Systematic reviews and integrated report on the quantitative, qualitative and economic evidence base for the management of obesity in men

Development of a thematic index

To develop a thematic index, one researcher (DA) coded the findings reported in the included qualitative studies linked to interventions line by line for content and meaning and categorised these according to whether they corresponded to the a priori themes or whether they appeared to represent emergent themes unconnected to the a priori themes.

After undertaking this process for all studies linked to interventions the coding was cross-checked by another researcher (FD). The data then underwent a second importing process within NVivo 9 into framework matrices for comparison of the a priori and emergent themes for effective and non-effective interventions. Framework matrices were used to facilitate the use of the constant comparative method to search for patterns and relationships and assist with developing theory. All qualitative researchers (DA, FD, PH and EvT) then developed the descriptive thematic index over a series of meetings, with a tree structure of themes and subthemes to enable us to remain close to the reported study findings. The thematic framework was discussed at a meeting with the Men’s Health Forum representatives to ascertain service users’ perspectives. The qualitative researchers decided that one of the 10 a priori themes (‘Are programmes deliberately involving partners/families more effective?’) was not supported by the data and it was rejected. To develop a thematic index, one researcher (DA) coded the findings reported in the qualitative studies line by line.

The development of interpretive themes

We then generated a more refined set of interpretive themes from the a priori and emergent themes for the effective management of obesity in men and the barriers to and facilitators of engaging in weight management programmes. In meta-ethnography these are described as ‘third-order interpretations’. 81 Following the completion of the analytical cycle for studies linked to interventions, the theoretical studies not linked to interventions were then read to ascertain whether they provided disconfirming evidence or added any new perspectives and all relevant data from these studies were extracted. This process was conducted to test the robustness of the synthesis and has been recommended in previous narrative synthesis methods literature. 82

The final stage of the analysis involved integrating the qualitative findings with findings from the quantitative reviews. This was achieved through a process of in-depth reading of each results chapter by all members of the research group to identify where qualitative findings were supported or refuted by quantitative findings. The supporting or disconfirming quantitative data were then integrated into the findings.

Researcher perspectives

It is important to be aware of the researchers’ backgrounds and associated perspectives when interpreting our findings. DA is a health services researcher with a background in sociology and mixed-methods research methodologies. FD is a public health researcher with a background in health promotion and nursing, with an interest in health inequalities and the social determinants of health outcomes and behaviours. EvT is a medical sociologist with a background in qualitative and mixed-methods research and an interest in public health in general and health promotion in particular. PH is an academic general practitioner (GP) with expertise in qualitative and mixed-methods research, particularly when applied to RCTs of complex interventions. None of the qualitative researchers can be considered obese. Throughout the study reflexivity took place through weekly research team discussions until a consensus was reached.

Quantitative outcomes

All studies reported either baseline and end weights or changes in weight. Four studies reported change in BMI. 84,85,89,92 Three studies reported waist circumference87,89,90 and one reported waist-to-hip ratio. 85

For cardiovascular risk factors, only two studies reported total cholesterol,85,91 three studies reported HDL cholesterol,84,85,91 one study reported LDL cholesterol,84 two studies reported triglycerides85,91 and three studies reported systolic and diastolic blood pressure. 84,85,89 Systolic and diastolic blood pressure data for the trial by Wood and colleagues93 were reported in a linked publication. 115 The trials by Borg and colleagues84 and Esposito and colleagues85 reported results for fasting plasma glucose.

Only two trials reported a male-specific disease outcome: erectile function following weight loss. 85,87 One trial92 reported adverse events and one trial84 reported lack of adverse events. The trials did not report quality of life, HbA_1c levels or any economic outcomes.

Risk of bias

The risk of bias assessment for the individual trials is shown in Appendix 8 (see Table 56 ). Figure 2 summarises the assessment.

FIGURE 2.

Summary of risk of bias assessment of trials included in the systematic review of men-only RCTs.

It was unclear whether random sequence generation was adequate in all but three studies,89,90,93 which were judged to have adequately randomised participants. Five studies84,85,90,91,93 were judged to have successfully concealed allocation, whereas this was unclear in the remaining trials. No trial was able to carry out blinding of participants and the majority did not blind health-care providers, although this item was uncertain in the trial by Benassi-Evans and colleagues83. This is because blinding of participants and health-care providers would have largely been impossible because of the nature of the interventions considered by the trials. Four trials did blind outcome assessors,85,88 ^– 90 whereas the trial conducted by Borg and colleagues84 did not. It was unclear whether outcome assessors were blinded in the remaining trials.

The majority of the trials (10/11, 90.9%) were judged to have treated participants similarly apart from the given intervention in each arm of the trial, with only the trial conducted by Esposito and colleagues85 causing uncertainty for this item. Over half of the trials were judged to be at low risk for incomplete outcome data,84 ^– 87,89,92 with this item being unclear in the other trials. 83,88,90,91,93 Only four trials85,86,89,90 carried out intention-to-treat analysis with the majority (6/11, 54.5%) of the trials analysing data for trial completers only. It was unclear whether or not an intention-to-treat analysis was carried out in the trial conducted by Benassi-Evans and colleagues. 83

The trial by Khoo and colleagues87 was judged to be at high risk of selective reporting as the authors did not report HbA_1c outcomes. This was unclear for the SHED-IT trial89 as outcomes considered in the trial protocol were not reported in the 6- and 12-month reports [e.g. Short Form questionnaire-12 items (SF-12) and sexual function data]. All other trials were judged to be at low risk for this item.

It was unclear for four studies83,84,91,93 whether or not the trials were at risk of any other bias. Five trials85,87,88,90,92 were judged to be at low risk for this item, with the trial by Jeffery and colleagues86 considered to be at high risk as the participants paying larger monetary contracts in both the individual arm and the group arm had a higher mean baseline weight than the participants paying smaller deposits. The authors acknowledged that randomisation may have failed to equally distribute participants with respect to their weight. The authors also experienced difficulty recruiting people to this trial, as only 50% of those eligible signed monetary contracts. The SHED-IT trial89 was also judged to be at high risk of other bias because of the limited generalisability of the participants, who were all staff or students recruited from a single university, and because of the potential for contamination between participants because of the work/student environment. It was unclear if the trial conducted by Pavlou and colleagues91 was at a similar risk of workplace contamination.

Assessment of equity and sustainability

Figure 3 summarises the equity assessment. Results for the individual trials are detailed in Appendix 8 (see Table 57 ).

FIGURE 3.

Summary of the equity and sustainability assessment of trials included in the systematic review of men-only RCTs.

The majority of equity items were not considered, were unclear or were not reported by the trial authors. The trials did not report on diversity, sustainability or political context and did not describe any partnerships.

Only three trials were considered to have been conducted in a way that could have excluded specific groups of men, therefore influencing equity of access. The trial conducted by Jeffery and colleagues86 was considered to have excluded men who could not afford to pay a financial deposit. The SHED-IT trial89 was also judged to have potentially limited the inclusion of non-academic men because of the university setting and the use of internet and mobile phone technologies. Similarly, the trial by Patrick and colleagues90 required men to have internet access and to be computer literate.

Three trials considered unrepresentative samples of obese men: the trial by Esposito and colleagues85 recruited men from a weight-loss outpatient clinic who all had erectile dysfunction; men in the SHED-IT trial89 were all university staff or students; and men in the Pavlou trial91 were all police officers. Similarly, the trials by Wood and colleagues93 and King and colleagues88 had very narrow inclusion criteria and therefore excluded a wide number of potentially relevant men from their samples. It was unclear whether or not the trial by Jeffery and colleagues86 included a representative sample of obese men as there was insufficient detail in the report. Inclusion criteria were not reported by Patrick and colleagues. 90

The SHED-IT trial89 reported differences in compliance by occupation and age, with non-academic staff members and older men showing greater compliance. It is unclear whether or not any socioeconomic differences existed between withdrawals from each of the intervention groups. Patrick and colleagues90 reported that younger, non-white men were more likely to withdraw from their trial. The remaining trials did not report on any socioeconomic differences between withdrawals and exclusions.

Only three trials86,89,90 reported on categories relating to place of residence, race/ethnicity, occupation, gender, religion, education, socioeconomic status or social capital (PROGRESS). It was unclear in the trial by Patrick and colleagues90 whether or not the authors had attempted to address diversity or disadvantage with their trial intervention. All other trials did not address or report these items.

Four studies84,91 ^– 93 carried out fidelity checks for physical activity by monitoring heart rate or maximal oxygen consumption (VO₂ max.) of the men during the exercise sessions, thus ensuring that the men exercised at the required level of training intensity. The remaining studies did not report on fidelity checks, although the SHED-IT trial89 reported qualitative data concerning the enthusiasm of men for the internet and control interventions.

Most trials83,85,89,90,92,93 (6/11, 54.5%) reported details of who provided the interventions. The other trials did not report this item86 ^– 88,91 or it was unclear. 84 Providers were dietitians, exercise instructors or research staff.

The trial conducted by Borg and colleagues84 was partly funded by the Finnish pharmaceutical company Leiras Oy. As this company manufactures a weight-loss product it is unclear whether or not this created any author conflict. Patrick and colleagues90 also reported that three of the trial authors were co-owners of Santech, Inc. (San Diego, CA, USA), which developed products related to the described research at the time of publication. Both San Diego State University and the University of California approved this arrangement in accordance with their conflict of interest policies. The other trials did not report whether or not there was potential for author conflict.

An exercise programme compared with control

Wood and colleagues93 compared men in an exercise programme with a control group of men who made no change to their diet or level of physical activity. Men in the exercise programme participated in supervised exercise three times per week in 1-hour sessions with the aim of reducing their total body fat by one-third over a 9-month period. Exercise activities included calisthenics, muscle stretching, brisk walking and jogging. At 1 year the authors reported a significant difference in weight in favour of a programme of exercise compared with the control group (no diet or exercise) (mean difference –4.60 kg, 95% CI –6.18 kg to –3.02 kg) ( Figure 4 ).

FIGURE 4.

Effect of an exercise programme vs. control on weight (kg).

At 1 year, exercisers significantly lowered their triglyceride levels compared with the control group (reported p < 0.05) and significantly improved their HDL cholesterol compared with the control group (reported p < 0.01) ( Table 4 ). No other differences for cholesterol, LDL cholesterol or systolic or diastolic blood pressure (measured in the clinic) were reported as significant, although changes in the exercise group were all more beneficial than those in the control group.

A reducing diet compared with control

Wood and colleagues93 also compared the control group with men who followed a reducing diet (300–500 kcal per day deficit) aimed at reducing total body fat by one-third over 9 months; this group also made no alteration to their level of physical activity. At 1 year the diet group had lost significantly more weight than the control group, with a mean difference of –7.80 kg (95% CI –9.38 kg to –6.22 kg) ( Figure 5 ).

FIGURE 5.

Effect of a reducing diet vs. control on weight (kg).

At 1 year, dieters showed significantly lower triglyceride levels (reported p < 0.05) and HDL cholesterol levels (reported p < 0.001) than the control group ( Table 5 ). No other risk factor changes were reported as significantly different, although changes in the diet group were all more beneficial than changes in the control group.

An exercise programme compared with a reducing diet

A reducing diet was shown to produce a significant reduction in weight compared with an exercise programme at 1 year, with a mean difference of –3.20 kg (95% CI –4.78 kg to –1.62 kg) ( Figure 6 ). 93

FIGURE 6.

Effect of an exercise programme vs. a reducing diet on weight (kg).

None of the risk factor differences between the reducing diet group and the exercise programme group ( Table 6 ) were reported as statistically significant.

A low-fat reducing diet with behaviour therapy and exercise advice compared with control

Three studies examined behavioural therapy, exercise advice and low-fat reducing diets compared with a control group. In the SHED-IT trial by Morgan and colleagues,89 participants were given support and advice through a free study website. The control group received an information booklet only. At 12 months the internet group had lost more weight than the control group, with a mean difference of –2.20 kg (95% CI –5.65 to 1.25 kg), but the difference in weight reduction was not significant.

Similarly, Patrick and colleagues90 used the internet to deliver dietary and physical activity advice and behavioural therapy. The trial authors held interviews with two male weight-loss experts and held focus groups with men to tailor the intervention specifically for men described as overweight. The results of this developmental work showed that men wanted an intervention that was individualised, fact based, flexible and simple to understand. Men also indicated that they preferred the use of ‘businesslike’ language. Pedometers were provided to encourage physical activity and were enjoyed by the men for their novelty and assistance with self-monitoring of their behaviour. Men in the control group were given access to a website detailing general male-related health advice that was unlikely to lead to lifestyle changes that would promote weight loss (e.g. dealing with stress, hair loss, worksite injury prevention). Men receiving the weight-loss intervention lost more weight than men in the control group but the difference between the groups at 12 months was also not significant.

The trial by Esposito and colleagues85 examined behavioural therapy, advice on how to increase physical activity and a low-fat reducing diet. The men were recruited because they were obese and had erectile dysfunction [determined by a score of ≤ 21 on the International Index of Erectile Function (IIEF)126]. The men met in groups but received advice tailored to their individual requirements. The control group in this trial received general oral and written advice regarding healthy food choices and exercise at baseline only. At 2 years the intervention group had lost significantly more weight than the control group, with a mean difference of –13.00 kg (95% CI –16.18 kg to –9.82 kg) ( Figure 7 ).

FIGURE 7.

Effect of a low-fat reducing diet with behaviour therapy and exercise advice vs. control on weight (kg).

At 12 months the only significant difference between groups with regard to risk factors was in systolic blood pressure in favour of the intervention group (reported p < 0.03)89 ( Table 7 ).

At 24 months the intervention group in the trial by Esposito and colleagues85 showed significant improvements compared with the control group for total cholesterol, triglycerides and fasting plasma glucose (reported p ≤ 0.05). BMI, HDL cholesterol, systolic and diastolic blood pressure and waist-to-hip ratio were also significantly improved (reported p ≤ 0.01) (see Table 7 ).

Esposito and colleagues85 also reported that 17 out of 55 men in the intervention group compared with three out of 55 men in the control group reported an IIEF score of ≥ 22 (indicating regained sexual function; reported p = 0.001).

A diet and exercise programme compared with a diet only

Two trials reported the effect of adding an exercise programme to a diet compared with diet only. The trial by van Aggel-Leijssen and colleagues92 was a small trial in which men were randomised to a diet only or a diet and exercise programme. All men followed a 10-week diet. For the first 6 weeks men followed a very low energy (500 kcal per day) formula diet (Modifast) of 50 g of carbohydrate, 52 g of protein and 7 g of fat. For weeks 7–8, men consumed 330 kcal per day of the formula diet and 840 kcal per day from foods of their choice. During weeks 9–10, men consumed 170 kcal per day of the formula diet and 1170 kcal per day from their chosen food. The men were then instructed to stabilise their body weight for weeks 11–12. Men in the diet and exercise group also participated in a low-intensity exercise programme (40% VO₂ max.) for 12 weeks, which was then continued to week 52. The men trained four times per week in 1-hour sessions. Three of these sessions were supervised by a personal trainer in the research laboratory and the other session was unsupervised at home. The exercise sessions consisted of cycling, walking and aqua-jogging. Attendance for supervised exercise sessions was 57% (SD 20%). Reasons for non-attendance included illness, holidays and work commitments. Two of the men in the exercise group had to withdraw from the study because of knee injuries. There were no other reported adverse events. Weight data for men at 12 months were reported in a linked report by Lejeune and colleagues. 119 Men in the diet and exercise group did not lose as much weight as men in the diet-only group (mean difference 4.20 kg, 95% CI –1.47 to 9.87 kg).

The study by Pavlou and colleagues91 consisted of a pilot trial and the main trial. The pilot trial compared a low-calorie diet of 1000 kcal per day coupled with an exercise programme with a low-calorie diet only. The pilot trial also compared a low-calorie, low-carbohydrate diet of 1000 kcal per day plus an exercise programme with a very low-calorie diet only. The main trial compared the same two diets, with and without an exercise programme, as the pilot trial. The main trial additionally compared the effect of adding exercise to two types of very low-calorie diet (420 kcal per day and 800 kcal per day).

Combining results for all diet groups, the effect of adding an exercise programme to a diet was highly significant at 18 months (mean difference –7.63 kg, 95% CI –10.33 to –4.92 kg) and at 36 months (mean difference –8.22 kg, 95% CI –15.27 to –1.16 kg). There were no significant differences between the 1000 kcal per day low-carbohydrate diet and the 1000 kcal per day low-calorie diet, however, or for the two forms of very low-calorie diet (420 and 800 kcal per day) at 18 or 36 months ( Figure 8 ).

FIGURE 8.

Effect of a diet and exercise programme vs. a diet on weight (kg, SDs assumed).

In the main trial,91 systolic (mean difference –8.90 mmHg, 95% CI 13.65 to –4.15 mmHg) and diastolic (mean difference –12.10 mmHg, 95% –15.20 to –9.00 mmHg) blood pressure were significantly lower in the diet and exercise groups compared with the diet only groups at 18 months.

Type of diet

A high-protein reducing diet compared with a high-carbohydrate reducing diet

The trial by Benassi-Evans and colleagues83 compared a high-protein diet with a high-carbohydrate diet. Both diets consisted of 1670 kcal per day with some adjustment in energy to achieve an approximate weight loss of 1 kg per week. At 12 months the high-protein diet group had not lost as much weight as the high-carbohydrate diet group, although the difference between groups was not statistically significant (mean difference 1.55 kg, 95% CI –4.70 to 7.80 kg kg) ( Figure 9 ).

FIGURE 9.

Effect of a high-protein diet vs. a high-carbohydrate diet on weight (kg) at 12 months.

Monetary contracts: individual compared with group

The trial conducted by Jeffery and colleagues86 recruited 89 men to receive a 15-week behaviourally oriented weight reduction programme with a goal of achieving a total weight loss of 30 lb (13.6 kg) at a rate of 2 lb (0.2 kg) per week. Using a factorial design, men in the trial were randomised to pay monetary deposits of US$30, US$150 or US$300 and to either a group contract or an individual contract. Refunds were made at a rate of US$1, US$5 or US$10 per lb lost up to a maximum cumulative weight loss of 2 lb per week. Men in the individual contract groups received refunds based on individual weight loss, whereas those with group contracts were refunded based on the mean weight loss of their group. Group contracts produced significantly more weight loss than individual contracts both at 1 year and 2 years (reported p < 0.05). Effects of contract size at 1 year86 and 2 years117 were reported as not significant. Table 8 shows the mean weight loss for group and individual contracts by contract size at 1 and 2 years.

Weight maintenance

A behavioural intervention for weight maintenance compared with control

The trial conducted by King and colleagues88 randomised men from the diet and exercise arms of the trial by Wood and colleagues93 at the end of the 1-year trial period. These men were randomised within their original intervention groups to receive behavioural support based on their original weight-loss method or to two assessment-only control groups. The behavioural support intervention comprised monthly mailed information packs including a supportive letter and a list of coping strategies for problems relevant to their original intervention, for example holiday eating for dieters or finding time to engage in physical activity for exercisers. The men were also telephoned regularly to discuss any concerns or questions related to their problem areas and were weighed at 6-monthly intervals. Men in the assessment-only groups were given written information about their original weight-loss method at the start of the weight-maintenance period. The men received no other contact apart from the 6-monthly weight assessments.

The behavioural therapy intervention produced greater weight-maintenance success for the exercise group compared with the control group than it did for dieters. After 1 year, exercisers who received the behavioural intervention weighed significantly less than control participants (–3.10 kg, 95% CI –5.04 kg to –1.16 kg). Dieters in the behavioural intervention group were not significantly different from control participants after 1 year (0.60 kg, 95% –1.27 kg to 2.47 kg) ( Figure 10 ).

FIGURE 10.

Effect of a behavioural intervention for weight maintenance vs. control on weight (kg) at 12 months.

Attrition in men compared with women

Eight trials95,96,98,100,102,103,106,111 provided data that could be included in the analysis comparing the numbers of men and women who completed the trials. In total, there were 3813 participants; 1197 were men and 2616 were women ( Tables 11 and 12 ). The results shows that men were 11% (95% CI 8% to 14%, p < 0.001) more likely to complete the trial than women, suggesting men are highly motivated once commencing weight loss.

Quantitative outcomes

All trials reported either baseline and end weights or changes in weight by sex except for the FDPS,104 which reported incidence of diabetes by sex. Two studies reported waist circumference by sex106,109 and two reported BMI by sex. 106,108 Cardiovascular risk factors were reported by sex in five trials: two reported HbA_1c,97,108 three reported systolic and diastolic blood pressure,101,106,109 three reported triglycerides106,108,109 and three reported total and HDL cholesterol. 106,108,109 LDL cholesterol was reported by Volpe and colleagues109 and Ross and colleagues106 and fasting plasma glucose was reported by Vanninen and colleagues108 and Ross and colleagues. 106

Process outcomes

Hakala96 evaluated inpatient rehabilitation and health centre weight-loss programmes. The only reported difference by sex was that men in the health centre group were more likely to find GP appointments useful than women (63% of men vs. 23% of women). Both programmes met the expectations of participants. Participants in the rehabilitation group considered counselling by a nutritionist, physiotherapist and physician to be necessary and counselling from a social worker to be useful, although 62% wanted more individual counselling from the nutritionist. The majority (72%) of the rehabilitation participants felt that the 3-week inpatient period was satisfactory. The remaining 28% felt that it was too short. In comparison, the majority (72%) of the health centre participants felt the longer 10-week course was too short and 77% stated that they wanted more individual counselling with a greater emphasis on practical physical activity and dietary advice. No data were provided by sex for process outcomes.

Similarly, all nurses delivering the 6-week weight reduction programme in the trial by Karvetti and Hakala101 and the majority (58%) of the participants responding to the evaluation assessment (86% response rate) felt that the weight-loss programme in this trial was too short. Most nurses suggested that the programme should be extended to 10 weeks. The programme met the expectations of 66% of the participants. The only reported difference by sex was that most men (80%) and just under half of the women (42%) felt that six weight assessments during the intervention year were adequate. The remainder would have preferred more frequent assessment. Group support was also reported as being important for successful weight reduction.

Of the GPs participating in the trial conducted by Hakala,96 65% responded to the evaluation exercise. Only 41% considered their role as a GP to be suitable for the follow-up of obese participants, with 47% feeling that they were not suitable. Many felt that the task was uninteresting and useless, with only 35% showing that they were motivated to support participants in their weight reduction efforts. Similarly, the authors of the Lighten Up trial100 noted that GPs may have less faith in their ability to produce positive weight change in obese participants. Participants in this trial also noted that difficulty arranging regular appointment times with their general practice contributed to their failure to complete the full programme. Commercial companies, on the other hand, were able to offer weekly meetings at the same time each week. Neither trial reported differences by sex.

Risk of bias

The risk of bias assessment for the individual trials is shown in Appendix 10 (see Table 58 ). Figure 11 summarises the assessment.

FIGURE 11.

Summary of risk of bias assessment of trials included in the systematic review of men and women.

Five trials99,100,104 ^– 106 were judged to have adequate sequence generation but this was unclear for the majority of trials (14/20, 70%). Only one trial97 was judged as having a high risk for this item. Similarly, only five trials97,99,100,104,106 demonstrated successful allocation concealment and were considered at low risk for selection bias. Richelsen and colleagues105 compared orlistat therapy with placebo for weight maintenance; therefore, blinding was possible in this trial, although it was unclear whether or not those administering treatment were blinded to the nature of the interventions and whether participants were unblinded by gastrointestinal effects or not. Blinding of participants or health-care providers was not possible for any of the remaining trials. Four trials100,104,107,111 carried out blinding of outcome assessors. This was unclear for the remaining trials.

In three trials, groups were treated differently apart from the interventions received. For the trial conducted by Shai and colleagues,107 fewer men in the low-fat diet group participated in the ancillary substudy; in the DPP trial111 the standard lifestyle group received a placebo tablet whereas the intensive lifestyle group received no tablet; and in the trial conducted by Wing and colleagues113 the very low-calorie diet group received physician monitoring, which was not given to the low calorie diet group. Although only four trials95,96,101,111 were judged as being at high risk of attrition bias because of incomplete outcome data, the risk of bias for this item was unclear in just under half of all trial reports (9/20; 45%). 94,97 ^– 99,103,105,108,109,113 Only seven trials94,99,100,104,106,107,110 carried out a full intention-to-treat analysis and for three trials97,98,109 the method of analysis was unclear. The remaining trials analysed data for trial completers/compliers only. All trials were considered to be at low risk for selective reporting and only two trials were considered to be at unclear risk of other bias. Jeffery and colleagues98 enrolled men from a previous weight-loss trial conducted by the same authors. Gorin and colleagues94 provided free exercise equipment for use in the home and reimbursed participants in the intervention group for home delivery of grocery items; the control group received no similar ‘incentives’ or additional reason to feel obligated to complete the trial.

Assessment of equity and sustainability

Trial-level results are detailed in Appendix 10 (see Table 59 ). Figure 12 summarises the equity assessment.

FIGURE 12.

Summary of equity assessment of trials included in the systematic review of men and women.

As with our review of men-only RCTs, the majority of equity items were not considered, were unclear or were not reported by the trial authors. The trials mainly did not report on diversity, sustainability or political context and did not describe partnerships. The majority of trials included a representative participant spectrum with the exception of those conducted by Karvetti and Hakala,101 Jeffery and colleagues,98 Volpe and colleagues,109 and Gorin and colleagues,94 but none reported sociodemographic differences between participants withdrawing or excluded and those continuing. Four trials were considered to have occurred in settings that could have excluded specific population groups. The trials conducted by Jeffery and colleagues98 and Wing and colleagues112 both required financial deposits at trial entry, thus potentially excluding less affluent participants. Similarly, Jeffery and colleagues99 recruited participants with private health-care insurance only. Shai and colleagues107 conducted their trial in a work setting with on-site medical clinic facilities. Only two trials100,110 carried out fidelity checks. Three trials reported adverse harms. 100,105,106 As discussed earlier, only three trials evaluated process outcomes. 96,100,101 None of the trial authors were considered to have conflicting interests although this was unclear for five reports. 95,101,103,105,106

An exercise programme compared with a low-fat reducing diet

Volpe and colleagues109 compared a supervised exercise programme, a low-fat reducing diet, and a supervised exercise programme plus a low-fat reducing diet. The goal was for participants to lose 0.5–1.0 kg per week, although it is unclear whether this related to the dietary prescription alone or also took account of the exercise programme. By 12 months the exercise group and the diet group had gained weight. Men in the exercise group gained more weight than men in the diet group whereas the reverse was true for women ( Figure 13 ).

FIGURE 13.

Effect of an exercise programme vs. a low-fat reducing diet on weight change in men and women.

At 1 year, all groups showed little difference in changes to waist circumference. All groups showed a large deterioration in HDL cholesterol and an increase in LDL cholesterol. Similarly, all groups showed increases in triglycerides, with the exception of men in the diet group in which levels were lowered, although changes were more modest. Both men and women in the diet group and men in the exercise group lowered their systolic and diastolic blood pressure, with men showing greater decreases than women. Both types of blood pressure were raised for women in the exercise group, especially diastolic blood pressure. The difference in diastolic pressure between women in the diet group and women in the exercise group was significant after 1 year (p = 0.01) ( Table 13 ).

A low-fat reducing diet plus an exercise programme compared with a low-fat reducing diet

In the same trial109 a combination of a low-fat reducing diet and an exercise programme was compared with the same diet but without the exercise programme. At 12 months, men in the diet and exercise programme group gained less weight than men in the diet group. The protective effect of diet and exercise was repeated for the women ( Figure 14 ).

FIGURE 14.

Effect of a low-fat reducing diet plus an exercise programme vs. a low-fat reducing diet on weight change in men and women.

The low-fat reducing diet and exercise programme resulted in modest reductions in waist circumference. Although both groups showed poorer LDL cholesterol levels, the diet and exercise group showed less deterioration than the diet group for both sexes. Women in the diet and exercise group and men in both groups reduced their triglyceride levels, with men in the combined group showing the greatest reduction. Men in both groups reduced their systolic and diastolic blood pressure with the greatest reduction shown in the diet and exercise group. Conversely, both diastolic and systolic blood pressure slightly increased in women in this group, whereas diastolic and systolic blood pressure slightly decreased in women in the diet group ( Table 14 ).

A low-fat reducing diet plus an exercise programme compared with an exercise programme

The trial by Volpe and colleagues109 also compared diet plus exercise with exercise alone. The addition of a diet to exercise again provided greater benefits than exercise alone in terms of less weight gain for both sexes, although the difference was statistically significant for men only (p = 0.04) ( Figure 15 ).

FIGURE 15.

Effect of a low-fat reducing diet plus an exercise programme vs. an exercise programme on weight change in men and women.

Modest reductions in waist circumference were also found for this comparison. LDL cholesterol levels in all groups worsened. The decrease in triglyceride levels for the diet and exercise group was not repeated in the exercise group, with both sexes showing increased levels. The combination of both diet and exercise produced the greatest reduction for this outcome. Again, only men benefited from a reduction in systolic and diastolic blood pressure, with the greatest reduction occurring in the combined intervention group. Both diastolic and systolic pressure were raised in women in both groups at 1 year ( Table 15 ).

A low-fat reducing diet with exercise advice compared with control

Vanninen and colleagues108 investigated a low-fat reducing diet and exercise advice compared with basic conventional educational materials only. Details of the exact dietary prescription were not provided. All participants in this trial were non-insulin-dependent type 2 diabetics. After 1 year men in the intervention group had lost significantly more weight than men in the control group (p = 0.04). Women in the intervention group also lost more weight than women in the control group although the difference was not significant ( Figure 16 ). It should be noted that women had a higher mean BMI than men at baseline (34.2 and 33.4 kg/m² for women vs. 30.1 and 31.1 kg/m² for men for the control and intervention groups respectively).

FIGURE 16.

Effect of a low-fat reducing diet with exercise advice vs. control on weight change in men and women.

Men in the intervention group showed a greater improvement in total cholesterol, triglyceride and HbA_1c levels than men in the control group, although the control group had a greater improvement in HDL cholesterol levels. Fasting plasma glucose levels increased for both groups of men but the control group saw the greatest increase. No significant differences between groups were reported for men ( Table 16 ).

Women in the intervention group showed greater improvements than women in the control group for total cholesterol and triglycerides and also for HDL cholesterol. The control group showed a greater decrease in fasting plasma glucose levels, but the improvement in HbA_1c was the same in both women’s groups. However, women in the control group had higher HbA_1c and fasting plasma glucose levels than women in the intensified diet and exercise group at baseline (see Table 16 ). In a later report,125 the authors report a continuing reduction in BMI for the intervention groups at 15 months, with some reduction also seen in the control groups.

A low-fat reducing diet plus behavioural therapy compared with control

Karvetti and Hakala101 investigated a low-fat reducing diet plus behavioural therapy delivered in the primary health-care setting. The dietary prescription of 1200 kcal per day for weight reduction and 1800 kcal per day for maintenance was not reported to differ by sex. Participants in the intervention group were supported by public health nurses with three lectures from a physician, a psychologist and a physiotherapist. Participants were also given encouragement and support. The control group received no instruction and was contacted for yearly assessment only. After 1 year both men (–11.80 kg, 95% CI –16.86 to –6.74 kg) and women (–5.60 kg, 95% CI –8.74 to –4.57 kg) in the intervention group had lost significantly more weight than men and women in the control group respectively ( Figure 17 ).

FIGURE 17.

Effect of a low-fat reducing diet plus behavioural therapy vs. control on weight change in men and women.

Women in the intervention group showed a significant reduction in systolic blood pressure compared with women in the control group (reported p < 0.05) ( Table 17 ). No other changes in risk factors were reported as being statistically significant between groups according to sex.

A low-fat reducing diet plus exercise advice, behavioural therapy and home environment modification compared with a low-fat reducing diet plus exercise advice and behavioural therapy only

The trial by Gorin and colleagues94 randomised overweight and obese participants and an overweight household member, willing to act as a support partner, to a low-fat reducing diet with exercise advice and behavioural therapy or to the same treatment package but with modifications made to the home environment. Only participants received treatment in the standard programme whereas both participants and partners received treatment in the modified programme. Most participant–partner pairs were spouses or significant others (77.2%), with the remainder being participant–adult child (17.4%), participant–other relative (3.0%) and participant–roommate (2.5%) pairings. Modifications targeted physical and social cues in the home. At 18 months, women in the modified programme lost significantly more weight than women in the standard programme [–8.1 kg (SD 1.1 kg) vs. –4.2 kg (SD 1.1 kg), reported p = 0.014]. However, men in the standard programme lost more weight than men in the modified programme [–10.0 kg (SD 2.3 kg) vs. –4.6 kg (SD 2.2 kg)] although differences were not significant (reported p = 0.065). Partners in the modified programme lost more weight than partners in the standard programme at 18 months, regardless of sex. The authors reported that sex did not moderate weight regain for participants or partners, with men and women in both groups regaining weight at equivalent rates.

A low-fat reducing diet plus an exercise programme and behavioural therapy

Wing and colleagues112 randomised obese type 2 diabetic participants to receive a behavioural weight-loss programme either with their obese spouse (together) or without their spouse (alone). All participants received behavioural therapy consisting of stimulus control, problem solving, assertion, goal setting and cognitive techniques. Participants were also advised to monitor calorie intake to between 1200 and 1500 kcal per day and to set stepwise goals for walking. Participants in the together group attended with their spouses and both were targeted for weight loss. Participants in the alone group attended by themselves with spouses attending assessment sessions only. The weight loss of participants treated alone and together was not significantly different after 1 year, although men lost more weight when treated alone whereas women did better when treated together ( Table 18 ). There was no evidence that marital satisfaction affected weight loss. Spouses of both sexes lost more weight in the together group than in the alone group (p < 0.05) but there were no differences between treatment groups for change in calorie intake or exercise.

The Look AHEAD study110 recruited overweight or obese type 2 diabetics to a trial comparing an intensive lifestyle intervention comprising a low-fat reducing diet, some meal replacements and exercise advice and intensive behavioural therapy with diabetes support and education. The intensive lifestyle intervention was designed to produce a minimum weight loss of 7% of initial body weight during the first year, with dietary instructions tailored to initial body weight.

Wadden and colleagues110 reported weight data by sex for the active intervention group. The men in this group consistently lost more weight than the women at each annual assessment up to 4 years’ follow-up ( Table 19 ). The prescribed calorie intake was based on weight but it is not clear whether or not the calorie intake also took account of sex. Attendance and treatment contacts were similar for men and women in the first 4 years.

Effect on body image of a low-fat reducing diet with exercise advice and behavioural therapy compared with control

Several ancillary studies have reported sex effects in the Look AHEAD study. Stewart and colleagues123 investigated changes in body image in men and women. The authors recruited participants from one centre (Pennington Biomedical Research Centre, Los Angeles, CA, USA) and reported weight data by sex for both groups at 1 year. Women in this study had a higher baseline BMI than men in both treatment groups. Women in the intervention group had a slightly higher discrepancy between their current and their ideal body size (body dissatisfaction) than men (20.9 vs. 19.6), whereas men in the control group were slightly more dissatisfied than women with their body size (20.0 vs. 19.0). Both men and women in the intervention group lost significantly more weight than men and women in the control group respectively (reported p < 0.001) ( Figure 18 ). Both men and women in the intervention group showed a significant reduction in body image dissatisfaction compared with the control group after 1 year (p < 0.05 and p < 0.01 respectively). Men in both the intervention group and the control group showed a greater reduction in dissatisfaction than women [–8.1 (standard error 1.59) vs. –6.3 (standard error 0.94) for the intervention group and –3.3 (standard error 1.66) vs. –2.3 (standard error 0.96) for the control group].

FIGURE 18.

Effect of a low-fat reducing diet with exercise advice and behavioural therapy vs. control on weight change in men and women.

Effect on erectile dysfunction of a low-fat reducing diet with exercise advice and behavioural therapy compared with control

Wing and colleagues122 investigated the effect of weight loss on erectile function. Men with erectile dysfunction (n = 153 in the intervention group and n = 150 in the control group) were recruited from five of the 16 Look AHEAD centres. Sexual function was evaluated by self-reported completion of the IIEF. 126 The IIEF is a validated scale of erectile function, with five domains, with scores ≤ 10 denoting severe dysfunction, scores of 11–21 denoting moderate dysfunction, scores of 22–25 indicating mild dysfunction and scores ≥ 26 indicating no dysfunction. After 1 year men in the intervention group showed a greater weight change than men in the control group ( Figure 19 ).

FIGURE 19.

Effect of a low-fat reducing diet with exercise advice and behavioural therapy vs. control on weight change (kg) in men with erectile dysfunction.

After adjusting for baseline differences in erectile function score, the authors reported an increase (improvement) in erectile function of 1.3 (SD 4.7) with the intensive lifestyle intervention at 1 year and 0.03 (SD 5.7) in the control group at 1 year (reported p = 0.06).

Using the cut-off values of ≤ 21 and ≥ 22 to denote worsening or improvement in erectile function, respectively, the greater weight loss achieved by the intervention group did not produce a significantly greater improvement in erectile function in men with moderate or severe dysfunction compared with the control group. For mild or no dysfunction, significantly fewer men in the intervention group than in the control group reported worsening of function (reported p < 0.001). The intervention group also showed significantly greater changes in HbA_1c, systolic and diastolic blood pressure and HDL cholesterol (all reported p ≤ 0.01) ( Table 20 ).

TABLE 20 - Mean change in risk factors from baseline in the diet and exercise with behavioural therapy and control groups after 1 year (men only)

Between-group significance reported as p ≤ 0.01.

Outcome	Diet and exercise with behavioural therapy (n = 153)	Control (n = 153)
Total cholesterol (mmol/l)	–0.26 (n = 148)	–0.22 (n = 144)
LDL cholesterol (mmol/l)	–0.18 (n = 148)	–0.14 (n = 144)
HDL cholesterol (mmol/l)a	0.09 (n = 148)	–0.03 (n = 144)
Systolic blood pressure (mmHg)a	–7.5 (SD 16.3)	–1.5 (SD 14.9) (n = 150)
Diastolic blood pressure (mmHg)a	–4.7 (SD 7.9)	–1.0 (SD 7.6) (n = 150
HbA1c (%)a	–0.7 (SD 1.0)	–0.3 (SD 1.1) (n = 150)

Effect on spouses of a low-fat reducing diet with exercise advice and behavioural therapy

Gorin and colleagues121 assessed the impact of the intervention and control treatments on the untreated spouses of the Look AHEAD participants from three sites. There were no specific eligibility requirements for spouses other than a willingness to participate in the research. Spouses were not formally involved in either treatment group and were not expected to attend group meetings. Participants in the active intervention group were taught ways to enhance social support to promote their weight-loss efforts (e.g. how to communicate assertively with family members about desired food modifications). Participants in the control group received no such training. As in the full trial, intensive lifestyle participants lost more weight than control participants during the first year [–9.9 kg (SD 7.6 kg) vs. –1.2 kg (SD 4.9 kg), reported p < 0.001].

After 1 year, spouses of the intensive lifestyle participants had a weight change of –2.4 kg (SD 4.5 kg) compared with –0.2 kg (SD 3.3 kg) for spouses of control participants. The authors reported no effect by sex or baseline weight of the spouse.

A low-fat reducing diet with an exercise programme and behavioural therapy compared with placebo

The DPP111 randomised individuals at high risk of diabetes to an intensive low-fat reducing diet with an exercise programme and behavioural therapy, metformin or placebo. For the purposes of this review, we present data for the intensive intervention and placebo groups only. The aim of the intensive lifestyle programme was to lose 7% of initial body weight and maintain this weight loss throughout the trial. The calorie goals were calculated based on initial weight loss and a deficit of 500–1000 kcal per day, together with an increase in physical activity equivalent to 700 kcal per week. By 30 months, both sexes had lost more weight in the intensive group than in the placebo group (reported p < 0.001).

The authors reported that, within the lifestyle treatment group, black women lost significantly less weight than all other race–sex groups (reported p < 0.01) with the exception of black men ( Table 21 ). A higher proportion of women than men were obese at baseline (74% vs. 59%) and a higher proportion of black people than Hispanic and white people were obese at baseline (74% vs. 67% and 68% respectively).

Comparisons of different types of diet

Shai and colleagues107 investigated the effectiveness of a low-fat reducing diet (1500 kcal per day for women, 1800 kcal per day for men), a Mediterranean diet with equivalent calories and a low carbohydrate (20 g per day initially increasing to 120 g per day) non-restricted calorie diet in the Dietary Intervention Randomized Controlled Trial (DIRECT). At the end of the 2-year trial, the only significant difference was for women in the Mediterranean reducing diet group, who lost significantly more weight than women in the low-fat reducing diet group (p = 0.01) ( Figures 20–22 ).

FIGURE 20.

Effect of a low-carbohydrate diet vs. a low-fat reducing diet on weight change (kg) in men and women.

FIGURE 21.

Effect of a low-carbohydrate diet vs. a Mediterranean reducing diet on weight change (kg) in men and women.

FIGURE 22.

Effect of a Mediterranean reducing diet vs. a low-fat reducing diet on weight change (kg) in men and women.

Effect on Dietary Intervention Randomized Controlled Trial wives

Golan and colleagues120 conducted a parallel study describing the effect of the DIRECT dietary interventions on 74 wives of men participating in the trial. The wives were not randomised to any treatment group but were invited to attend the 90-minute support group meetings held every 2 months for the DIRECT participants. The aim of the meetings was to update the wives about the principles of the diet strategy to which their husbands had been randomised rather than treating the wives directly. At the end of the trial, men whose wives had attended support meetings lost more weight than men who did not have spousal support, both as an entire group and within each diet group ( Figure 23 ).

FIGURE 23.

Effect of spousal support on weight change (kg) in DIRECT husbands at 2 years.

Golan and colleagues120 investigated whether or not the intervention had any indirect influence on the DIRECT wives, termed ‘halo’ effects by the authors ( Figures 24–26 ). Differences in weight loss between groups were statistically significant between the low-carbohydrate diet and low-fat reducing diet groups only (reported p < 0.05).

FIGURE 24.

Effect of a low-carbohydrate diet vs. a low-fat reducing diet on weight change (kg) in DIRECT wives.

FIGURE 25.

Effect of a Mediterranean reducing diet vs. a low-carbohydrate diet on weight change (kg) in DIRECT wives.

FIGURE 26.

Effect of a Mediterranean reducing diet vs. a low-fat reducing diet on weight change (kg) in DIRECT wives.

Wing and colleagues113 compared an intermittent very low-calorie diet (400–500 kcal per day) with a low-calorie, low-fat diet (1000–1200 kcal per day) in a 1-year trial including participants with type 2 diabetes. Both groups also received behavioural therapy and exercise advice and deposited US$150, which was refunded depending on compliance. Women in the very low-calorie diet group lost significantly more weight after 1 year than women in the low-calorie, low-fat diet group (14.1 kg vs. 8.6 kg, reported p < 0.023) whereas men showed comparable weight loss in both treatment groups (15.4 kg and 15.5 kg respectively).

An on-demand diet compared with a regularly repeated diet

After 16 weeks of a 450 kcal per day diet, Lantz and colleagues103 randomised participants to receive either an on-demand very low-calorie diet (450 kcal per day) or a regularly repeated diet. After the initial 16 weeks, participants in the intermittent on-demand group followed a 500 kcal per day deficit diet but changed to the 450 kcal per day diet when their individual body weight reached a predetermined cut-off level throughout the trial period. Participants in the regularly repeated group followed the same 500 kcal per day deficit diet but used the 450 kcal per day diet for a fortnight every third month.

At 2 years, men in the on-demand intermittent diet group showed significantly more weight change than men in the regularly repeated diet group (mean difference –10.50 kg, 95% CI –4.84 to –16.16 kg). There was no significant difference in weight loss between diets for women (mean difference 1.80 kg, 95% CI 5.23 to –1.63 kg) ( Figure 27 ).

FIGURE 27.

Effect of intermittent vs. repeated very low-calorie diets on weight change in men and women.

Types of behaviour change for weight loss

Heitzmann and colleagues97 randomised participants with type 2 diabetes to a behavioural, cognitive or cognitive–behavioural therapy or a control group, who received muscle relaxation training and factual diabetes information only. Participants in all groups received dietary advice from a registered nutritionist and were given individual exercise advice. At 18 months across all intervention groups, it was reported that men lost an average of 3.63 kg whereas women gained an average of 0.04 kg ( Table 22 ). That men may benefit more from weight reduction programmes than women was shown by a borderline significant interaction (reported p = 0.057). Men also experienced a significantly greater reduction in HbA_1c than women (reported p < 0.05) but this difference was not significant between experimental groups. The effects of the individual programmes by sex were not reported.

An intensive inpatient rehabilitation setting compared with a community setting

Hakala and colleagues95,96 investigated the effectiveness of an intervention carried out in an initial inpatient rehabilitation setting compared with an intervention carried out in a community setting for people who were at least 50% overweight. The rehabilitation intervention included intensive behavioural and educational group sessions along with a prescribed physical activity programme and occupational therapy, as well as individual nutritionist (1200 kcal per day) and physician counselling. The community intervention involved the same dietary intervention but included either individual physician counselling95 or group-based counselling delivered in the health centre setting. 96

In the earlier trial,95 men did better in the community setting than in the inpatient setting, possibly because of more individual counselling, although differences were statistically significant only for years 1 and 2 (p < 0.01). There were no significant differences between groups for women ( Figure 28 ).

FIGURE 28.

Effect of an intervention in an intensive rehabilitation setting vs. an intervention in a community setting, including individual counselling, on weight change in men and women.

In the later trial by Hakala and colleagues,96 a similar comparison was carried out between an intervention in an initial intensive inpatient rehabilitation setting and an intervention in a community setting, delivered in group format only. When both rehabilitation and community interventions were delivered to men in groups, the rehabilitation setting produced favourable results, although differences were again statistically significant only over the first 2 years (p = 0.02 and p = 0.04 respectively). For women, the rehabilitation setting produced no significant benefit in weight loss over the community intervention for any time point from 1 to 5 years ( Figure 29 ).

FIGURE 29.

Effect of an intervention in an intensive rehabilitation setting vs. an intervention in a community setting on weight change in men and women.

Type of provider and tailoring for dietary intervention

A tailored nurse intervention compared with a doctor-provided leaflet for weight loss

Korhonen and colleagues102 randomised type 2 diabetic patients to receive care from either a doctor or a specialist nurse to investigate the effect of provider and individual tailoring of a diet and weight reduction intervention on achievement of weight loss. Participants randomised to the care of a doctor were given general concise written information on diet and weight reduction with no further instruction thereafter. Participants randomised to the care of the nurse were given individual assessments and tailored dietary instruction. Nurse interventions were repeated at follow-up visits. At 12 months there were no significant differences between groups for either sex for weight change ( Figure 30 ).

FIGURE 30.

A tailored nurse intervention vs. a doctor-provided leaflet for men and women.

Commercial providers compared with NHS providers

The Lighten Up trial100 randomised participants to one of three weight-loss programmes run by commercial companies (Weight Watchers, Slimming World and Rosemary Conley) or to one of three programmes delivered through the NHS (NHS Size Down, a GP or a pharmacist) or to a choice group in which they were able to choose one of the six programmes depending on their preference. For the control group (minimal interventional) participants received vouchers for 12 free sessions at a council-run leisure centre. All programmes lasted for 12 weeks and provided advice on exercise, but only the Rosemary Conley group had an exercise class provided. The Weight Watchers and Rosemary Conley programmes tailor advice by sex. The investigators did label some of the commercial groups as ‘male friendly’ so that men would know that they would not be the only ones present. For the Rosemary Conley programme, a group walk was available for people who did not want to undertake the group exercise. Women were more likely to choose a commercial provider than men (81% vs. 47%). Men in the choice arm were more likely to choose a NHS programme.

Statistically significant weight loss at 1 year from baseline was found for all groups except for the general practice and pharmacy groups (complete case analysis, baseline observation carried forward and last observation carried forward). Only the Weight Watchers group was significantly different from the control group for men and women combined (adjusted mean difference –2.49 kg, 95% CI –4.15 kg to –0.83 kg). The authors found no statistically significant interaction between sex and weight-loss programme.

Further data supplied by the authors show significant weight loss from baseline for women in the choice, NHS Size Down, Rosemary Conley, Slimming World, Weight Watchers and control groups, for all methods of analysis. For men, the NHS Size Down, Rosemary Conley and Weight Watchers programmes produced significant weight loss from baseline for all analyses. For men, the control and Slimming World programmes also produced significant weight loss from baseline, but only in the last observation carried forward analysis ( Table 23 ).

Telephone compared with mail advice and behaviour change techniques

Jeffery and colleagues99 compared the effectiveness of an intervention including weight reduction advice, physical activity advice and behaviour change techniques delivered by telephone or mail. A control group received usual care. Details of the dietary and exercise advice are unclear. Men in both the telephone group and the mail group had lost significantly more weight at 1 year than men in the control group (p = 0.03). By contrast, there were no significant differences between women in the telephone or mail group and women in the control group. There were no significant differences for either sex for the telephone group or the mail group ( Figures 31–33 ).

FIGURE 31.

Effect of telephone advice and a behaviour change intervention vs. control on weight change in men and women.

FIGURE 32.

Effect of mail advice and a behaviour change intervention vs. control on weight change in men and women.

FIGURE 33.

Effect of telephone advice vs. mail advice and a behaviour change intervention on weight change in men and women.

Physical activity advice, a healthy diet and behavioural therapy compared with usual care

The Prevention and Reduction of Obesity Through Active Living (PROACTIVE) trial106 randomised abdominally obese participants to receive an intervention offering physical activity and behavioural therapy or to usual care. Participants in the intervention group received individually tailored counselling based on motivational interviewing to provide the knowledge and skills to increase physical activity and adopt a healthy diet and eating patterns. Calorie reduction was not explicitly mentioned. Participants receiving usual care were given lifestyle advice for reducing obesity from their primary care physician following the usual appointment schedule and counselling approach. After 2 years, men in the intervention group had lost significantly more weight and significantly reduced their BMI and waist circumference compared with men in the usual care group. Women in the intervention group significantly reduced their waist circumference compared with women in the usual care group after 1 year but the effect was lost by 2 years. There were no other significant differences between groups ( Figure 34 and Table 24 ).

FIGURE 34.

Effect of physical activity advice, a healthy diet and behavioural therapy vs. usual care on weight change in men and women.

Varying monetary contracts for weight loss

Jeffery and colleagues98 investigated the effect of financial contracts on weight loss and weight maintenance in men and women recruited from a previously identified population or people self-referred through newspaper advertisements. All participants paid a US$150 deposit at the start of a 16-week weight-loss phase consisting of nutrition, exercise and behaviour change technique education sessions. Details of the dietary and exercise advice are unclear but the aim was to lose 2 lb (0.9 kg) per week. Participants randomised to the control groups were refunded their entire deposit at the initial session. Participants in the constant contract groups were refunded $30 for each successive group average weight loss of 5 lb (2.27 kg) and participants in the increasing contract groups were refunded US$5, US$10, US$20, US$40 and US$75 for successive 5-lb group weight losses. Following the weight-loss phase, 17 men and 25 women were randomised to receive either intensive or non-specific weight-maintenance sessions.

Those enrolling in the maintenance phase paid a US$100 deposit, which was returned in US$25 increments for attendance at quarterly group sessions. Those not enrolling in the maintenance phase were contacted at the 1-year follow-up assessment only. Table 25 details the weight loss for all participants at 1 year. Eleven participants gave a self-reported weight at this assessment. The trial authors added 5 lb to these weights for their analyses. Two participants who were lost to follow-up were recorded as having lost 0 lb and three participants were excluded from the analyses.

The authors reported that weight loss at 1 year was not statistically associated with recruitment source, contract type or sex. Analysis of percentage change in weight showed that women lost significantly more weight than men (reported p < 0.05). During weight maintenance it was reported that the only significant effect was for women in the intensive maintenance condition who outperformed men for this contract type (reported p < 0.006).

Orlistat compared with placebo for weight maintenance

Richelsen and colleagues105 investigated the effect of orlistat in people with type 2 diabetes, impaired fasting glucose or dyslipidaemia. Before randomisation, participants all initially lost at least 5% of their body weight by following a very low-calorie diet of 600–800 kcal per day over an 8-week period. Participants were then randomised to receive lifestyle counselling with either 120 mg of orlistat three times daily or matching placebo capsules. Weight change from the start of the diet to 3 years, analysed using the last observation carried forward for dropouts, was reported as significantly greater for women in the orlistat group than for women in the placebo group [–9.7 kg (–8.4%) vs. –6.3 kg (–5.3%), p < 0.02]. For men the difference between groups was not significant [orlistat vs. placebo: –8.9 kg (–8.3%) vs. –8.1 kg (–7.5%)].

Comparison between weight loss in men and weight loss in women across trials

For the analysis comparing weight loss between men and women a total of 11 studies had data available,95,96,98,101,102,106,108 ^– 112 including a total of 5519 participants, 3493 women and 2026 men. Two analyses were carried out comparing mean weight change and percentage weight loss between men and women. Both analyses show that there were no significant differences in weight change between men and women recruited to these studies ( Figures 35 and 36 ). However, few studies provided sufficient data to allow us to be sure that men and women were prescribed the same calorie deficit. Whether men or women adhere better to lifestyle prescription is unclear.

FIGURE 35.

Difference in mean weight loss (kg) between men and women.

FIGURE 36.

Percentage weight loss from baseline in men and women.

Weight-loss programmes and the workplace

Two studies investigated the delivery of dietary interventions to promote weight loss in a male-orientated workplace setting. The first of these146 was conducted in a single petrochemical worksite in Glasgow over a 24-week period. Men were randomised to receive either an individualised energy-deficient diet or a general low-calorie diet. Men in the energy-deficient diet group received individualised energy prescriptions calculated in accordance with their age, sex and body weight to produce a 600-kcal daily energy deficiency. Men in the general diet group consumed 1500 kcal per day. The weight-loss phase of the trial lasted 12 weeks. During weeks 12–24, diet and energy requirements were recalculated for weight maintenance rather than weight loss. Within each group, men were also randomised to consume meat or no meat. Men in the meat groups consumed red meat at least five times per week. Men in the non-meat groups substituted red meat with fish, eggs and cheese. Intention-to-treat analyses showed no significant difference in weight loss between the energy-deficient diet and the general low-calorie diet during the weight-loss phase [–3.7 kg (SD 3.4 kg) vs. –3.9 kg (SD 3.5 kg) respectively, p = 0.78]. Differences between groups for other risk factors were also non-significant ( Table 29 ). For those who completed the programme, both groups significantly gained weight during the weight-maintenance phase but the difference between groups was not significant [+0.9 kg (SD 2.0 kg) vs. +1.4 kg (SD 1.6 kg), p = 0.27].

Data for meat/non-meat participants were not reported by the original diet group. Meat consumption did not significantly affect weight loss or biochemistry. Average weight loss during the weight-loss phase for those men who completed the meat and non-meat programmes was –4.2 kg (SD 3.7 kg) and –5.0 kg (SD 3.2 kg) respectively. Again, both groups gained weight during the weight-maintenance phase [+0.9 kg (SD 1.6 kg) vs. +1.4 kg (SD 2.0 kg)]. The study authors commented that weight loss was less than expected and the dropout rate was significantly greater in the general low-calorie diet group.

The study conducted by Drummond and colleagues140 aimed to recruit male and female taxi drivers, but only men volunteered. Men followed a low-fat, high-carbohydrate (sugar-containing), energy-restricted diet that was tailored to produce a daily energy deficit of 600–700 kcal for each man over a 12-week period. Men were also advised to consume specialist sugar-containing Carb Boosters snacks to prevent a ‘starve and binge’ dieting pattern. The study was funded by the Sugar Bureau. By 12 weeks both BMI and weight were significantly reduced (p < 0.05). The average weight loss in completers was –5.5 kg, representing a 5.2% reduction in total body weight. The waist-to-hip ratio was reduced from 1.00 to 0.97. The men also perceived that their quality of life had increased over the 12-week period.

Sports clubs

Four studies138,139,141,149 investigated the use of male-orientated sports clubs to deliver nutritional advice and an exercise programme in men-only groups. The study by Brady and colleagues138 included male season ticket holders at Glasgow Rangers and Celtic football clubs. Both clubs were part of the SPL at the time of the study. Interested men were asked to provide details of their height, weight, general fitness level and approximate level of general health. Men were graded according to their BMI measurement and those with the highest BMI were selected first. A total of 20 men were invited to each club for the first programme cycle. There were two early withdrawals before the initial assessment but these places were readily filled by others. The authors reported that almost all men underestimated their true weight and overestimated their height. The men attended 10 weekly sessions lasting 2 hours at their respective club stadiums. The first hour covered discussion of health issues, with health lectures delivered by a physician. Mediterranean-style low-fat dietary advice was delivered by research dietitians and nurses, with calorie restriction when required. Emphasis was placed on changing lifestyles and adopting healthy behaviours for the men and their families. The second hour consisted of exercise classes run by professional Rangers and Celtic coaching staff. Heart rate monitors were used and the men were instructed to exercise at an appropriate heart rate for 30 minutes three to four times a week. At the start of the programme only six men could jog round the stadium football pitch without stopping (distance of around 350 m). After the 10-week programme all of the men could complete one lap and some were able to complete multiple laps without stopping. The programme attracted 100% attendance. Furthermore, some of the men arranged to exercise in small groups after the programme finished. Others encouraged the setting up of exercise programmes at their workplace. Data for 36 out of 40 men were available 15 months after the programme finished ( Table 31 ).

The study authors deliberately targeted their intervention at men who had a passion for their football club. Participants ranged from manual workers and office workers to a company director but all shared an enthusiasm for their club. The authors reported that every man considered their participation in the programme to be one of the most rewarding experiences of their lives.

Gray and colleagues141 delivered a similar intervention at two SPL clubs (Hearts and Kilmarnock) as part of a pilot RCT lasting 12 weeks. Men were randomised either to receive the FFIT weight-loss programme or to a waiting list control group. The recruitment target (n = 60) was achieved in the larger city-based club (Hearts) but not in the smaller town-based club. As with the study by Brady and colleagues,138 men in the FFIT programme attended 12 weekly sessions at their club training ground where they received personalised dietary and healthy eating and behaviour change advice followed by structured exercise classes. Men in the waiting list group were told that they would receive the FFIT programme 4 months later. The majority of men were white (99%), married (71.8%) and in full-time employment (76.7%). The authors reported that there were no baseline differences between the FFIT group and the waiting list group (Dr Cindy Gray, University of Glasgow, October 2012, personal communication).

The FFIT men achieved significant weight loss and a significant reduction in waist circumference compared with the waiting list men, who showed increases for both outcomes ( Table 32 ). The FFIT men also reported significant improvements in self-esteem, quality of life (as measured by the SF-12) and physical activity and healthy eating. These changes were also significantly different from those in the waiting list group (reported p = 0.001 to 0.048).

The attrition rate was reported as low, with 83.5% of the men completing the FFIT programme. The authors reported that the men were very positive about their participation in the programme. The affiliation with the football clubs was highlighted as being the main incentive as many men indicated that they would not have attended a weight-loss programme in an alternative setting. The professional coaching staff also gave positive feedback about their involvement with the programme.

Similar weight-loss programmes have been aimed at men in the rugby league setting. The Tackling Men’s Health (TMH) initiative was developed out of a partnership between the Department of Health, Leeds Rhinos Rugby League Club and Leeds Metropolitan University. The initiative was promoted in partnership with male health-related projects, the Yorkshire Man Mini Manual (Men’s Health Forum) and Change4Life (Cancer Research UK), and targeted men attending rugby matches at the Leeds Rhinos stadium with the aim of improving various areas of health and well-being (e.g. mental health, diet and nutrition, exercise and sexual health). TMH staff found it difficult to recruit men to the weight-loss group, highlighting difficulties in identifying men in a crowd of supporters as overweight without causing offence. The recruitment strategy was therefore altered for the weight-loss group: men were contacted by the Leeds Rhinos club through existing communication strategies and offered the opportunity to attend the weight-loss programme at the club training ground. Group sessions were split between discussing diet, physical activity and behaviour change and taking part in physical activity involving aerobic and muscle resistance training exercises. The programme was run over a 6-week period during the 2009 rugby league season for a group of seven men149 and was repeated over an 8-week period during the 2010 season for a group of 12 men (10 completed the course). 139 The main facilitator for the theory sessions was male but the intervention was delivered by a mixed-sex team. The average weight loss for the 2009 and 2010 groups was –2.1 kg (SD 3.3 kg) and –2.4 kg (SD 1.4 kg) respectively. Furthermore, data for the 2010 group showed an average 4.3-cm (SD 2.3 cm) reduction in waist circumference and an average BMI reduction of 0.8 kg/m².

Men in the 2010 group139 gave their evaluation of the course. The key reasons that the 12 men gave for joining the programme were the chance to use the club training facilities (five men); that the sessions were held in the evening (four men); the accessible location (three men); and the men-only group (two men). The men enjoyed the sessions, particularly the inclusion of physical activity exercises and the ability to discuss sensitive issues in a male-only environment. The initiative also provided the opportunity to refer the men to other local health initiatives and two men were referred to a specialist diabetes group.

Commercial weight-loss providers

We identified seven reports of men-only weight-loss groups provided by commercial organisations: Weight Watchers,33 Slimming World34 and LighterLife. 35,143 ^– 145,148 Both Weight Watchers and Slimming World delivered their standard programme in men-only groups. Men in the Weight Watchers report were referred by their NHS health-care provider. Weight Watchers provide a dietary plan that allocates points to certain foods whereas Slimming World allows unlimited consumption of low-calorie ‘free foods’ and a ‘no hunger’ diet plan. A 30-minute exercise add-on class was provided to two-thirds of the Weight Watchers groups. The Weight Watchers programme is designed to create a caloric deficit for a healthy weight-loss rate of up to 2 lb a week, taking into account an individual’s sex, age, height and current weight. Slimming World does not differentiate by sex in its dietary prescription.

The LighterLife ‘Man Plan’ programme was aimed at men and participants were provided with a booklet using male-friendly language and humour, using a football team analogy to describe the benefits of the programme. The LighterLife men abstained from conventional food and alcohol for 8 weeks and consumed very low-calorie diet formula foods known as ‘Man Plan’ food packs (although the diet is the same for men and women, irrespective of body weight). Abstinence from conventional food is reported by the LighterLife authors as providing clarity around boundaries for food and drink consumption. Group behaviour change work is also used to explore reasons for overeating and develop practical and psychological strategies for future weight maintenance.

Men in the LighterLife, Weight Watchers and Slimming World reports were followed up for a period of 8, 12 and 24 weeks, respectively, although men in the Weight Watchers group were able to complete the 12 sessions over a longer time period if necessary. For the Slimming World programme, data were analysed for men who had attended for at least 8 weeks.

LighterLife data were available at 8 weeks from five reports. 35,143 ^– 145,148 Details are presented in Table 33 .

The majority of men completed the Weight Watchers programme (63%, 39/62) and the majority achieved a weight loss of ≥ 5% [77% of programme completers (30/39) and 52% of all participants (32/62)]. 33 Average weight loss for all men was –6.3 kg (SD 4.1 kg) and average BMI reduction was 2.1 kg/m². The authors reported that the men were positive about the Weight Watchers programme and unanimously preferred the men-only meetings.

Weight-loss data for the Slimming World programme for those men who had been members for at least 24 weeks were available for 16 men only. 34 These men lost an average of 13.2 kg (SD 3.6 kg) representing an 11.4% (SD 4.2%) change in weight (69% achieved 10% weight loss and 31% achieved at least 5% weight loss). At the point of data collection, the average BMI reduction was –3.4 kg/m². There were no significant differences in weight loss between men employed in shift work and men employed in non-shift work.

NHS primary care setting

We identified four reports of weight-loss programmes delivered in the NHS primary care setting. 31,151 ^– 153 Three of these reports describe an evidence-based computer programme, ProHealthClinical. The programme provides instant access to practical eating plans and strategies for increasing physical activity as well as motivational information, such as behaviour change techniques, and tools for tracking patient progress for health-care practitioners. The Counterweight programme offers patients a prescribed energy-deficient diet, behavioural therapy and advice for increasing physical activity.