programme, although not having access to the Slimming World resources made it more difficult to stay on track and resulted in weight gain for women who were no longer attending. However, all women described a healthy diet as incorporating the key principles of the weight management intervention, which they could use moving forward:

A healthy diet is the one that I think a lot of the recipes at Slimming World give you. I do like a lot of naughty things but I think interspersed with, you pull out these tools out of your toolkit when you need to. It was a really really good thing to do.

Heather, 12 sessions, lost 2.6 kg, 12 months

Three further women who had attended ≥ 10 sessions also mentioned similar themes in the questionnaire response. One noticed the impact on her health [‘health has definitely improved; weight loss has noticeably improved my fitness’ (Cyclamen, 12 sessions, lost 9.8 kg, 12 months)], a second commented on lifestyle change [‘improve healthy lifestyle choices’ (Azalea, 11 sessions, lost 1.8 kg, 12 months)] and a third mentioned the impact on her family [‘the advice was very easy to follow, even used on kids dinners, and I lost over 1.5 stones and felt good’ (Pansy, 10 sessions, lost 4.5 kg, 12 months)].

Despite not attending the full programme, four women who had attended six to nine sessions and completed the questionnaire indicated that they too had learned to eat a healthier diet and how to have a healthier lifestyle, although two of these admitted to only partially following the advice given. A similar pattern was found with the interviewed women who attended six to nine sessions and gained weight: they described following the plan as difficult and commented that they had not made significant changes to their eating habits. For several women in this group, reluctance to fully implement the plan was related to beliefs about the efficacy of the intervention [see Beliefs and expectations (motivation: psychological capability)] and physical capability (i.e. the ease of following the plan, and for some, the physical opportunity in relation to timing of the intervention):

Knowing that . . . it’s quite a low-sugar, low-fat diet. I was also really reluctant to cut my fat levels down too much so I still tried to keep in some full-fat milk and I would still eat more fatty avocados and stuff like that . . . I just took the free food and went with it without actually counting syns . . . I am not convinced that sticking to the plan would actually help me maintain longer term.

Lavender, nine sessions, lost 0.4 kg, 12 months

For several women, knowing that they had gained weight (e.g. owing to holidays) affected their motivation to attend sessions:

When you know you are not losing weight then you are less likely to go. I think I didn’t go the week after the holiday because I know I’m not going to lose, so I missed the holiday week and the week after.

Lavender, nine sessions, lost 0.4 kg, 12 months

Two women, despite not attending many sessions, reported in the questionnaire that they had learned new skills that could help with their weight loss: ‘Learnt a lot, how to eat healthy, mix food types, etc. I know what I need to do’ (Carnation, three sessions, lost 4 kg, 12 months); ‘Slimming World has taught me to eat smarter; when I go back to work I would definitely like to try their recipes’ (Stock, one session, gained 4.1 kg, 12 months).

The social context (social opportunity/motivation)

The social context of being part of a group-based intervention was a key factor influencing how women perceived the intervention. There were two prominent subthemes: (1) perceived group identity and the group support approach and (2) the role of the consultant. A third theme relating to the wider social context – social support from partners – was also identified.

Group identity and the group support approach

Most of the interviewed women who completed the programme, and one who attended seven sessions and achieved her weight loss goal, had very positive views about their weight management group, in terms of both the other members attending and how the intervention was delivered. Several described group members as welcoming, friendly and accepting of them bringing along their baby, including sometimes offering practical help when they attended. Contact with other members afforded women with a safe, affirming, supportive and low-pressure environment where they felt comfortable to attend and share their experiences:

It’s that low-pressure, comfortable, happy environment that you are invited into, a hotchpotch of grannies and mums . . . Yes, it’s not just losing the weight, it’s everything that goes with it and I think that’s the Slimming World thing almost. It’s a strange club that you get invited to.

Heather, 12 sessions, lost 2.6 kg, 12 months

Drawing inspiration from others was often seen as motivating, particularly when women felt they were struggling:

The main thing is that it gives me the motivation to keep going. And seeing other people have up and down days, it wasn’t so bad when I had my up and down days. I think it was more just drawing from the inspiration of how much weight some of the other members had lost.

Violet, 11 sessions, lost 17.5 kg, 12 months

For some women, attending groups provided them with additional social benefits, for example feeling connected to others and making new friends and consequently feeling part of a community:

I was positively surprised how much the group was supportive and I’m also new in the area so it was a really good way to meet people, like meet new mums. I keep bumping into members in the supermarket and in the park exercising, enjoying the social side of things. I felt part of a little community and the ladies were all very welcoming. It was a bit like going to meet some mates at the pub without calories.

Orchid, 10 sessions, lost 3.7 kg, 12 months

For two women who attended Slimming World beyond the trial intervention period, the group continued to be a source of valued support. Both women had continued to lose weight:

Yes, I did go back . . . I wanted to stay in the same group because the consultant is really nice and I got to know the other ladies, so they became friends.

Orchid, 10 sessions, lost 3.7 kg, 12 months

They [group members] are very good . . . and we all got to know each other well and it’s just nice seeing a familiar face and all being a support for each other.

Violet, 11 sessions, lost 17.5 kg, 12 months

In the questionnaire responses, although most of the women who had attended ≥ 10 sessions concurred with these positive views about the group support, a few did not. Two women stated they did not find attending the groups useful: ‘Did not find group particularly helpful as most information available on the online support repository’ (Jasmine, 11 sessions, lost 16.8 kg, 12 months); ‘I would generally only stay if I had any concerns about my weight that week’ (Violet, 11 sessions, lost 17.5 kg, 12 months).

There were also some that were unhappy with some of the diet recommendations. For example, one woman said her consultant suggested she eat Angel Delight^® (Premier Foods, St. Albans, UK) and another said that she disagreed with the attitude to losing weight presented: ‘The advice in my group wasn’t really about improving health. I’d prefer more info on this than “image therapy”, which I don’t like. Meetings mainly plugging food or how to cheat syns! Or eat treats’ (Lotus, 10 sessions, lost 8.1 kg, 12 months).

For several women who attended < 10 sessions (one to five sessions or six to nine sessions) and gained weight, the group support approach was not seen as beneficial and several felt that they did not identify with others attending:

The people who came and the host were of lower quality than I expected. I didn’t want to bond with them. I also don’t need to come every week to listen to other people saying ‘oh, I’ve lost weight this week and this is what I ate’ and then we all clap and say well done. I didn’t need that type of encouragement.

Anemone, one session, gained 2.7kg, 12 months

Women also perceived others attending their groups to have long-standing and established social connections with each other, which meant that they felt isolated. One woman compared her group to a well-known television sitcom that presented a comical and stereotypical view of weight management groups:

The meeting itself wasn’t bad, to be fair. It doesn’t quite fit with my expectations of motivation, to get motivation to slim, but it was interesting to probably hear some other people. A lot of them were older people and with all honesty I felt like this is more like a club for them. They were lovely, I can’t deny that, even the person that was hosting the event, she was lovely. But it’s still very – how can I put it? – like American funny, I don’t know, a bit like they went to one but sometimes on the comedy shows. But maybe it’s me going there with these ideas of Little Britain [BBC, 2003–7] kind of thing. I had some ideas what to expect and in the end it felt to me like at times that I was in a comedy sketch.

Allium, three sessions, gained 5 kg, 12 months

Comments on group identity/the groups were similarly mixed in the questionnaire responses. Some were very positive: ‘I think attending the sessions post pregnancy is a positive thing and encourages you to start to think with a healthier mind set’ (Tulip, six sessions, gained 3.6 kg, 12 months); ‘The groups were very useful to incentivise me to start to lose weight’ (Lavender, nine sessions, lost 0.4 kg, 12 months); ‘loved the group setting and sharing of experience’ (Carnation, three sessions, lost 4 kg, 12 months). Other women were less positive about this aspect of the weight management group. One woman described her group as ‘not entirely welcoming of small baby/noise’ and felt that it was too evangelical and that there was too much marketing of ready meals and products (Aster, six sessions, gained 2.8 kg, 12 months). Another was disappointed that the groups did not have ‘more cognitive behaviour therapy type info’ (Lavender, nine sessions, lost 0.4 kg, 12 months).

Other comments from those that attended < 10 sessions included that they felt the meetings were not useful or insightful: ‘Meetings are not that insightful or particularly helpful to me personally. Meetings also seem a bit disorganised and lack focus’ (Geranium, three sessions, gained 9.9 kg, 12 months). Several women commented specifically that they did not like sharing their experiences of weight loss and listening to others: ‘Didn’t want to spend an hour listening to others’ perspectives. It didn’t feel like a good use of my time’ (Bluebell, three sessions, lost 5.4 kg, 12 months); ‘Felt a bit like alcoholics anonymous, wasn’t my cup of tea . . . Don’t really like the idea of sitting down and talking about your problems as reasons why one is overweight. I felt like I don’t need a group setting to help me lose weight – not really my style’ (Geranium, three sessions, gained 9.9 kg, 12 months).

Several specifically mentioned that they struggled to identify with the group: ‘I couldn’t relate to any of the other people’ (Poppy, two sessions, lost 7 kg, 12 months); ‘The atmosphere was very city/work based and not welcoming to new mothers’ (Poinsettia, one session, gained 1.8 kg, 6 months).

Only 11 women (30% of those who attended the weight management sessions and responding to the questionnaire) provided information about how many times they had stayed for the whole session (only five women returned the log of attendance that they were asked to complete for each group attended, as described earlier).

All 11 women explained why they left sessions after being weighed. Most of them (n = 9) did not have the opportunity to stay because the timing did not fit with their schedules, the sessions lasted longer than they could stay, they needed to look after their baby or they had other family commitments. One woman explained that the group she attended only had weigh-ins (did not have image therapy). In several cases, women did not stay because they did not like the groups and/or the consultant running them: ‘didn’t feel the sessions were very well structured. Others also left before the end so didn’t feel the need to stay’ (Geranium, three sessions, gained 9.9 kg, 12 months).

The consultant–participant relationship

The personality and approach of the consultant and the quality of support they provided to women was also an important factor influencing their perception of the intervention. For most women who completed the programme (attended > 10 sessions), regardless of whether they gained or lost weight, and the one woman who attended only seven sessions and reached her target weight, their Slimming World consultants were perceived as being friendly, welcoming and encouraging. They perceived their consultants as skilful at being able to judge and pick out what they themselves and others needed in terms of support. Several valued what they perceived as personalised support that reflected their needs as mothers:

She offered support and encouragement. She also sends me special messages regarding breastfeeding, so she encouraged me to have the extra calcium and the other things that you are allowed when you are breastfeeding, so she just made sure I was clear with the information.

Orchid, 10 sessions, lost 3.7 kg, 12 months

However, for several women who attended between six and nine sessions and gained weight, and for one woman who attended > 10 sessions, their contact and interaction with their consultant was less positive. These women described feeling scrutinised and chastised for either not being able to stay for the whole group session (e.g. leaving after being weighed) or not losing weight. They felt that their consultant was not understanding or sensitive to their postnatal situation and the challenges of being a new mother:

The woman who ran the group wasn’t very nice, she was very critical . . . The first time I gained she said to me ‘why did you gain?’ I explained that things are harder when you have a baby and to be able to eat healthier things in the time that you get to eat, and she said ‘well, how important is your weight gain to you?’ I think you need to be able to read people because some people were fine with those sorts of questions, whereas me, I found it quite disheartening . . . It had an impact on my motivation to attend the body image therapy.

Hyacinth, seven sessions, gained 13.7 kg, 12 months

A similar pattern of responses was noted in the questionnaire responses with the majority of those who attended ≥ 10 sessions providing positive comments about their consultant: ‘the consultant is a very helpful person and very motivated’ (Nemesia, 10 sessions, gained 0.9 kg, 12 months); ‘Consultant very positive, made baby welcome and was supportive’ (Peony, lost 3.1 kg, 12 sessions, 12 months). Those attending < 10 sessions provided mixed views on this aspect, with three very dissatisfied with their consultant [describing their consultant as ‘abrupt and quite rude’ and ‘not very kind with her words when members gained’ (Hyacinth, seven sessions, gained 13.7 kg, 12 months), as having a ‘lack of empathy with new mum challenge’ (Aster, six sessions, gained 2.8 kg, 12 months) or as ‘patronising and demoralising . . . She had an attitude of “telling off” ’ (Amaryllis, 11 sessions, gained 1.4 kg, 12 months)]. The last woman persevered with attending despite this owing to her positive view about the Slimming World programme and the support she got from her mother, who also attended with her.

Social support

All participants who attended ≥ 10 sessions on the programme, and one woman who attended seven sessions and reached her target weight, described their families, most notably their partners (and for one woman her mother, who also attending Slimming World), as providing support in their weight management efforts. This was mostly described in relation to food preparation; several women described their partners as being happy with what they cooked and/or participating in cooking low-fat options.

One woman described how she and her husband negotiated the cooking to ensure that she could follow the Slimming World programme:

Well, actually, since I’ve been doing Slimming World my husband has started doing more cooking because he doesn’t mind the food but he’s like OK I need to be more creative and I think he’s trying to tell me that he doesn’t really want to do Slimming World as well, so he cooks and we have lots of curries, so the trade-off is he’ll cook reasonably low fat and I’ll eat what he’s cooking but I will bulk it up with vegetables, so like a hybrid of his cooking and me applying Slimming World.

Heather, 12 sessions, lost 2.6 kg, 12 months

Several women described their partners as also being keen to lose weight, and one had persuaded her husband to join her by the time of the 12-month interview:

I took my husband with me. He lost quite a lot of weight as well. So it was good.

Orchid, 10 sessions, lost 3.7 kg, 12 months

However, for one woman who gained weight again at 12 months, support from her partner had waned:

It’s a lot easier to do [follow the plan] on the nights he [husband] is working. I suppose he just likes certain food, especially when he works a lot of evenings and when he is off he likes to have a proper meal . . . steak, basically.

Hibiscus, 10 sessions, gained 0.3 kg, 12 months

By contrast, women who attended < 10 sessions and gained weight described partners who continued to eat as they had before. Many of these women described making small changes to their diet but generally continuing to cook food in the way they had always done:

If for example we’re going to have a pasta dish I would have mostly vegetables with a very small amount of pasta whereas my husband would have the normal combo. I would still eat more fatty avocados and stuff like that. I didn’t go down the whole slimmer’s swapping, I would just eat more of the free foods. I still needed to have a bit of oil so I was still cooking with oil.

Lavender, nine sessions, lost 0.4 kg, 12 months

In the questionnaires, two women specifically mentioned that it was hard for them to find the motivation to change their eating habits because they did not receive support from their partners and needed to accommodate their families: ‘Hard on portion control because I had to make food for the family, didn’t want to cook just for me. Husband wasn’t keen on programme restricting his diet’ (Bluebell, three sessions, lost 5.4 kg, 12 months). Several women stated that other aspects of support limited their participation. This included lack of child care (in relation to looking after their babies for the duration of the session and/or their older children). Three women felt that it was unsuitable to take their babies to the weight management group (meetings were too long, lights were too bright and the babies became irritable and cried). A further three women noted that they benefited particularly from the remote provision of advice using Slimming World’s online support repository, the Slimming World app or contacting the consultant through Facebook (Facebook, Inc., Menlo Park, CA, USA) or WhatsApp (Facebook, Inc.). These were an additional source of support when they were unable to attend or stay for the duration of the meeting because of child care needs (although one respondent stated that the online system did not recognise the extra foods she could eat while breastfeeding).

Timing of commencement of the intervention

Women’s views about the timing of commencement of the intervention varied across the interview sample and some mentioned this specifically in their questionnaire responses also. Their views about this did not appear to be related to the number of sessions that they attended or changes in their weight. In the interview sample, all the women who completed the programme, two women who attended between six and nine sessions and three who attended fewer than six sessions found the timing of the intervention acceptable:

I thought it was the right time. I think I went at 7 weeks but I thought if I left it any later maybe you get too involved with your life with your baby and you just fall into looking after the baby and not looking after myself. I had recovered from my C-section so I was more mobile . . .

Orchid, 10 sessions, lost 3.7 kg, 12 months

Those who felt that timing of the intervention was not acceptable included three women who attended fewer than six sessions and two who attended between six and nine sessions. These women described the difficulties they experienced with breastfeeding, concerns over their baby’s health and not feeling ready to commit to the intervention because they were struggling with their transition to parenthood:

I simply didn’t have the capacity or the energy at times. Again, I think that reflects the timing of the intervention. Had that taken place at 5 or 6 months when I was weaning, when I’d begun weaning, then that actually might have been far easier because it’s something I could have incorporated into the weaning and the food preparation that I was going through then.

Aster, six sessions, gained 2.8 kg, 12 months

One woman felt she would have benefited from starting earlier; however, she seemed unclear that the window was between 8 and 16 weeks and started the weight management programme when her baby was 4 months old:

I would have wanted to have joined a bit sooner but I didn’t receive the information and obviously you only did it for 16 weeks onwards . . . I personally would have wanted to join it a little bit sooner because I’d already got into a routine, my routine had already changed into what it was. So, changing it again with the baby was really difficult. For me personally, maybe a week after I’d got home from hospital . . . obviously, I understand starting it straight away was really unrealistic. In the beginning I did benefit from it but even if I was just a month earlier I think I might have benefited from it.

Hyacinth, seven sessions, gained 13.7 kg, 12 months

Only four women mentioned the timing of commencement in the questionnaires, one stating that she would have preferred to start the programme during pregnancy and the other three stating that it would have been better to start later when their babies were a bit older. The third, when explaining why she did not go to more sessions and stay for the group, said the following: ‘couldn’t concentrate while had baby with me as breastfeeding . . . offer too early in postnatal period, bad timing of groups with baby, exhausted’ (Primrose, two sessions, lost 0.7 kg, 12 months).

Two women experienced barriers when attempting to join the weight management group. One woman contacted two different consultants at different groups but said that she did not receive a response from either and therefore decided not to pursue joining a group. A second woman (mentioned at the start of the chapter), who attended one session, reported that she was asked for her BMI score and weight details at joining, and as she did not have this information she was not weighed and instead asked to obtain this information before she would be allowed to register. Despite trying to get hold of this information she gave up and did not return to the group:

The issue I had with Slimming World. I think maybe when my initial, when I did try to join a group it took me a long time to actually find one that was close to me and then when I did find one they were asking me for my BMI figures and at the time I didn’t have it to hand and they said they wouldn’t weigh me because obviously I think I came through a referral but there were other people obviously attending that were paying members.

Iris, one session, gained 4.4 kg, 12 months

Group location and timing

In the interview sample, most women who completed the programme chose group locations and times that fitted with their postnatal routines or suited their preference of attending either with their baby or on their own. Only one woman mentioned that ‘sessions in the area were limited’ (Amaryllis, 11 sessions, gained 1.4 kg, 12 months). Some women chose groups that were further away from where they lived so that they could make it an outing and incorporate some exercise that they could do with their baby:

The Slimming World group that I joined was on the other side of [area] and so when I started I’d get my buggy and it was about a 45-minute walk from one side of [area] to the other which when I was doing it I really enjoyed it . . . It became a nice thing because I found a nice route to go via a park and there is a fruit shop on the way that I would pick up loads of fresh fruit and I spoke to the owner of the shop and it was a little morning thing, I made it into a whole morning’s activity.

Heather, 12 sessions, lost 2.6 kg, 12 months

However, two women who did not complete the programme described difficulties finding groups that were close enough to their home at times compatible with their postnatal routines. These women attended groups that were either further away from home but at convenient times or closer to home but at inconvenient times. Regular attendance at group sessions was often affected by holiday plans, child care difficulties that arose from changing postnatal routines, and baby’s or mother’s illness. For some of these women distance and changing postnatal routines were cited as reasons for missing sessions or not completing the programme:

That [group] was the only one within walking distance from my house . . . On most days, I didn’t stay for the after discussion . . . because of the inconvenient timing she was, by that point in time she would be really ratty and ready for sleep, so I didn’t stay. Towards the end as well I got postnatal depression so that made it harder for me to leave the house.

Hyacinth, seven sessions, gained 13.7 kg, 12 months

One woman mentioned that a terrorist attack in London meant that she could not get to her group and that she was not aware she could go to other groups:

I wasn’t sure whether I could attend other groups . . . I wasn’t sure whether I could attend in other parts of London.

Daisy, three sessions, gained 3.3 kg, 6 months

The questionnaire responses show that, even among women who managed to attend ≥ 10 sessions, some women found that the timing of groups (in the evening) coincided with their children’s bedtime, which presented a considerable barrier to them staying for the whole session, and three women who attended only one to five sessions stated that it was the timing of the groups that presented a significant barrier to attendance. The sessions were felt by some women to be too late in the day, and others stated that going to the sessions with their babies did not work as, for example, ‘the clapping was too loud’ (Poppy, two sessions, lost 7 kg, 12 months).

Continuation beyond the intervention

According to Slimming World data, a total of nine women continued to attend beyond the 12 sessions that were offered as part of the intervention. In the 12 months subsequent to them joining Slimming World, these women attended a total of 13–49 visits over a period of 20–52 weeks. All had attended at least 10 sessions in the intervention period.

In the interview sample, several women who completed the programme also talked about wanting to continue beyond the intervention period. They suggested that the intervention should last longer than 12 weeks because they felt that it took time to understand the plan and adjust to life with a baby:

I think the programme didn’t last long enough so you are running around and I know they say if you do something for 3 months then you are in the habit but I think you need 6 months, so that you get used to it. Your life starts getting in a position where you are more in control with your baby . . . I don’t feel that it went on long enough that it became second nature.

Amaryllis, 11 sessions, gained 1.4 kg, 12 months

However, the cost to continue with Slimming World after the trial was a barrier for some women, particularly when they had limited resources while on maternity leave (which was also raised in the questionnaire by another woman):

Once the 12 weeks were up and it came around to paying for it weekly, that was the time when I had no money, so £5 a week ended up being quite a lot of money that I couldn’t really afford. It all came down to money sadly and I think maybe if it was possible to do it for a year, like which is the year that the women have got their maternity leave, that would be fantastic.

Azalea, 11 sessions, lost 1.8 kg, 12 months

One of the respondents to the questionnaire specifically mentioned that the duration was not long enough: ‘My only criticism is that I don’t think 12 weeks is long enough to see any real benefits and to establish healthy eating/lose weight long term. I would say 6 months is optimal’ (Tulip, six sessions, gained 3.6 kg, 12 months).

In the 12-month survey sample there were four women (11% of the 37 women who attended) who continued attending sessions after the end of the 12-week programme. In the 12 months after the birth of their babies they had each lost between 1.7 kg and 17.5 kg (data collected by research midwives).

Weight loss aspirations (psychological capability/motivation)

All women interviewed expressed aspirations for weight reduction and/or positive lifestyle changes as reasons for deciding to participate in the trial. As for the women allocated to the intervention arm, the strength of these aspirations for the women in the control arm varied across the sample. Aspirations were often related to whether or not women felt that weight loss after birth was important and whether or not they perceived a need to lose weight in relation to whether or not they felt they had gained weight during pregnancy.

Two women who had lost weight at both 6 and 12 months felt that weight management after birth was important primarily because of health reasons, beliefs about the link between weight gain and postnatal depression and the longer-term impact of weight on the mother and child. One of these women had an overweight BMI score and the other had an obese BMI score at antenatal booking. In this context, these women were disappointed about being allocated to the control arm and felt that they would have benefited from having some support with weight management after birth:

Well, I have always struggled but I have PCOS [polycystic ovary syndrome], so weight was always something that I kept an eye on and also my mum put on a lot of weight when she had me and has never really lost it, so I was aware that that’s something that I didn’t want to happen. But I certainly had a bigger appetite. I’m absolutely sure that I gained quite a bit. I was slightly disappointed not to be one of the random ones selected.

Beech, lost 5.9 kg, 12 months

Two women, one who had an overweight BMI at antenatal booking (BMI 25.3 kg/m²) and the other who had a normal BMI at antenatal booking (BMI 24.9 kg/m²) and EGWG, had lost weight by 12 months. These women also felt that weight management was important and wanted to return to their pre-pregnancy weight:

I’m not skinny but I’m not fat either, I’m medium. I would say slim-ish, kind of slim, a bit slim. Normal? I’m just normal. So, yes, I wanted to go back to my normal weight.

Chestnut, lost 1.6 kg, 12 months

For one woman her immediate concern after birth was being healthy for breastfeeding but she acknowledged the importance of weight management for future pregnancy:

I was more concerned about my breast milk and just being healthy enough to feed my baby . . . The other thing is that, it’s probably relevant to the study as well, is that I’d probably like to have another baby in the next year to 18 months. So I’d like to be at the same baseline or I’d like to be a few kilos less even than I was for this pregnancy.

Rowan, lost 1.0 kg, 12 months

By contrast, most women who gained weight at both time points did not feel that they had gained much weight during their pregnancy and/or were more accepting of their body weight post birth. One woman who had an overweight BMI at antenatal booking emphasised that she had not gained much weight during her pregnancy and felt that her weight was acceptable at the time of the interview despite having gained further weight postnatally:

To be honest, when I was pregnant I didn’t really gain that much weight because I wasn’t eating well. I don’t really want to be big. With me I don’t want to be very fat, so I’d rather be slim, keep fit than to be fat . . . I said to you my weight is kind of OK, it’s kind of balanced at the moment for an African woman, for a black person, we’re mostly big, very big. But for me I’m just in-between. I was a size 16 but now I’ve gone to 14, which is good. I am very happy with that.

Maple, gained 3.1 kg, 12 months

Several of these women also felt that weight management was not a priority after giving birth. Breastfeeding and the focus on general health and well-being was more of a priority. Women referred in particular to their postnatal mental health needs, but some also had physical health issues, meaning that these women were reluctant to think about weight loss, particularly when they prioritised exercise as the preferred method for weight loss:

In the first year after having [name of son] I didn’t do any exercise and that’s mainly because I was breastfeeding and I just didn’t feel comfortable doing much exercise whilst that was happening . . . It’s not even just so much the weight I think it’s just about getting posture back, so for me I had very bad stomach muscles and a very bad back and I was recovering from PGP [pelvic girdle pain], and I put off doing exercise because I was too scared to do more damage. It’s not so much the losing weight, it’s the getting back the strength, it’s the core strength, and then I think you can only do that bit first and build on that to then lose weight.

Birch, gained 6.2 kg, 12 months

One woman acknowledged that her BMI score was in the overweight category but felt that there was a gap in her care and did not feel emotionally ready to focus on weight management after birth:

I stopped drinking in order to get pregnant and then I let myself off the hook and ate lots of cake because I figured that was better than a bottle of wine. But the focus of the study was about weight management, it was never about my health and well-being prior to becoming pregnant, during my pregnancy or post pregnancy. Post pregnancy my priorities have changed and I’m still breastfeeding, so it was more about doing activities that I could bond with my child. I was definitely not in a good physical and emotional healthy state.

Pine, lost 7.7 kg, 12 months

Weight management beliefs (capability and motivation)

Interviews revealed that women’s previous experiences of healthy lifestyle management and their understanding of what constituted a healthy diet influenced their beliefs about the most efficacious ways in which to manage weight.

Women who lost weight at both 6 months and 12 months during the trial period had a good understanding of what constituted a healthy diet and were actively engaging in managing their weight by focusing on what they ate and combining this with exercise. Women talked about using apps, going to the gym and joining fitness programmes with a dietary element included. A healthy diet was seen as including lean meat, chicken and fish, with lots of fruit and vegetables and a reduction of sugar. One woman had previous experience of Weight Watchers, so this earlier learning may well have contributed to her beliefs. For these women, there was a belief that diet and exercise contributed to their weight loss:

So I’ve managed the weight, to be honest mainly with food. Sometimes I’ve been better than others, sometimes I’m 4 lbs over but then I quite quickly pull that back. Portion sizes, cutting back on alcohol, cutting back on carbohydrates in the evening, bulking up with more vegetables and that sort of thing. I see quite immediate effects from that sort of thing.

Beech, lost 5.9 kg, 12 months

I started this weight-lifting-based programme for my back, my shoulder and my leg, that I really actively did anything like full on top of everything else. I picked that up. It’s made a huge difference to my physical well-being . . . They also have a dietitian as part of the programme, or a nutrition specialist, so they tell me what I should be eating and how many calories to help through the whole process. They recommend MyFitnessPal^® [Under Armour, Inc., Baltimore, MD, USA], and I don’t track it every day, but it’s helped me work out what calories are in things. So it’s been quite helpful.

Pine, lost 7.7 kg, 12 months

Women who lost weight at 12 months had similar views. They were actively engaged in watching what they were eating and one had used the Weight Watchers app for support. A healthy diet was perceived as controlling portion sizes; focusing on fruit and salad; making food healthier by not cooking with oil; avoiding takeaway meals, sweet foodstuff and junk food; and being mindful of the importance of the longer-term goal of being healthy for your family. Participating in the trial was also seen as an incentive to lose weight. Women understood the role of diet and exercise for weight management:

In all honesty I really believe that it’s more diet weight loss is 90% and 10% exercise. I think portion control is the biggest one. I think it’s being creative with what healthy eating is. I think people need to be re-educated about what healthy eating is. I tell you one benefit of the study though: I think it gave me more incentive to get back to my normal eating and to get the weight down so that by the end of the study I was back to my normal weight.

Chestnut, lost 1.6 kg, 12 months

In the group of women who gained weight throughout the trial period, views were mixed. Two women talked a lot about focusing on exercise during the trial period, although there was some mention of changes in dietary habits. For one woman there was a recognition that her weight loss attempts through exercise had not been successful and that she needed to monitor her calorie intake:

I’ve been exercising a lot for about 6 months and I haven’t really lost any weight, so I need to look at my diet and do both together I think. It will be probably be some kind of calorie-controlled diet. I think what I’m going to do is try and have an app that tracks calories, so that I can have in the same app see how many calories I’m burning through exercise versus how many I’m eating.

Birch, gained 6.2 kg, 12 months

Two other women in this group felt that diet was important but they were not actively managing their weight in this way, although they had some idea of what they needed to do, for example cutting down on certain sweet foods and drinks such as chocolate and sugary drinks, eating more vegetables and doing some exercise. They described completing the trial questionnaires as an incentive to think about weight management:

Every time I opened the questionnaire and answered the questions about the healthy food and I remember that I have to do that . . . Sometimes when I leave work, I get the bus but I was walking in, 20–25 minutes, so I was doing that. I was reducing bread and chocolate.

Ash, gained 3.5 kg, 12 months

But with your exercise [i.e. completion of the study questionnaires] I got to ask myself, do I really need that much in a week? How many vegetables do you eat? Bread, salad, pizza, pasta? All those things, I do eat them, but I’ve never think about it, how many do I eat? But right now with the study I’ve been able to work things out myself. So that has really helped me a lot to be honest. I was kind of eating more vegetables, more fruit and I do walk 40 minutes to work and 40 minutes back home. It’s different to what I used to do.

Maple, gained 3.1 kg, 12 months

Social support (opportunity)

All women who had lost weight at 12 months described their partners as taking an active role in supporting them by sharing the same meals, offering practical support in relation to cooking, providing childcare or generally encouraging them to stay fit:

My partner is 100% supportive. He’ll eat whatever I cook. He’s very happy to follow the same sort of meal plan that I eat. He’s helped with the childcare. I wouldn’t have been able to do it [attend a fitness programme] without him. I couldn’t have done it without him. That’s in short, full stop.

Pine, lost 7.7 kg, 12 months

By contrast, women who gained weight described that their partners had been happy to eat what they were cooking but had not made significant changes to their dietary patterns:

Whatever I cook in the house is what everyone eats. My husband is happy with it, he can’t be bothered. Whatever mummy is eating is what everyone is eating.

Maple, gained 3.1 kg, 12 months

One woman felt that being back at work after birth made managing her weight more difficult:

I mean I work in a place where there is a culture that someone has to bring something every week like a cake or sweet and sometimes you really don’t want to have the cake but because someone made the effort of bringing that for us you don’t want to let anybody down, so you end up eating. I would say OK, just a little cake.

Ash, gained 3.5 kg, 12 months

Access

Access to additional weight management support was assessed at 6 and 12 months postnatally. A total of 83 women (85%) allocated to the intervention arm completed the relevant question in the questionnaire at 6 months, of whom 71 also did so at 12 months. A total of 75 women (79%) allocated to the control arm completed the question at 6 months, of whom 71 also did so at 12 months.

In total, 25 out of 83 women (30%) allocated to the intervention arm and 28 out of 75 women (37%) allocated to the control arm accessed additional support at 6 months postnatally. Similar rates were reported at 12 months postnatally (32% of women allocated to the intervention arm and 37% of women allocated to the control arm). Looking at both time points, a total of 37 women allocated to the intervention arm and 39 women allocated to the control arm accessed additional weight management support at either time point. Eleven women allocated to the intervention arm and 15 women allocated to the control arm accessed additional weight management support at both time points.

Timing of commencement

Women assigned to the control arm were asked when (if at all) they had accessed additional weight management support. Support most commonly commenced between 5 and 6 months postnatally (that is, for 40% of women who accessed support), with fewer women (only 20% of women who accessed support) reporting that support commenced earlier than 5 months postnatally. This suggests that women were ready to access weight management support slightly later than the time at which the intervention commenced.

Types of additional support and risk of contamination

Joining a gym was the most popular type of weight management support for participants in the intervention and control arms (30% and 50%, respectively; Table 18). In the intervention arm, using a fitness tracker such as a Fitbit^® (Fitbit, Inc., San Francisco, CA, USA) was the second most popular type of support (13%), whereas in the control arm weight management apps and Weight Watchers were each used by 15% of women who responded. In relation to risk of contamination, a total of five women allocated to the control arm joined Slimming World (three women chose to join when their baby was 6 months old and two women when their baby was 8–9 months old).

In the intervention arm there were participants using the Fitness Blender website (www.fitnessblender.com; accessed 3 January 2019), the Body Boss website (www.bodyboss.com; accessed 3 January 2019), the FatSecret^® app (FatSecret, Melbourne, VIC, Australia), the Samsung Health^® app (Samsung, Suwon, South Korea) and the 7 Minute Workout^® app (Johnson & Johnson, New Brunswick, NJ, USA). In the control arm one participant used the Sweat^® app (Sweat, Adelaide, SA, Australia), Weight Watchers app and MyFitnessPal app and one women used the Body Boss website.

The ‘other’ weight management support accessed by women is presented in Table 19.

Views on additional support

All women allocated to the intervention arm were positive about the additional measures they had chosen to help their weight loss, and several benefits were mentioned. Women commented that going to the gym improved their health and physical strength, and that ‘whilst my weight hasn’t gone down my measurements have’ (Poppy, two sessions, lost 7.0 kg, 12 months). It was noted by two women that the gym had the advantage of easy access and flexibility and one said that having a personal trainer gave her additional motivation. Two women mentioned that they use a gym and swimming pool at which there is a creche, which makes it easier for them to access it, with one finding that she enjoyed exercising and having that time away from her baby. Another found that attending yoga and Pilates classes were important for her core strength and pelvic floor muscles.

One woman found a Zumba class to be useful; another went to two drop-in exercise groups that were held when her daughter was at nursery, which made them easy to attend. However, there were also several comments about the convenience of personal exercise at home using online support: ‘easy to access, felt comfortable to exercise within my own space’ (Magnolia, lost 4.6 kg, 12 months); ‘as the onus is on me, I found that I made time to exercise. Committing to classes with two children is not ideal’ (Freesia, lost 2.6 kg, 12 months). There were also those who liked walking with their baby and found a fitness tracker useful to track how many steps they had done. Informal social support from friends and family was mentioned as very helpful by two women; one woman said that it also resulted in improved well-being and mental health.

In contrast to women allocated to the intervention arm, women allocated to the control arm presented mixed views about their experiences of finding ways to lose weight. Some women found useful and enjoyable ways to exercise individually, including walking to work, swimming, going to the gym or using YouTube (YouTube, LLC, San Bruno, CA, USA) videos. One woman described exercising at home with her husband as useful. One participant found Slimming World useful and easy to fit in to her lifestyle; another woman said that Weight Watchers was great for support, particularly the app she used to track what she ate (whereas one woman said she disliked the app and therefore stopped using Weight Watchers) and the motivation of attending weekly meetings:

I find going to Weight Watchers meetings really helpful as members support each other without judgement . . . However, after several good months it is easy to go back to bad habits. I am not sure it is a long-term solution as I am an emotional eater and this is frustrating.

Elm, lost 12.8 kg, 12 months

Others commented that they had seen the benefits of their activities, including weight loss, improvement in health and mood, strengthened core muscles, increased energy, and motivation to take better care of themselves. On the other hand, there was a group of participants who acknowledged that they found it difficult to continue with their activities, even though they saw the benefits:

[The Fitbit] has improved my health; however, I haven’t been constant.

Oak, gained 34.0 kg, 12 months

The inconsistency this participant described was primarily due to struggling to find time. One participant stopped due to ill health, and another was finding it expensive. One woman admitted that, despite some effort, she did not manage to lose weight: ‘the gym and step tracker helped me get my energy up but didn’t help with weight loss. Other than these, I just followed my own diet plan but still didn’t manage to lose weight’ (Poplar, lost 7.1 kg, 12 months).

Understanding of the purpose of the trial

One woman allocated to the control arm admitted understanding very little about the trial, but most other participants understood that it was about supporting post-partum women to lose weight. However, only a few women understood that it was specifically about trialling Slimming World as a postnatal weight management intervention, as Slimming World was not mentioned in trial recruitment literature in order to minimise risk of contamination. Several women mentioned improving nutrition, well-being and monitoring the psychological impact of weight loss as additional objectives.

Recruitment strategy

Participants thought that the trial recruitment process was straightforward and the timing was appropriate:

I think because in the context of when you are pregnant, and you know that you are going to have to have vaccines appointments, it’s just another interaction with the health service, so it was like your mind is already thinking along those things because you are really focused on those interactions. So it’s not like maybe another time of my life when I would never need to go to a doctor or nurse, but when you are pregnant you are in and out pretty much every month for something.

Willow, gained 1.0 kg, 12 months

One woman suggested that it would have been better to have personal contact with researchers than to find out about the trial through a leaflet given with other paperwork during a midwife appointment; another suggested recruiting at the health centre where women also have antenatal appointments as well as at the hospital.

Recruitment information

Women in both the intervention and control arms had mixed views about the information provided during recruitment to the trial. Some were satisfied with it, saying it was sufficient; others said that it did not provide enough information about the trial or the intervention. More specifically, women felt there could have been more information about the rationale for the intervention and about randomisation into trial arms. They also would have wanted to know that the intervention involved Slimming World specifically and was focused on diet and that they would be expected to attend meetings:

I thought it would be around eating as well as physical activity. I did maybe have that expectation, but obviously when I went to the session I realised that it was more around the diet and the food intake. They didn’t say you couldn’t exercise, obviously, but I think that would be something that you had to do on your own.

Iris, one session, gained 4.4 kg, 12 months

One woman also commented that she would have liked a more nuanced approach to weight loss and to have been asked what her barriers are and what would personally help her.

Time to decide to participate

Most women said that they had sufficient time to make their decision about participating, knew they could opt out and did not feel pressure to agree. However, one woman allocated to the intervention arm who did not attend any sessions said that she made the decision too quickly, without thinking about it, and in hindsight thinks that she would not have agreed to participate if she had properly considered how difficult it was going to be to attend a group with a baby and her toddler.

Lifestyle leaflet

Of the nine women allocated to the intervention arm interviewed, seven were asked specifically about the lifestyle leaflet component of the intervention (provided at the beginning of the trial). Most did not recall much about the lifestyle leaflet. Five women remembered the leaflet, three of whom remembered some of the content (mainly about breastfeeding) and thought that some of it was useful but that other aspects were not relevant to them. One of these five woman did not remember any of the content, and one said that she did not read the leaflet. Two of the seven women who were asked about the leaflet could not remember receiving it.

Randomisation

Many of the trial participants understood that they were randomly assigned to two arms to assess the impact of the intervention. Two women said that they did not understand it, one thought that there were three trial arms (although could not remember what they were) and one thought that participants were allocated according to questionnaire responses.

Allocation to the control arm

Several women said that they did not mind being allocated to the control arm because they did not feel they had missed out and that it did not change anything for them. They mentioned that being weighed by the midwives and completing the questionnaires was still a positive experience. Only one woman said that she was disappointed not to be selected for the intervention but felt that the questionnaires and interviews kept her focused on her weight loss.

Women allocated to the intervention arm had more varied views. Three said that they would still have been happy to take part in the trial had they been allocated to the control arm, which would have meant ‘not having the benefit of weight loss, but research in itself is a good thing’ (Hibiscus, 10 sessions, gained 0.3 kg, 12 months). One woman said that she would have been less interested in participating; another said that she would have felt disadvantaged and would have gone to Slimming World herself to try it out. One woman who did not understand randomisation said that she would have questioned not being allocated to participate in the intervention. Two women allocated to the control arm stated that they would not have wanted to participate in the intervention. One woman hoped when she first joined the trial that she would become part of a group with other new mums, which would help her to find a social environment, and she was not too concerned about the weight loss aspect. Another said that she would prioritise using the time for exercise:

Actually now thinking about it, I think I would have struggled to go to weekly meetings with my baby [because they would be too frequent]. I think I would have been more likely to have preferred using that time to go for a swim or go to exercise class than go to a meeting where I sit around talking about my weight.

Rowan, lost 1.0 kg, 12 months

Incentives

Seven women discussed the use of incentives (vouchers given to thank women for their time in participating). Four of these women, who were allocated to the intervention arm, thought that the shopping vouchers were a good incentive:

I mean, £10 is OK. That’s alright. It’s just to say a thank you to have that opportunity to get the time out of your busy schedule to help. So to me I think it’s not that bad. I appreciate that.

Marigold, lost 2.5 kg, 12 months

However, the three women from the control arm who spoke about the vouchers did not think they were a good enough incentive to convince them to participate if they did not want to do it anyway:

I mean, probably not in terms of my time and the effort and all of that kind of thing, probably not. If I wasn’t going to do the study that wouldn’t have made me do it.

Birch, gained 6.2 kg, 12 months

I know that £10 is standard for a questionnaire and that wasn’t motivation for me, but I don’t know if that is enough motivation for a busy mum . . . If you want the really wider demographic I think you might have to incentivise, think of another way to incentivise. Maybe baby products. I know you can’t do that, but I think it’s going to have to be something else because 12 sessions and going to meetings, that’s a lot of time commitment.

Rowan, lost 1.0 kg, 12 months

There was also one participant who said she gave her vouchers away as she does all her payments contactless on her phone or Apple Watch^® (Apple Inc., Cupertino, CA, USA), so for her a digital format voucher would have been better.

Choice of intervention

Two women allocated to the intervention arm spontaneously discussed the choice of intervention (i.e Slimming World) and were critical of the programme in relation to meeting the needs of women in the postnatal period. One thought that the Slimming World consultants needed to be trained to better deal with postnatal women, their dietary needs and mental health vulnerability. The other would have preferred a more holistic approach with more support:

It did feel very much as though it was a, not even a signpost, but an on-passing. So, ‘Go over there, have your intervention, see what Slimming World manage to do to you, for you, with you and then come back’. It didn’t feel particularly connected or joined up . . . It didn’t feel particularly structured or managed within that setting . . . It was just, ‘Right, you’ve got a baby and 3 months in get rid of the weight’. It felt a little disconnected.

Aster, six sessions, gained 2.8 kg, 12 months

Questionnaire

All women (from both arms) provided feedback on the questionnaires they were asked to complete. As reported in Chapter 5, there was a high completion rate of trial questionnaires and included measures. Some women liked the questionnaires and found them suitable for the trial; they specifically mentioned that they could answer about the food they had eaten, the pictures of the types of food were helpful, the length of the questionnaire was OK and they were glad that it took mental health into consideration. Two women allocated to the control arm said that responding about what they ate was helpful because it made them think more critically about it:

The questionnaires were really very good. How often do you eat starter? How often do you eat bread? How often do you eat salad? How often do you drink Coke? These are questions that I’ve never asked myself . . . But right now, with the study, I’ve been able to work things out myself. ‘OK, do I really need this much on a weekly basis? No.’ So that has really helped me a lot to be honest.

Maple, gained 3.1 kg, 12 months

However, there were many comments that suggested that the questionnaire could be improved. A number of women thought that it was too long (one needed two sittings to do it; another said that she would not have had the time to do it if she was not on maternity leave) and required a lot of mental effort. However, several women suggested that it would have been useful to add some more questions about exercise and well-being (one said that there was too much emphasis on food) and about factors that influence diet, for example support at home, birth experience and recovery, which would provide a more holistic picture:

It wasn’t really asking about health and well-being, it was asking very specifically about weight. I felt like the focus was on weight management and not on health and well-being. Maybe I had more to say about that.

Willow, gained 1 kg, 12 months

One woman allocated to the intervention arm felt that there was too little space for her comments on attending a commercial weight management group:

It was very quantitative and not much about whether I actually thought the whole programme was helpful or not. It was almost too objective . . . I actually felt that I had found the programme helpful and I had learnt things that would change my behaviour, but I just didn’t feel – the benefit that I thought I got from the intervention – I didn’t feel that was reflected in what I was answering . . . So there wasn’t really anything asking me on how good I thought the programme was or anything like that.

Hibiscus, 10 sessions, gained 0.3 kg, 12 months

Another woman said that she would have preferred to have a food diary for 2 weeks because she had no detailed memory of what she had eaten.

Many women mentioned that the questions were difficult to answer and that they needed help from the research midwives to know how to fill it in correctly for the following reasons:

• inconsistency in question design and therefore easy to make a mistake

• responses were not in a logical order from strongly agree to strongly disagree

• no dairy-free food options to tick, so not possible to properly enter dairy substitutes (mentioned by two women).

Finally, two women suggested that the questionnaire could have been made available in electronic format, which would have made it more efficient and convenient for them to fill in:

So first the administration of it being paper and pen was surprising to me in this day and age. Nearly any kind of market research is done on an iPad^® [Apple Inc., Cupertino, CA, USA] with a Google [Google Inc., Mountain View, CA, USA] document or something. It felt really archaic to be using paper and pen and then also I was a little, not that I would question anybody’s ability, but obviously there is room for error in that because it then has to be transcribed and re-entered into a system and so mistakes could be made or there’s just another point of potential human error that’s getting in the way.

Willow, gained 1 kg, 12 months

Contact with research midwives

Women were overall very satisfied with their contact with the research midwives, reporting that their visits provided additional reassurance in terms of their physical and mental health and that being visited and weighed was motivating. Some said that they had a choice of whether they had the meetings at hospital or at home and this was very helpful. Most felt that the meeting places were appropriate, with home visits being particularly complimented because this was felt to be very accommodating in a difficult period:

She was very nice about coming at a time that suited me. Actually the home visits are the things that I think were really agreeable. Not having to go to hospital because we’re all kind of trying to get back with our careers and stuff and I was working from home.

Heather, 12 sessions, lost 2.6 kg, 12 months

Several women mentioned that they were happy to go to hospital for their appointments as they added them on to antenatal appointments (baseline visit) or enjoyed having the excuse to go out without the baby and have time to themselves. However, two women said that they had to come to the hospital and that this was difficult for them; in one case this also made the meeting feel rushed.

Weighing frequency

Four women said that they were happy with how frequently they were weighed; however, one woman allocated to the control arm said that she could not make it in to be weighed a second time as she was working full time from 6 months postnatally onwards. Several women from both the intervention and control arms reported that they would have liked to have been weighed more often to help keep them on target or to reflect the more frequent significant fluctuations in weight that they experienced. Suggestions included that weighing could be carried out:

• monthly

• after baby weight checks at the clinic

• every 3 months

• between the two meetings with research midwives

• at the end of the 12-week programme.

More frequent weighing of the women allocated to the control arm could have resulted in an intervention effect because this is not current standard care.

The women allocated to the intervention arm did not seem to realise that the data on their weight from Slimming World would also be taken into account and so were concerned that the fact that they did lose weight when on the programme would be lost by the time they were weighed again for the trial:

So I felt that after the 12 weeks I had lost weight and then I put some of that back on by the time I got weighed again, which I suppose is what you are trying to see if the weight loss is maintained. But at the same time I think it would be useful to have more weighs in order to see what the pattern is.

Hibiscus, 10 sessions, gained 0.3 kg, 12 months

It was also commented that the weighing was not standardised and therefore not an accurate reflection of their weight:

I was weighed at completely different times of the day. I’d just had lunch before one, I hadn’t had lunch, it was sort of like nothing was on par, like one I had a heap of jewellery on and the other one I didn’t. It was a whole lot of things like if you were to normally measure weight you would try and standardise.

Pine, lost 7.7 kg, 12 months

Slimming World voucher at end of trial (control arm)

Women who were in the control arm of the trial were given a voucher to attend 12 Slimming World sessions over a 14-week period at the end of the trial. Some were positive about this and two women were considering attending. One woman said that it was a really good incentive to look forward to; however, she had already lost weight so did not need it any more. Three women said that they would not be able to make it to meetings because they had to go back to work and it would have been better to be able to use it during maternity leave. Two women said that they were not interested in attending as they did not like the sound of it:

And that’s not really my bag. Even the brand of the title, Slimming World, I just think, it’s not something that I would be interested in. I’ve joined a CrossFit^® [CrossFit, Inc., Santa Cruz, CA, USA] gym and they are kind of more my people . . . I know friends who have done it. For me it sounds like a fat-people version of AA [Alcoholics Anonymous] meetings. It just sounds sad, and the weigh-ins, and it just . . . Also there’s something very kind of middle-aged housewifey about it that just because I’m a mum does not mean that I am one of those types of mums.

Willow, gained 1 kg, 12 months

Suggestions for other forms of support

Two women (allocated to the intervention arm and control arm, respectively) offered suggestions for further support that would be useful to them: a creche that they could make use when undertaking fitness activities and having an NHS ‘centre’ where women could come for support with managing their diets.

Acceptability of intervention

• Of the 98 women assigned to the intervention arm, 47% attended at least one session, but only 19% attended the full Slimming World recommended programme of ≥ 10 sessions.

• Key barriers related to a lack of opportunity to attend. Women struggled to find the time, to fit attendance into their family routines or to find childcare when they felt it was inappropriate to attend with their baby/children. Motivation was also a factor, with some women either disagreeing with some aspects of the approach or not feeling that the groups provided a positive and supportive environment for them. There was also a number of women who were positive about the intervention but unable to attend many sessions owing to unforeseen circumstances.

• According to data from Slimming World, nine women (20% of women who attended at least one session) continued to attend sessions subsequent to the 12-week intervention package. A key barrier to attending sessions after the intervention ended for others was cost, especially where the timing coincided with the end of maternity-leave benefits.

Timing of commencement of intervention

• The majority of women (77%) who attended at least one session did so when their baby was aged ≥ 10 weeks. Feedback from women indicated that there were a significant proportion who felt that the commencement period was too early in the postnatal period and that it was difficult for them to find the time or energy to attend sessions and/or implement the proposed lifestyle changes.

• A total of 40% of women allocated to the control arm who accessed additional support did so in the first 5–6 months postnatally, with only 20% accessing it before 5 months postnatally. This suggests that women were ready to access weight management support slightly later than the time at which the intervention commenced.

Experience of attending commercial weight management groups sessions

• Five key themes emerged that mapped to various aspects of the COM-B framework: weight loss aspirations, beliefs and expectations, understanding and implementing the intervention, the social context, and accessibility of the intervention. Analysis compared women who completed the programme by attending ≥ 10 sessions and women who attended fewer sessions but lost weight with women who attended < 10 sessions, highlighting key differences in relation to these themes.

• Women who completed the programme by attending ≥ 10 sessions and women who attended < 10 sessions but lost weight were more likely to perceive themselves as overweight, understood why they gained weight and were more desirous to lose weight after birth. Their experiences of the programme and resulting weight loss helped to change pre-existing ideas about how to lose weight, particularly in relation to exercise. They understood the dietary plan, found it easy to follow, planned meals and used online resources to support their weight loss journeys. They considered the plan to be unrestrictive, sustainable and compatible with their postnatal lifestyle, leading to positive benefits for not only themselves but their babies and families. These views persisted when several were interviewed again at 12 months.

• Social context also played a positive role in facilitating women’s acceptance of the programme. Identifying with group members, forming social bonds at a time when women wanted to be out with their babies meeting others, and opportunities to share experiences in a safe environment were key. Women formed positive relationships with group consultants who provided personalised support in relation to breastfeeding needs, which enhanced their belief in the programme and therefore its acceptability. Women in this group were supported to lose weight by their partners, who helped on a practical level or were happy to share Slimming World-friendly meals. Although some women continued with Slimming World beyond the trial intervention period, cost was a key barrier for others. Several women suggested extending the intervention period. Women interviewed at 12 months were still adopting aspects of the programme, although one woman with some difficulty.

• By contrast, most women who attended between six and nine sessions (and gained weight) acknowledged that they were overweight and understood why, but viewed weight management after birth as less of a priority. Adapting to life with a baby was perceived as more important than attending a weight management programme. These women were more mistrustful of the intervention in part because they felt it failed to prioritise exercise in line with their pre-existing beliefs; they saw it as unsuitable for breastfeeding mothers and felt that there was a gap in their care, particularly in relation to postnatal mental well-being. Despite attending sessions, the women did not appear to understand the Slimming World programme and experienced difficulties making dietary adjustments. Weight gain and lack of perceived benefit may have affected regular attendance. The social context was not seen as positive. Irregular attendance may have been why women felt they did not bond with others in the group, leading to isolation, although several felt that they did not socially identify with others. Some women in this group also felt that their group consultants were unsympathetic to their postnatal circumstances and felt scrutinised or chastised for not following the plan or gaining weight. Social support was conceptualised as partners being happy to eat what was cooked by women; however, several women did not make major adaptations to their dietary habits and used only selected aspects of the Slimming World programme.

• There were some similarities between women who gained weight and attended between six and nine sessions and women who attended fewer than six sessions. Weight management was not a priority in either of these groups. For some women there was a degree of acceptance of their weight or they felt that they had not gained much during their pregnancies despite being classified as overweight at antenatal booking. Women in this group felt that the intervention failed to prioritise exercise, particularly when exercise was perceived as a main method of previous weight management attempts. Women in this group did not attend enough sessions to fully comprehend the principles of the Slimming World programme because most attended fewer than three sessions. Two women in this group experienced difficulties with joining Slimming World.

• All women allocated to the control group expressed aspirations for weight reduction and positive lifestyle changes after birth. Women allocated to the control group who lost weight acknowledged the importance of weight loss after birth primarily for health reasons, of belief in the link between weight gain and postnatal depression and of weight loss for future pregnancies. They had a good understanding of what constituted a healthy diet and actively engaged with weight loss strategies through diet and exercise. They were also supported or encouraged by their partners in their weight loss attempts. By contrast, women allocated to the control group who gained weight tended to perceive that they had not gained much weight during their pregnancies or were more accepting of their weight. Breastfeeding and the focus on general health and well-being, particularly postnatal mental health, were considered more of a priority and, for several women, exercise was seen as key to weight loss. Most women made changes to their dietary patterns but this did not have an impact on their weight. Social support was conceptualised as partners being happy to eat what was cooked by the women; however, several women did not make major adaptations to their dietary habits. The work environment added an additional barrier to weight management.

Variation in group attendance

• Most women attended only one group, with the same consultant, for the duration that they attended the weight management sessions, though some women appeared to move around groups and consultants.

• The majority of women attended sessions on weekdays, with an almost equal split between attendance in the morning and evening.

• Feedback indicates that some women may not have known about the opportunity to change groups/consultants when they were not satisfied with their initial choice.

Additional weight loss support

• Around one-third of women in both arms stated they had accessed additional support for weight management. The most popular type of support was joining a gym, followed by using a fitness tracker (intervention arm) and weight management apps and Weight Watchers (control arm).

• Women allocated to the control group were most likely to access support when their babies were aged 5–6 months (only 20% of women accessed support before then).

• The majority of women who sought additional support (in both the intervention and control arms) were still accessing it at the time of completing the questionnaire.

Contamination between trial arms

• There is a low risk of contamination between the arms of the trial: only five women from the control arm joined Slimming World independently (choosing to join 6–9 months postnatally).

Acceptability of trial processes and procedures

• Most women understood the trial design but some did not realise that the intervention involved attendance at Slimming World sessions, and some had expectations of a postnatal-specific intervention focused on more than solely weight management.

• Few women remembered the lifestyle leaflet or its contents.

• Completion of the questionnaire was felt to be useful and acceptable by some women, but others found it too long and complicated, or commented on the absence of questions on some aspects of lifestyle behaviours that they considered important. Despite this, there were generally high completion rates of questionnaire measures other than aspects of physical activity during pregnancy.

• The appointments with the research midwives were valued; women had no objection to being weighed and some asked for this to be more frequent.

Intervention, comparator and time horizon

The economic evaluation assessed the total incremental costs and QALY gains of usual postnatal care combined with enrolment in the weight management programme (intervention arm) versus usual care only (control arm). Incremental costs and QALYs were estimated for the 12-month follow-up, inclusive of service contacts and quality of life changes aggregated over the two 6-month periods following the baseline assessment at trial entry (36 weeks’ gestation).

No discounting of costs or QALY impacts was undertaken given that the evaluation period did not extend beyond 12 months. The economic analysis did not consider any longer-term programme impacts linked to future pregnancies and wider health risks to mother and infant. Issues concerning the modelling of these impacts in any future definitive trial are discussed in Modelling out-of-trial programme impacts.

Analytical approach

The economic evaluation was conducted on an ITT basis including only those randomised participants who were followed up at 12 months (n = 140). Stata version 15 was used for all reported analyses. Population mean differences in total cost and total QALYs at 12 months were estimated using ordinary least squares regression modelling including a binary ‘treatment’ explanatory variable indicating whether a participant was randomised to the intervention or control arm of the feasibility trial. Both models made baseline covariate adjustments for any differences between intervention and control participants in relation to age at trial registration, ethnicity (‘black’, ‘Asian’ and ‘other’; reference group ‘white’, plus three cases recorded of ‘unknown’ ethnicity), weight (in kilograms) at trial registration, obesity identified at the first antenatal visit (defined as a BMI score of ≥ 30 kg/m²; reference group, BMI score of < 30 kg/m²), number of children prior to the birth of the index infant, EQ-5D-5L community weighting scores at baseline and costs of service contacts reported at baseline with respect to the period of pregnancy prior to 36 weeks’ gestation (pregnancy- and non-pregnancy-related hospital and community health service contacts and costs of inpatient admissions were entered as separate covariates).

Missing data

Missing data on specific service items and baseline regression covariates (excluding missingness due to lack of follow-up at 12 months) reduced the size of the potential estimation sample (from 140 to 85) for the economic analysis. Where an admission to hospital and reason for admission were reported but data on number of days spent in hospital were missing (mother or infant), the average cost of a finished consultant episode (FCE) reported in NHS Reference Costs 2016/1749 for the currency code that best matched the reason for admission was applied. More generally, to avoid discarding relevant data collected from participants and to mitigate the risk of evaluating programme cost-differences on a biased subsample of those followed up to 12 months, missing data were handled using multiple imputations. This involved using the chained equations approach,90 applying an imputations model drawing on variation in observed baseline sociodemographic data and participant weight, costs and EQ-5D-5L quality of life data measured at follow-up. Imputations were made across each trial time point for all participants followed up at 12 months. This generated 20 simulated fully imputed data sets each containing 140 observations for all relevant cost, quality of life and baseline covariates, included in the regression modelling.

Sampling uncertainty

Sampling uncertainty was analysed using the distribution of bootstrapped replications of incremental cost and QALY outcome pairs. This uncertainty is presented in two stages: first, using a visual presentation of the simulated distribution of cost and QALY outcome combinations in the ‘cost effectiveness plane’ and then using this distribution to determine the proportion of simulations located in each cost-effectiveness ‘quadrant’; second, by identifying the proportion of simulated cost and QALY outcome pairs that result in a positive net QALY benefit (indicating that the programme is cost-effective) and repeating this for different assumed cost-effectiveness thresholds. This is used to trace a cost-effectiveness acceptability curve (CEAC),91 which plots the probability that the intervention is cost-effective at different assumed values for the cost-effectiveness threshold.

Additional sensitivity analysis

The appropriate level of the cost-effectiveness threshold for guiding decisions over the funding of new health-care programmes is uncertain. 92,93 Therefore, in addition to generating a CEAC, we also estimated the net programme QALY benefit across variable thresholds, some higher and some lower than the existing value adopted by NICE. The sensitivity of the main findings to the exclusion of high-cost outliers (e.g. owing to infrequent instances of unusually lengthy periods reportedly spent in hospital during the follow-up period by mother or infant) that could risk skewing findings in favour of either the weight management programme or usual care was also examined.

To explore the sensitivity of findings concerning costs and QALY outcomes to the inclusion of participants who were not successfully followed up at either 6 or 12 months, multiple imputations were used to impute missing cost and QALY data for trial dropouts. This test assumes that loss to follow-up is random (conditional on observable participant characteristics) and not related to either QALY outcomes or levels of health-care usage during the follow-up period.

Quality and completeness of economic data

There were no reported concerns from the research midwives that participants had difficulties responding to questions regarding contacts with community- or hospital-based services over the follow-up period. Analysis of the data collected to meet objective 5 broadly supported collection methods used as a feasible basis for evaluating intervention cost-effectiveness in a definitive trial. Tables 31–33 in Appendix 4 provide details on completeness of cost data against individual items of service use during pregnancy and the first and second 6-month periods. Self-reported data on contact with community-based health care from the 6- and 12-months interviews were complete. Only three cases had missing data from clinical case notes for costing labour and birth. In a small number of cases, data for costing A&E contacts at 6 and 12 months were missing (6% and 3% of participants interviewed, respectively) and for outpatient service contacts (3% of participants interviewed at each time point). Missing information was more apparent for costing inpatient admissions over follow-up, the proportion of participants with missing inpatient admission costs ranging between 4% (infant admissions at 12 months) and 11% (maternal admissions at 6 months) of those who were followed up. Unit cost data suitable for use in economic evaluation and for costing service use measured via the AD-SUS (including mode of birth) were universally available across all service items included in the scope of the economic analysis. Data on self-reported health status using the EQ-5D-5L measure (for QALY estimation) were provided by all women who completed follow-up. All component items of the EQ-5D-5L were fully completed by all women followed up at 6 and 12 months, enabling identification of the relevant EQ-5D-5L health state and its associated utility weighting.

Despite high levels of complete data for most service items, aggregation across individual items of cost led to missing data on total costs and missing paired data on total costs and QALYs for 33% of participants at 6 and 12 months. Missing cost data were successfully imputed using multiple imputation methods.

Reported service use by time period

Descriptive statistics on service contacts during pregnancy and at 6- and 12-month follow-up for women allocated to either arm are shown in Tables 31–33 in Appendix 4. All the information reported in these tables refer to the sample of women who had complete data on service contacts at each time point.

During pregnancy a wide range of services were reportedly used by women in both trial arms. Midwives, GPs, GP practice nurses and obstetricians were reported to be the most frequently contacted clinical professionals. As could be expected, antenatal visits for ultrasound scans and blood tests were almost universally reported across the trial sample. Hospital admissions during pregnancy were comparatively rare in both arms: 8% and 12% of the intervention and control arm samples, respectively, reported at least one admission to hospital.

The overwhelming majority of women (> 90%) continued to report having some level of contact with GPs and midwives at the 6-month follow-up, averaging between three and four contacts over 6 months for those who reported engaging with these services. Health visitors were also widely utilised at follow-up (> 90% of women). Reported use of midwives dropped off at 12 months, though GPs and health visitors continued to be widely utilised. Social workers, mental health professionals and employment and housing workers were contacted by only a minority of participants over the entire follow-up period.

Contact with an A&E department was reported by 26% and 30% of women at 6 and 12 months, respectively, with a slightly higher reported use in the intervention arm at both follow-up points. At 6 months, a higher percentage of women allocated to the intervention arm reported at least one outpatient visit relating to their infant’s health compared with women allocated to the control arm (28% vs. 17%, respectively), though this differential narrowed at 12 months. For those who reported outpatient contact for the infant, there was a reported average of two visits.

The percentage of women reporting any contact with outpatient services relating to their own health was higher in the control arm (23% at 6 and 12 months) than in the intervention arm (18% at 6 months and 16% at 12 months), with a mean of two to three visits for those reporting at least one contact at each of the follow-up interviews.

Health-related quality of life

Utility weights applicable to the health states described by participants using the EQ-5D-5L instrument are reported in Appendix 4, Tables 31–33. Mean health state utility scores improved in both groups from baseline over the follow-up period. Differences in mean utility scores between the intervention and control arms at 6 and 12 months were negligible.

Cost-effectiveness

Pre-intervention baseline costs accrued over pregnancy prior to 36 weeks’ gestation (inclusive of imputed values where data are missing) are presented in Appendix 4, Table 34. These estimates relate exclusively to participants included in the cost-effectiveness analysis who were followed up for 12 months. The costs of pregnancy- and non-pregnancy-related service contacts were broadly similar over this period for the intervention and control participants, though costs of hospital admissions were considerably higher for control arm participants.

The mean cost of service contacts for the intervention and control arms over the follow-up period (again for cases included in the economic evaluation) are presented in Table 20, along with mean cost differences adjusted for baseline covariates. Unadjusted differences (not presented) were similar and in practice adjustment for baseline covariates made little difference to the main cost-effectiveness findings.

The mean total cost per participant was £5718 in the intervention arm and £4922 in the control arm. The costs of birth contributed most to participant total costs over the follow-up period (a mean of £3000 in both groups) followed by community-based, A&E and outpatient service contacts (a mean of £884 and £758 in the intervention and control arms, respectively) and hospital admission costs. The cost of weight management programme enrolment (≈ £50 per randomised participant) was < 1% of the overall mean total cost for that group.

Total costs at 12 months (inclusive of the cost of commercial weight management group enrolment) were £741 higher in the intervention arm than in the control arm, adjusting for baseline covariates. In terms of types of service use, the biggest difference in mean costs arose in relation to infant hospital admissions at 6 months (a mean difference of £487 between intervention and control arms). The mean costs of contact with community-based and hospital outpatient services at 6 months were £106 higher in the intervention arm than in the control arm. The mean cost of delivery was also marginally higher for participants in the intervention arm. Mean cost differences across all service categories generally reduced between 6 and 12 months. At 12 months participants in the weight management group were estimated to have lived a single life-year of marginally higher quality after adjusting for health state utility scores measured at 6 and 12 months. This amounted to a mean difference of 0.01 QALYs at 12 months (see Table 20).

Table 20 also presents summary cost-effectiveness results based on the mean difference in total costs and QALYs at 12 months. The ICER for the intervention arm was estimated to be £74,100 per QALY gained. This translates into a 0.032 net QALY loss at 12 months.

Sampling uncertainty and sensitivity analysis

Figure 5 presents 100,000 simulated pairs of cost and outcome (QALY) differences in the cost-effectiveness plane along with a 95% confidence ellipse and the lower end of the threshold currently used by NICE to determine health programme value for money (£20,000 per QALY gain). A total of 63% of simulated pairs are in the north-east quadrant of the plane (the weight management programme has a higher cost and is more effective than usual care), with 26% located in the north-west quadrant (the weight management programme is cost-increasing and less effective) and a small percentage located in either the south-east or south-west quadrants (8% and 3%, respectively).

FIGURE 5.

Simulated cost and QALY pairings at 12-month follow-up.

Figure 6 translates this information into a CEAC.

FIGURE 6.

Cost-effectiveness acceptability curve.

At a cost-effectiveness threshold of £20,000 per QALY gained there is an estimated 0.17 chance that the intervention will be cost-effective at 12-month follow-up. At higher thresholds this probability increases, although it remains considerably below 0.5 even when assuming a maximum threshold of £60,000 per QALY gained. Figure 7 presents varying estimates of the net QALY benefit for different cost-effectiveness thresholds. Over the selected range of thresholds, the net QALY benefit of the programme as estimated at 12 months never exceeds 0.

FIGURE 7.

Net benefit of intervention at 12 months assuming varying cost-effectiveness thresholds.

When testing the sensitivity of these findings to the exclusion of a high-cost outlier (due to infant-related hospital admission costs in excess of £15,000 at 6 months) in the intervention arm, the mean difference in total cost between both trial arms fell from £741 to £625. This did not significantly affect the estimated ICER and the overall conclusions regarding programme cost-effectiveness at 12 months. Through use of multiple imputations, there was evidence that loss to follow-up could substantially bias estimated differences in mean costs: imputing for missing data conditional on observed values of covariates at baseline and over the follow-up period, the covariate-adjusted excess total cost per participant for the intervention compared with the control arm increased from £741 to in excess of £1500. This would serve to strengthen the conclusion that, over a 12-month follow-up period, the intervention would exceed the existing NICE maximum cost per QALY thresholds.

Search 1

(Economic AND model*) AND (obes* OR overweight OR ‘excess weight’ OR ‘gestational weight gain’ OR ‘gestational obesity’) AND method*

Search 2

((obes* OR overweight OR ‘excess weight’ OR ‘gestational weight gain’ OR ‘gestational obesity’) AND (pregnancy OR birth OR natal OR partum OR intrapartum OR interpregnancy) AND (‘weight loss’ OR diet* OR ‘weight reduction’)) AND human AND (econom* OR effect OR cost OR qaly OR ‘quality of life’ OR qol OR eval*)

Search 3

(obes* OR overweight OR ‘excess weight’ OR ‘gestational weight gain’ OR ‘gestational obesity’) AND (pregnancy OR birth OR natal OR partum OR intrapartum OR interpregnancy) AND (‘long term’ OR lifetime OR ‘life course’) AND human AND outcome* AND (mother OR baby OR child OR infant)

Inclusion criteria (common to all searches):

• English language

• human subjects only (no animal studies)

• journal articles only (no conference abstracts)

• published any time

• Organisation for Economic Co-operation and Development (OECD) countries.

Additional inclusion criteria for search 1 (PICO):

• problem – economic modelling

• intervention – obesity model

• comparison – other methodologies/conventional modelling approaches

• outcomes – modelling methodology.

Additional inclusion criteria for search 2 (PICO): problem/population, intervention, comparison, outcome

• population – women who experience EGWG or who are obese at conception

• intervention – any intervention to reduce bodyweight after pregnancy or between pregnancies

• comparison – any other intervention or usual care/natural weight loss

• outcomes – effect sizes, costs, quality of life measures.

Additional inclusion criteria for search 3 (PICOS):

• population – women who experience EGWG or who are obese at conception and their children

• intervention – any intervention

• comparison – any comparison

• outcomes – long-term outcomes in mother and child as a result of excessive weight gain

• study design – RCT, longitudinal study.

Epidemiological evidence

Because the purpose of this review was to identify key considerations for future modelling work, it was necessary to consider how the epidemiological evidence would fit into the outcomes observed as part of any definitive trial. A key piece of information is the reference measure, which establishes a link between a trial’s short- to medium-term outcomes and longer-term impacts, whether or not there is a dose–response relationship for a given effect size and how it is quantified. This might be usefully expressed as a RR, hazard ratio or odds ratio for each relationship. Existing models (such as those that underpin NICE guidance)94,95 have established evidence for general population risks of diseases stemming from long-term obesity, but these rarely have subgroups other than by sex. The aim here was to gain a general insight into what should be, on the strength of available evidence, the key adverse health outcomes (for woman or infant) in an economic model of pregnancy-related weight management.

We identified 39 studies96–134 that contained potentially relevant estimates of costs of achieving behavioural change and resource impacts linked to relevant long-term health outcomes. Fourteen studies118,135–147 reported health state utility scores relating to maternal outcomes, obesity and long-term conditions associated with weight gain.

The primary health risks for women linked to pregnancy weight gain reported in the reviewed studies were type 2 diabetes and cardiovascular disease. The evidence underpinning these comes mostly from large epidemiological registry reviews and regression models of large data sets, for example Eriksson et al. 105 There was limited evidence of elevated risks of cancer-related mortality and morbidity, though the authors warn that these effects are of only borderline statistical significance. The Eriksson et al. 105 study sample was drawn from the Finnish population; therefore, pregnancy-specific risks specific to the UK population were not addressed.

There was no evidence of a significant differential between elevated health risks linked to general weight gain and elevated health risks linked to weight gain during pregnancy. Therefore, for modelling purposes, general hazard ratios for the UK female population already used in existing general obesity impact modelling could be used for modelling long-term benefits of improved postnatal weight management.

In terms of subsequent pregnancy-related outcomes, the evidence points to an increased risk of higher maternal BMI scores on pre-eclampsia, gestational diabetes, caesarean section and other birth complications,133,134 and there is evidence that long-term weight retention in the postnatal period leads to negative outcomes in subsequent pregnancies. 7,8

The rapid evidence review did not identify a substantive pool of evidence for quantifying the link between pregnancy-related weight gain and risk to long-term infant development, suggesting that modelling these impacts will carry considerable uncertainty.

Economic end points

Estimates of long-term costs and QALY impacts attached to adverse health events of relevance to pregnancy-related weight management are well established in the UK and have been used in models underpinning NICE guidance (specifically guidelines PH25,134 PH15148 and CG181149 on cardiovascular disease, PH35 on diabetes,150 and PH5394 and NG795 on long-term models of weight loss). This review identified eight economic studies containing relevant evidence on economic points that could support modelling of long-term benefits. These included studies evaluating pharmacological approaches to delivering weight loss in patients with obese BMI scores135,151,152 and surgical interventions, including bariatric surgery. 136 The review also identified four epidemiological studies containing relevant economic data: two studies undertaken to estimate the burden of obesity118,153 and two studies that estimated the life-time burden of childhood obesity. 124,154

Programme clinical effectiveness and cost-effectiveness

We included 43 studies that estimated the clinical effectiveness and/or cost-effectiveness of behavioural programmes. 7,10,13,137,138,155–193 The primary outcome of the behavioural interventions identified in the review concerned their clinical effectiveness at reducing women’s weight in the first year post pregnancy. Most interventions identified used the absolute number of kilograms lost as the primary outcome of interest. In two studies, weight outcomes were also converted into a binary outcome identifying achievement of clinically significant weight loss postnatally, of which returning to within 5% of the woman’s pre-pregnancy weight was the most consistently used measure. 187,188 Significant 1-year outcomes in weight loss terms were found in only two studies. 189,190 One study191 was designed to capture adverse pregnancy events from a weight management programme and did not report weight loss outcomes; another study192 captured the impact of exercise interventions on other measures of cardiac and endocrine health through blood tests and waist measurement. Several other papers168,171,175,179,184,187 reported on initial acceptance of ideas or settings for interventions or provided outcomes for children rather than mothers.

Overall, the long-term clinical effectiveness and cost-effectiveness of behavioural programmes beyond the period of a clinical trial is uncertain. One paper193 offered evidence on weight maintenance over a 2-year follow-up period, with the remaining papers having follow-up periods no longer than 1 year. Evidence on longevity of effect, which is a crucial parameter for determining long-term economic outcomes, is lacking.

Existing economic models

The final search strand identified papers that would help in informing the methodological approach to modelling long-term impacts. A total of 34 papers139–147,194–219 were included. This search was deliberately broader in scope to try to capture the best practice for modelling obesity in general, not just in the postnatal population.

A total of 38% of the studies were based in the USA, 18% in the UK and 6% in the Netherlands. Canada,99,198 Germany,102,124 Ireland,115,135 Italy125,145 and Mexico195,214 each had two studies, and Australia,103 Belgium,152 Denmark,116 France,105 Greece,200 Malta,197 Portugal201 and Switzerland202 each had one study. One study98 used United Nations data to estimate worldwide impacts of obesity. The economic perspective of the studies was extracted, with 63% of papers taking a health service perspective and 37% taking a broader societal perspective.

Because this was a methodological search, there was a wider range of topics than only behavioural interventions. A total of 34% of models concerned modelling baseline or long-term epidemiological impacts of obesity without focusing on a specific intervention. Behavioural interventions were the second largest group, at 28%. A total of 22% modelled surgical interventions, one of which was to determine whether or not bariatric surgery improves pregnancy outcomes. A total of 10% modelled pharmacological interventions and 6% examined the population-level impact of policy interventions, including taxes on the sugar or fat content of food products. None of the papers covered by the review included a model dealing specifically with post-partum weight loss. One-fifth of the included studies covered only methodological approaches or were systematic reviews of A total of 32% of models used a microsimulation approach to model intervention impacts, 30% followed a Markov state-transition structure, 19% were based on regression analysis and 19% used a decision analytic approach.

Reviews of existing obesity models found that sophisticated long-term models are still rare, with modellers adopting strong assumptions about sustained duration of effect of interventions. 215–217

We identified one study218 that looked to capture the nuances of weight loss and regain in weight in key subgroups of people with obesity. This model used microsimulation methods to estimate four weight loss and gain scenarios over 4 years based on longitudinal data from a long-running diabetes study. This type of modelling structure could prove informative for extrapolating the impact of weight management programmes observed in any definitive trial given uncertainty regarding sustainability programme effects.

One review of the wider obesity modelling literature214 was identified and its key recommendations included the need to capture cardiovascular disease as a key consequence of obesity and to use modelling approaches geared to capturing the impacts of varying time to disease onset as well as disease severity and time to death. Microsimulation approaches, such as individual patient simulation or a Markov model, provide methodological frameworks for achieving this. Griffiths et al. 215 also recommend that implementation of interventions be included in the modelling, because they found that the scale-up and roll-out of small-scale interventions to population level is ‘controversial’ in the plausibility of the effects holding at population level.

Epidemiological evidence

• The review found strong evidence for the inclusion of pre-eclampsia, gestational diabetes, cardiovascular disease and type 2 diabetes in a future economic model, with relative risks available for these diseases already established in the literature.

• The evidence linking maternal outcomes to childhood obesity is not well established and linking a mother’s weight loss post partum to her child’s adult outcomes would require strong assumptions. Therefore, a childhood-to-adulthood model for this intervention would not be recommended.

Economic end points

• The economic impacts of cardiovascular disease, type 2 diabetes and pregnancy complications are well established in the UK through NICE guidance and these should be used to inform the future model.

• Relevant costs and health state utility/QALY estimates are available and widely documented in the literature for key clinical end points.

Programme clinical effectiveness and cost-effectiveness

• Estimation of the clinical effectiveness and cost-effectiveness of weight management interventions has been largely restricted to periods of < 1 year. Evidence on long-term sustainability of any weight loss over longer periods is generally lacking, which poses challenges for modelling extrapolated intervention impacts beyond time periods that could be feasibly built into a main trial.

Existing economic models

• The economic literature included did not contain an economic model specifically dealing with the impacts of postnatal weight management. However, many of the modelling studies included, including those that examined the impact of obesity, offer insight into appropriate methodological approaches to take when modelling impacts arising from improvement in pregnancy-related weight management and in some cases sources of evidence and data (e.g. relating to economic end points).

Objective 1

The sample size required 130 women to be followed up for 12 months postnatally. To reflect a 30% drop-out rate, because women are particularly mobile and tend to change addresses in the weeks and months after giving birth, we aimed to recruit a total of 190 women over a 6-month period. We recruited 193 women over 8 months; additional time was taken because of a revision to the trial protocol regarding recruitment of women who had normal antenatal booking BMI scores and developed EGWG. Contacting women initially by letter and asking them to opt out if they were not interested appeared to work well and meant that women could make their own decision about whether or not to participate. Of all women who received an invitation letter, only two contacted the team to express concerns about receiving a letter, indicating that women were generally happy to be contacted by a named midwife who could explain the trial further for those who were interested.

The initial plan (to advertise at the trial site to ask women who thought that they might have gained more weight than needed to contact the research midwives) only attracted a handful of women, none of whom met IoM criteria for gestational weight gain at 36 weeks. 5 A decision was made, supported by the TSC, to revise the recruitment approach and reflect the same approach used for women with overweight and obese BMI scores at antenatal booking, namely initially contacting women by letter to ask if they would be interested in joining the trial. We extended recruitment duration by 2 months to accommodate changes to the protocol and time to obtain REC/HRA approvals. Despite changes, we recruited only four women with EGWG. Although more women came forward to be weighed, they did not meet IoM criteria.

An important barrier to recruitment of these women is that women are not routinely weighed at each antenatal appointment in line with current NICE guidance,6 presenting a challenge with respect to identifying women with a normal BMI score at antenatal booking who develop EGWG. This is a population of women who give cause for concern15 because those who develop EGWG and do not lose the weight postnatally may gain more weight prior to their next pregnancy, increasing their risk of adverse outcomes such as gestational diabetes. 219 Alternative approaches to identify and support women with EGWG are needed, and, if weight management interventions are to target these women in the postnatal period, all women attending trial units who commence pregnancy with a normal BMI score may need to be weighed at each appointment or at pre-defined times during pregnancy.

Achieving high follow-up rates at 9 and 12 months in both groups was reassuring and reflected use of several approaches to follow-up, including face-to-face appointments with the research midwives at the trial unit, or in the woman’s homes, at 6 and 12 months, which women seemed to value. That so many women allocated to the control arm, and allocated to the intervention arm but who did not attend any sessions, completed follow-up was encouraging.

An important part of the process evaluation was to explore the acceptability of the intervention and trial procedures from interviews with women allocated to the intervention on completion of the commercial weight management programme and with women from both trial arms at 12 months postnatally to explore experiences of participating in the trial. We recruited as planned at both time points, selecting women for interview to reflect diversity of weight change and ethnicity and, in the intervention arm, high or low uptake or non-attendance of sessions.

Objective 2

We aimed to estimate the effect size for a probable primary trial outcome in a future definitive trial, namely difference between the trial groups in weight at 12 months postnatally, expressed as percentage weight change from antenatal booking weight. This was an outcome that the Expert PPI Group supported because longer-term weight management benefits were considered to be important. Analysis of weight data showed a modest benefit in relation to weight change at 12 months postnatally among women in the intervention arm, which was not apparent at 6 months. This is likely to be because some women allocated to the intervention arm (i.e. those who started the programme nearer to 16 weeks postnatally) were still accessing the Slimming World programme at 6 months postnatally. It may also reflect the need for women who had completed their weight management sessions to fully adapt the programme into their daily lives beyond the initial intervention period. Nevertheless, the findings confirmed the initial plan to include a primary end point at 12 months. Per-protocol analysis suggested that the intervention had a ‘dose effect’ associated with engaging with the Slimming World programme, with the greatest benefit seen among women who attended ≥ 10 sessions.

Objective 3

Supporting an individual’s lifestyle behaviour change is complex, even more so when women who have high BMI scores at antenatal booking may have multiple poor health-related behaviours, for example smoking tobacco, not taking physical exercise and not commencing breastfeeding. 70,220 When planning the feasibility trial, we had to consider if evidence-based information on positive lifestyle behaviour offered to women in the intervention arm, together with access to commercial weight management sessions postnatally, could support women to alter their lifestyle behaviours in the shorter- and longer-term, and, if so, the extent to which this was likely to occur if measures selected to assess ‘change’ were completed by women, and the most appropriate measures to use in a future definitive trial. Few of the potential measures available had been previously validated in postnatal populations because they were developed for general population use, and, for other specific areas of interest, questions were developed by the trial team because no relevant measures were identified.

Assessment of secondary outcomes showed minimal differences between trial arms. There could be several reasons for this, including that motivation for joining the trial was similar across trial arms because women who agreed to participate were interested in their weight after having a baby. It was also reassuring that a high proportion of women in both trial arms completed follow-up questionnaires, with few data missing from included measures. When data items were missing, for example on extent of physical activity at baseline, it is understandable that women at 36 weeks’ gestation may have considered this to be not relevant to their lives at the time.

Women allocated to the intervention arm were more likely to have an EPDS score of ≥ 12 at 6 months, indicating that women were more likely to be at risk of developing postnatal depression. 37 Reasons for this are difficult to speculate but could reflect baseline differences in mode of birth in the sample or that women attending weight management groups felt more anxious or concerned about their health. However, lack of difference in other outcomes compared with the control arm including body image, maternal esteem and infant sleep patterns, which could be associated with potential mental health issues,66,73 does not suggest that women were affected in other ways by the intervention. Lack of statistical power may also mean that any differences that may have been present were not detected, and that measures were not analysed per protocol.

Lifestyle and behavioural outcomes we were keen to explore in terms of whether or not the intervention was likely to make a difference included women’s dietary intake. As described in Chapter 5, the team had several discussions when designing the trial as to an appropriate food frequency questionnaire, eventually selecting DINE, which had been used previously in studies that included women of reproductive age. 35 Measures of fat, fibre and unsaturated fat showed no change between groups from baseline to 6 or 12 months, with potential reasons for this identified from the process evaluation (see Objective 4), including that the intervention was potentially not long enough to allow changes in dietary and other lifestyle behaviours to occur. If undertaking a definitive trial, it would be useful to consider further what data on diet are important to capture (i.e. patterns of healthy eating or nutritional content of dietary intake) to inform what may be a more appropriate measure. Measures used should also reflect the dietary content of the intervention of interest to capture any change in intake over time, because items included in an existing measure may not reflect dietary advice offered as part of the weight management intervention. Of note (and an unexpected finding) was that a few women described that completing the DINE questions influenced them to alter their diet as it included items that made them realise how much ‘bad’ food they were eating.

That there were no differences in important lifestyle outcomes including breastfeeding and tobacco smoking between trial arms probably reflects more positive lifestyle behaviour in the diverse trial population and population of women who uses the local maternity services generally, as well as lack of statistical power. The inner-city area from which women were recruited has traditionally had high levels of breastfeeding uptake and duration (in 2016, 91% of women in the local area commenced breastfeeding),221 which may be attributed to the ethnic mix of women in the area. Similarly, few women living in the local maternity catchment area acknowledged that they smoked in pregnancy, although slightly more women in the intervention arm reported that they had never smoked. The recent introduction of carbon monoxide monitoring at the trial site could provide data of greater accuracy on the number of women who smoke tobacco in pregnancy in a definitive trial. Fewer women allocated to the intervention arm than allocated to the control arm reported drinking any alcohol at 6 and 12 months, with a significant difference at 6 months, but low overall rates of alcohol use probably also reflects the profile of the local population of women.

Objective 4

An important aspect of the feasibility trial was to assess the acceptability of the intervention and trial procedures. This aspect of the evaluation was addressed using mixed methods, including questionnaire data, semistructured interviews with women and data provided by Slimming World. Findings highlighted the complex relationship between the intervention and some outcomes of interest.

Of the women allocated to the intervention, few women could recall the content of the lifestyle information leaflet despite this being offered to all women allocated to the intervention arm at trial recruitment. The information reflected current NICE guidance on a number of lifestyle behaviours (breastfeeding, keeping active, healthy eating and stopping smoking),6,33 which were assessed in included measures. The leaflet was developed with the support of the Expert PPI Group, but, perhaps not unsurprisingly, for women in late pregnancy and then facing the pressures of new parenthood it was unlikely that reading and remembering the content of the leaflet was a priority. For a definitive trial, the inclusion of additional information alongside a weight management support intervention would have to be considered further, as would a more user-friendly and accessible format of presentation, for example an online cartoon, blogs, infographics or social media.

All women allocated to the intervention could access a Slimming World group of their choosing at any time between 8 and 16 weeks postnatally. Just under half of the women (46; 47%) attended at least one session, most doing so when their infant was aged ≥ 10 weeks, with several key barriers to not attending sessions identified, including lack of opportunity, difficulty organising child care, timing of sessions and family illness. It would have been encouraging to have had greater uptake of weight management sessions among the intervention arm, but, given that the trial sample included women with a range of child-care and other family responsibilities who may not have encountered a similar intervention before and who live in one of the most socioeconomically deprived boroughs in England, that just under half attended one session could be viewed as positive. Analysis of the women’s experiences of attending weight management groups using the COM-B framework46 pointed to several important themes and to issues concerning why some women completed the full ‘dose’ of ≥ 10 sessions, why some attended < 10 sessions and why some did not attend any sessions. Of note is that the most common reason for not attending sessions was a lack of opportunity, although a small number of women had misbeliefs about Slimming World.

An important finding was that, based on per-protocol analysis, women who attended ≥ 10 sessions had greater weight loss at 12 months postnatally than women who attended nine or fewer sessions, did not attend any sessions or were allocated to the control arm. Of interest was evidence of a dose–response effect, which reflected that the more weight management sessions attended, the more women could adapt the programme to their daily lives. The process evaluation findings provided very useful data on how we could potentially increase uptake of sessions in a future trial of postnatal weight management support. This includes the need to offer commencement later in the postnatal period, because some women found that the time and energy required to attend a weekly session was too much and too soon after giving birth. Women expressed the potential importance of later commencement of the sessions and some considered that having only 12 sessions as part of the Slimming World programme was not long enough because they had only just got to grips with how to adapt the programme to their daily lives. Feedback also reflected the need to ensure that women were aware that they could change groups if they were not satisfied with their initial choice. Women wanted more information on trial allocation on what the Slimming World programme included, such as the role of exercise and how well the plan suited the needs of breastfeeding women.

Around one-third of women in both trial arms accessed additional support for weight management, most of whom joined a gym. There was low risk of contamination between trial arms, an outcome of particular importance to inform a definitive trial. From the 12-month interviews it was apparent that most women in both trial arms understood the aims of the trial, what being randomised meant and found recruitment approaches to be straightforward. Women generally found the trial questionnaires and included measures easy to complete and enjoyed having longer-term contact with research midwives.

Objective 5

It was feasible to generate economic data to inform a larger definitive trial using a combination of participant self-reported information and maternity records. There were no reported problems with the acceptability of questions regarding service contacts from the women and self-reported data on use of community-based services during and after pregnancy was largely complete. However, resource use information was less complete for more costly types of hospital-based service contacts, including hospital admissions for mothers and infants. A future trial should consider use of patient records data to reduce missing data for these types of service contacts.

The economic analysis highlighted a wide range of services used by women and for their infants during the 12 months after birth. Women allocated to the intervention arm had on average a higher service use cost over the first 12 months following birth than women allocated to the control arm. The average QALY lived over the follow-up period was only marginally higher for women allocated to the intervention. Combined findings suggested that the extra cost of the programme per QALY gained would be in excess of the cost threshold normally used by NICE to assess programme value for money. However, as many potential beneficial impacts of effective weight management could not be captured directly in this type of trial-based analysis (e.g. life-years gained and costs to the NHS avoided through prevention of disease linked to obesity and avoidance of complications in subsequent pregnancies), economic modelling of wider out-of-trial impacts should be included in any future definitive trial.

This was supported by the findings of the rapid review, which demonstrated that modelling of longer-term impacts linked to clinically significant short-term weight loss observed in a definitive trial would be feasible. The review identified a sufficient evidential basis for modelling out-of-trial impacts of effective weight management. The review findings suggest that priority should be given to modelling the impacts of health outcomes linked to pregnancy-related obesity and weight gain with the strongest supporting evidence, including cardiovascular disease, type 2 diabetes, gestational diabetes and health complications arising in subsequent pregnancies.

Objective 6

The criteria to proceed to a definitive trial (objective 6) were met. It was feasible to recruit and retain women with overweight or obese BMI scores at antenatal booking to a trial of a commercial weight management group plus standard care compared with standard care only after birth. Approaches to recruit women with normal BMI scores who develop EGWG and to optimise the benefit of the intervention based on the process evaluation findings need to be further considered.

Inclusion of a nested mixed-methods process evaluation enabled the integration of data from all sources to provide an in-depth explanation regarding how, why and if the intervention had potential to ‘work’ in a future definitive trial. As described with respect to objective 2, quantitative data analysis suggested a modest benefit in relation to weight change for women allocated to the intervention arm compared with women allocated to the control arm at 12 months postnatally, but little difference in relation to other lifestyle and health outcome measures of interest. To provide further context for the weight outcome finding, it was examined in conjunction with the qualitative data and per-protocol analysis, which showed that the intervention had a ‘dose effect’: greater weight loss was reported among women who attended ≥ 10 sessions than among women who attended < 10 sessions or were allocated to the control arm. The integration of findings provided greater understanding of factors that influenced women to receive the full ‘dose’ or not, which will inform a future trial.

Some disparity in trial findings was as a consequence of quantitative and qualitative methods being fundamentally different in terms of aims and questions, meaning that ‘measures’ were not matched. ITT analysis necessarily treated every intervention woman the same, regardless of whether she received the intervention as intended, and treated every weight management group and session as the same. However, the qualitative data showed a dose–response effect and that group identity and consultant factors influenced how the intervention was experienced by women.

The process evaluation identified key explanatory factors as to why some women attended ≥ 10 sessions, including their motivations to participate in the trial and to lose weight, their beliefs about the intervention, their ability to follow the Slimming World programme, opportunity factors and the support they had for participation. Integration of findings will inform decisions about a future definitive trial, in line with objective 6, with several clear recommendations emergent from the process evaluation that could optimise the potency of the intervention in a future definitive trial.

Independent members

Professor Rona McCandlish (chairperson).

Professor Hora Soltani.

Dr Nicola Heslehurst.

Mrs Samantha Hainsworth (PPI representative).

Dr Richard Hooper.