Table 2. Summary details of the community level studies included in the review.

Study	Design & quality appraisala	Setting & participants	Interventionb	Inequalityc	Summary results:d ↑=increase;↓= decrease; ↔=no change		Impact on inequalities in obesitye
Erfurt et al.35	Cluster randomized controlled trial; 3 year follow-up (post-intervention); final sample=1883; quality=strong	4 Workplaces, USA; 39–43 years; predominately male	3 year nutrition and physical activity prevention intervention: screening only (control) vs screening+health education (A) vs screening+health education+follow-up counselling (B) vs screening+health education+follow-up counselling+organized activities (C); cost of full group programmes ranged from 50 to $100 per participant	Disadvantage: manufacturing worksites (predominately blue collar employees)	Body weight All overweight employees (n=690): Intervention A Intervention B Intervention C ; Participants of weight loss components (n=238): Intervention A Intervention B Intervention C	↔ ↔ ↓ ↔ ↓ ↓	+
Ockene et al.27	Randomised controlled trial; 1 year follow-up (post- intervention); final sample=288; quality=strong	Senior community centre, USA; mean age≈52 years; ≈74% female	1 year nutrition and physical activity prevention intervention (Lawrence Latino Diabetes Prevention Programme): community-based, literacy-sensitive & culturally tailored intervention; individual & group counselling sessions: $661per participant for intervention ($1399 per participant for standard care)	Disadvantage: low-SES area	Weight BMI	↓ ↓	+
Kisioglu et al.33	Randomised controlled trial; 6month follow-up (approximately 5months post-intervention); final sample=400; quality=strong	Setting unclear, Turkey; mean age=34 years; 100% women	25-day nutrition and physical activity prevention intervention: health training support, nutrition educational material; encouragement to participate in education programme; no cost data reported	Disadvantage: low SES	BMI	↓	+
Faucher30; Faucher and Mobley31	Randomised controlled pilot trial; 5month follow-up (post-intervention); final sample=19; quality=strong	Community centre, USA; 100% women; mean age=35 years; BMI ⩾25 obese	20week nutrition treatment intervention: aimed at portion control; culturally sensitive and foods prepared culturally/economically specific to low-income Mexican-American families; no cost data reported	Disadvantage: low SES by income	Weight loss	↓(CS)	0
Rickel29	Randomised controlled trial; 12month follow-up (post-intervention); final sample=224; quality=strong	Cooperative extension service offices, USA; 50–75 years; 100% female; BMI⩾30	12month extended care nutrition and physical activity treatment intervention: extended care after a culturally tailored lifestyle intervention – face-to-face or telephone contact; no cost data reported	Disadvantage: counties of low levels of educational attainment and low household incomes	Body weight (Caucasians) Body weight (AA) Both intervention conditions vs control (no differences in weight change between intervention groups)	↓ ↔	0
Auslander et al.32	Randomised controlled trial; 3-month follow-up (post-intervention); final sample=239; quality=strong	Community, USA; 25-55 years; 100% female; obese	3-month nutrition treatment intervention (Eat Well Live Well): peer education focussing on nutrition skills tailored to individuals stage of change; social support from group sessions; no cost data reported	Disadvantage: low-income African-American women	BMI Body weight	↔ ↔	0
Baron et al.34	Randomised controlled trial; 1 year follow-up (9months post-intervention; final sample=119; quality=strong	Diet clubs, UK; 16-70 years; 85% female; overweight	3-month nutrition treatment intervention: diet clubs with weekly group meetings; followed either a LCD or LFD; no cost data reported	Gradient: LCD more effective than LFD in lower SES at 3months but this effect was not observed at 1 year	Body weight	↔	0
Befort et al.26	Randomised controlled pilot study; 16week follow-up (post-intervention); final sample=33; quality=strong	Community health centre, USA; ⩾18 years; 100% female; obese (BMI=30–50)	16week nutrition and physical activity treatment intervention: culturally appropriate behavioural weight loss programme plus motivational interviewing or health education; calorie reduced diet; self-monitoring; and food and physical activity guidance; no cost data reported	Disadvantage: lower income African-American women	Body weight BMI	↔ ↔	0
Campbell et al. 36	Cluster randomized controlled trial; 18month follow-up (mid-intervention); final sample=538; quality=strong	9 worksites, USA; 100% women	5-year nutrition and physical activity prevention intervention (health works for women): individualized computer-tailored health messages; a natural helpers programme at the workplace (lay health advisor); no cost data reported	Disadvantage: low-SES workplaces	BMI	↔	0
Howard-Pitney et al.38	Randomised controlled trial; 18week follow-up (post-intervention); final sample=242; quality=strong	Community setting, USA; mean age=31 years; 85% female	18week nutrition prevention intervention: 6-week, classroom-based intervention followed by a 12-week maintenance intervention; culturally sensitive; nutrition education delivered by professional nutrition health educators; telephone contact; no cost data reported	Disadvantage: low-income population	BMI	↔	0
Olvera et al.37	Randomised controlled trial; 12-week follow-up (post-intervention); final sample=35; quality=strong	Community locations, USA; mean age ≈35 years; 100% female (mothers)	12-week nutrition and physical activity prevention intervention: group aerobic or sports sessions or free play recreational activities; nutrition sessions; behavioural counselling sessions; no cost data reported	Disadvantage: low-income women	BMI	↔	0
Reid et al.28	Randomised controlled trial; 6month follow-up (6months post-intervention); final sample=149; quality=strong	Community health centre, Australia; ≈40% female; mean age=41 years	Single session nutrition prevention intervention: group counselling sessions conducted by a trained community health nurse; covered smoking cessation, dietary modification and non-pharmacological lowering of blood pressure; no cost data reported	Disadvantage: low-SES area	Weight	↔	0
Wing and Jeffery25	Randomised controlled trial; 10month follow-up (6months post-intervention); final sample=136; quality=strong	Setting unclear, USA; 22–55 year olds; ≈50% female	4month nutrition and physical activity treatment intervention: 16weekly group meetings led by a behaviour therapist and/or a nutritionist; weigh ins, review of self-monitoring records; lecture or discussion period; behaviour techniques; no cost data reported	Gradient: employment did not affect overall weight loss	Weight	↔	0

Abbreviations: BMI, body mass index; CS, clinically significant; EPHPP, effective public health practice project; LCD, low carbohydrate diet; LFD, low fat diet; SES, socio-economic status.

^aGlobal quality appraisal from EPHPP.¹⁸

^bPrevention or treatment intervention.

^cDisadvantage/gradient approach to inequality.

^dP<0.05.This is the relative mean differences between intervention and control at the longest follow-up.

^e+, positive intervention effect so it reduces obesity-related outcomes in low-SES groups or reduces the SES gradient in obesity-related outcomes; 0, no intervention effect or no effect on SES gradient in obesity-related outcomes.