## Abstract

Background: Limited evidence exists on obesity in the former Soviet Union (fSU), particularly its micro- and meso-level determinants. The objectives of this study were to determine age- and gender-adjusted prevalence of self-reported overweight and obesity in nine fSU countries; explore the relationship between individual and household (micro-level) factors and obesity; and explore the relationship between features of nutritional and physical environments (meso-level) and obesity. Methods: Data were collected from 18 000 adults using household surveys and from 333 communities using community profiles in Azerbaijan, Armenia, Belarus, Georgia, Kazakhstan, Kyrgyzstan, Moldova, Russia and Ukraine in 2010. Individual- and community-level determinants of self-reported obesity (body mass index ≥30 kg/m2) were analysed using multi-level random intercept logistic regression models. Results: A total of 13% of the males and 18% of the females were categorized as obese. Factors associated with obesity in males were older age, increasing educational achievement, declining self-reported health, alcohol consumption and automobile ownership. Males who were current smokers, not married and perceived physical activity to be important were less likely to be obese. For females, obesity was associated with older age, completion of secondary-level education, declining self-reported health and average household financial situation. Unmarried women were less likely to be obese. Multi-level analysis indicated that individuals living in communities with higher presence of garbage were more likely to be obese. Conclusions: This is the first study to examine both micro- and meso-level influences on obesity in fSU using multi-level analysis. Findings indicate a similar obesity risk profile to countries in Western Europe and North America.

## Introduction

The global prevalence of obesity has doubled in the past 30 years, and there are currently more than 500 million obese adults and this figure is projected to increase to 1.15 billion by 2030.1 It is now acknowledged that patterns of body weight are influenced by the characteristics held by individuals (e.g. demographic, socio-cultural and economic) and the actual environments in which they live.26 Research on the latter set of influences has given rise to the concept of the obesogenic environment and the ways in which influences such as access to healthy food choices, the built environment and opportunities for active lifestyles may encourage energy expenditure.4,610

Research undertaken primarily in the USA and UK has provided evidence of an association between availability, accessibility and affordability of food products and obesity4,7,8,1117 and also how perceived subjective characteristics of the environment can influence obesity, such as low neighbourhood safety, absence of sidewalks, unpleasant community aesthetics and the presence of garbage.4,7,12,14,18 There are fewer studies examining objectively measured attributes of the environment but these have indicated associations between obesity and access to fewer recreational facilities; greater walking distance to stores;19 negative community support; no access to outdoor physical activity facilities20 and the presence of vandalism, crime and social disorder.21

The existing evidence on obesogenic influences comes largely from North America, Australia and the UK, and there is a need for a better understanding of the influence of the varied social contexts and physical and food environments in developing and transition countries.68,12,2224 The transitional countries of the former Soviet Union (fSU) have faced many health challenges since the collapse of Communism and are also facing a growing epidemic of obesity now.24,25 However, only a few published studies have examined individual-level determinants of diet and obesity in the fSU,2428 and no studies have investigated the role of community influences on obesity in the fSU.

This study uses data from the Health in Times of Transition (HITT) study (http://www.hitt-cis.net/) on self-reported height and weight in 18 000 individuals residing in nine fSU countries (Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, Kyrgyzstan, Moldova, Russia and Ukraine).

The study objectives were to determine age- and gender-adjusted prevalence of self-reported overweight and obesity in the fSU; explore the relationship between individual and household (micro-level) factors and obesity and to explore the relationship between features of nutritional and physical environments (meso-level) and obesity.

## Methods

The data are from two data sources simultaneously collected in 2010 as part of the HITT study. The micro-level data are from a series of nationally representative cross-sectional household surveys with adults using multi-stage stratified random sampling and standardized questionnaires in the nine countries on a range of health outcomes and behaviours and demographic, socio-economic and environmental characteristics. Multi-stage cluster sampling was used for the surveys. Response rates in countries varied from 47.3% to 83%. All respondents gave informed consent and the research was approved by the Ethics Committee of the London School of Hygiene and Tropical Medicine.

The source of the meso-level data was a series of ‘community profiles’ conducted in the nine countries using a randomly selected sample of 333 primary sampling units (PSUs) used in the household surveys. The data from the community profiles were collected at the same time as the household surveys used for the micro-level data. Thirty community profiles were conducted in each country, except Russia (N = 73) and Ukraine (N = 50), to reflect their larger and more diverse populations. The community profiles were based on a previously validated EPOCH instrument.12 In brief, trained data collectors systematically recorded aspects of the environment relating to the general socio-economic situation (e.g. litter and graffiti), nutrition and physical activity (e.g. food environment and walkability) and tobacco and alcohol use (e.g. availability and advertising) using a standard methodology and a structured survey instrument. In addition, photographs were taken of specified community characteristics, which were then coded using a structured-coding form (see Supplementary Appendix A1).

The primary outcomes of interest were being overweight and obese. Overweight was defined as body mass index (BMI) ≥25 kg/m2 and obesity as BMI ≥30 kg/m2.29 BMI was calculated using self-reported height (metres) and weight (kilograms) data. All analyses were limited to individuals with complete data for both height and weight.

The key independent micro-level variables from the household survey were age (18–29, 30–39, 40–49, 50–59, ≥60 years), marital status (currently married, not married), educational achievement (completed secondary education or less, secondary vocational/some higher education, completed higher education), self-reported health (a 5-point scale of good, very good, fair, poor and very poor collapsed into three categories: good/very good, fair and poor/very poor), household economic situation (a 5-point scale of good, very good, average, bad and very bad collapsed into three categories: good/very good, average and bad/very bad), current smoking status (yes/no), frequency of alcohol consumption (any type) (≥weekly, 1–2 times per month/once in 2–3 months, rare/never), perceived importance of healthy diet and physical activity (both on a 4-point scale of important, quite important, rather unimportant and unimportant collapsed into important and not important), car ownership (yes/no) and residence: urban (capital of the country/regional capital/city/urban settlement) and rural (village).

The meso-level variables (from the community profiles) addressed the nutrition and the physical environments. The nutrition environment variables included the total number of restaurants/food outlets; the number of restaurants serving local or international cuisine; the presence of unhealthy food advertisements (billboards and posters on shop windows, bus shelters or other locations for fast food, snacks and fizzy carbonated drinks) which were then categorized as ≤2 or >2 and the cheapest price of healthy (1 kg of apples) and not healthy (1 l of fizzy sweetened drinks) food. Local prices were converted to US dollars based on exchange rates at the time of data collection and the mean price was used as a cutoff point to dichotomize the variable. The meso-level variables for the physical environment included pedestrian density; presence of crosswalks; perceived safety/pedestrian friendly; presence of graffiti and garbage; and how appealing the community was aesthetically [response options of strongly agree and somewhat agree (combined) vs. strongly disagree and somewhat disagree (combined)]. Access to a range of facilities (food outlets, green spaces, etc.) was assessed as ‘accessible’ if <1.5 km walking distance and ‘not accessible’ where walking distances were >1.5 km. The variables and response categories are shown in tables 3 and 4, and further details are given in the Supplementary Appendix A1.

### Analysis

For the micro-level analysis, we calculated age- and gender-stratified prevalence estimates for overweight and obesity for each country and the region as a whole. Gender differences in prevalence of obesity were assessed using chi-squared tests. A priori, micro-level analysis was stratified by gender, given evidence of gender variances in obesity and its determinants. We fitted a two-level random intercept logistic model for females and standard logistic regression techniques for males, as strong evidence for clustering of obesity within communities was present for females [likelihood ratio test: between-community variance (σ2u0) = 0; P < 0.001] but not for males (P = 0.32). Age and household economic situation were established a priori as determinants. A univariate regression analysis was initially conducted and factors whose statistical significance reached an initial significance level of P < 0.1 (with a low level set given the exploratory model building purpose) were entered into the multivariate analysis according to their strength of association using forward stepwise techniques. Those that remained associated with obesity at higher significance levels (P < 0.05) were then retained in final multivariate model.

For the meso-level analysis, data were merged with the household survey data to connect individuals with corresponding community level data (3082 respondents lived in the 333 primary sampling units with community profile data). Separate analyses were then conducted for the physical and nutritional environments. To ensure adequate power in this now smaller sample, models did not further differentiate by gender but were fitted with an interaction term between gender and age. We fitted an unconditional two-level random intercept logistic model, with obesity as an outcome and the meso-level community indicator (primary sampling unit) as the random portion. From this model, we calculated baseline between-community variance (σ2u) in obesity using the variance partition coefficient (VPC).30

We then fitted two-level random intercept logistic models, individuals (level 1) and communities (level 2), to analyse the relationship between obesity and meso-level variables after accounting for micro-level variables. Multivariate models were developed using forward step-wise techniques. Retention in the final multivariate model was assessed by decreases in between-community variance (VPC) and statistical judgement (P < 0.05).30 Cross-level interactions between micro- and meso-level variables were beyond the scope of the analysis. VPC was re-estimated for final models.

To ensure adequate power, both micro- and meso-level analyses combined observations from the nine HITT countries.31 All results are presented as odds ratios (ORs), using the binary outcome measure (not obese <30 kg/m2 or obese ≥30 kg/m2), with corresponding 95% confidence intervals (CIs) and likelihood ratio test P-values. All models were estimated using STATA 11.0.

## Results

Of the 18 000 respondents, 1056 were missing either weight (N = 385) or height (N = 671) and so the individual-level analysis was therefore limited to 16 944 (94.1%) individuals residing in 1711 household survey PSUs. The prevalence of overweight and obesity in both males and females was high (table 1). A total of 47.6% of males and 47.9% of females exceeded normal (BMI <25 kg/m2) categorization. Women were more likely to be obese than males (18.3% vs. 12.6%; P < 0.001), whereas males were more likely to be classified as overweight (35.3% vs. 29.3%; P < 0.001). Levels of obesity were highest in Georgia (23.4% males, 19.2% females). With the exceptions of Moldova and Ukraine, obesity follows a curvilinear trend in males, increasing up till the age of 60 years and then reducing. A similar trend was seen among females (P < 0.001), aside from Armenia and Kazakhstan (table 1).

Table 1

Prevalence of overweight and obesity by country, gender and age group

All ages combined

Age group (%)

Gender differencea (P, chi-squared test)
N % (95% CI) 18–29 30–39 40–49 50–59 ≥60
Armenia
Females
Normalb 550 58.70 (54.59–62.81) 82.50 63.68 44.00 44.44 35.09 0.17
Overweightb 293 31.27 (25.96–36.58) 16.43 28.86 41.33 37.61 45.61
Obeseb 94 10.03 (3.96–16.10) 1.07 7.46 14.67 17.95 19.30
Males
Normal 409 51.84 (47.00–56.68) 70.45 50.00 39.24 43.56 40.16
Overweight 316 40.05 (34.65–45.45) 25.51 42.95 51.27 41.58 49.61
Obese 64 8.11 (1.42–14.80) 4.05 7.05 9.49 14.85 10.24
Azerbaijan
Females
Normal 460 56.03 (51.49–60.57) 88.26 59.04 40.83 31.82 37.33
Overweight 258 31.43 (25.77–37.09) 9.13 33.13 40.83 46.21 42.67 0.06
Obese 103 12.55 (6.15–18.95) 2.61 7.83 18.35 21.97 20.00
Males
Normal 499 63.24 (59.01–67.47) 77.46 64.06 51.45 38.78 50.63
Overweight 214 27.12 (21.16–33.08) 19.94 25.00 30.43 44.9 34.18
Obese 76 9.63 (3.00–16.26) 2.60 10.94 18.12 16.33 15.19
Belarus
Females
Normal 534 53.88 (49.65–58.11) 88.06 67.58 41.04 29.61 27.31 <0.001
Overweight 278 28.05 (22.77–33.33) 10.07 21.98 34.10 39.47 42.59
Obese 179 18.06 (12.42–23.70) 1.87 10.44 24.86 30.92 30.09
Males
Normal 406 53.13 (48.28–57.98) 75.11 51.63 42.03 37.11 40.56
Overweight 275 35.99 (30.32–41.66) 21.89 39.22 42.75 41.24 45.45
Obese 83 10.86 (4.17–17.55) 3.00 9.15 15.22 21.65 13.99
Georgia
Females
Normal 569 50.18 (46.07–54.29) 85.33 55.26 43.46 28.08 37.76 0.03
Overweight 347 30.6 (25.75–35.45) 12.00 32.46 36.29 36.95 35.27
Obese 218 19.22 (13.99–24.45) 2.67 12.28 20.25 34.98 26.97
Males
Normal 280 38.46 (32.76–44.16) 57.06 35.46 30.50 38.64 28.48
Overweight 278 38.19 (32.48–43.90) 33.13 41.13 36.88 33.33 46.36
Obese 170 23.35 (16.99–29.71) 9.82 23.40 32.62 28.03 25.17
Kazakhstan
Females
Normal 525 55.91 (51.66–60.16) 84.03 63.89 45.70 33.80 29.34 <0.001
Overweight 246 26.20 (20.71–31.69) 12.17 21.76 33.11 35.92 39.52
Obese 168 17.89 (12.09–23.69) 3.80 14.35 21.19 30.28 31.14
Males
Normal 511 60.05 (55.80–64.30) 82.04 55.61 47.90 40.74 46.88
Overweight 260 30.55 (24.95–36.15) 14.79 36.22 38.92 43.52 36.46
Obese 80 9.40 (3.01–15.79) 3.17 8.16 13.17 15.74 16.67
Kyrgyzstan
Females
Normal 598 64.30 (60.46–68.14) 89.00 68.20 47.56 43.18 41.03 0.09
Overweight 239 25.70 (20.16–31.24) 10.00 24.42 35.37 42.42 35.90
Obese 93 10.00 (3.90–16.10) 1.00 7.37 17.07 14.39 23.08
Males
Normal 558 64.14 (60.16–68.12) 83.02 58.59 54.12 46.08 47.56
Overweight 245 28.16 (22.53–33.79) 14.47 33.33 35.88 38.24 40.24
Obese 67 7.70 (1.32–14.08) 2.52 8.08 10.00 15.69 12.20
Moldova
Females
Normal 515 53.26 (48.95–57.57) 84.40 65.43 40.48 31.47 35.79 0.003
Overweight 281 29.06 (23.75–34.37) 12.40 24.69 39.29 36.04 38.42
Obese 171 17.68 (11.96–23.40) 3.20 9.88 20.24 32.49 25.79
Males
Normal 367 49.33 (44.21–54.45) 69.33 45.05 41.79 35.48 37.96
Overweight 284 38.17 (32.52–43.82) 25.21 43.24 42.54 47.58 43.8
Obese 93 12.50 (5.78–19.22) 5.46 11.71 15.67 16.94 18.25
Russia
Females
Normal 796 48.01 (43.92–52.44) 82.05 60.40 42.67 24.06 25.00 <0.001
Overweight 457 27.56 (23.46–31.66) 12.56 23.10 32.57 35.71 36.48
Obese 405 24.43 (20.25–28.61) 5.38 16.50 24.76 40.23 38.52
Males
Normal 573 50.26 (46.17–54.35) 71.43 47.85 45.88 35.65 41.53
Overweight 404 35.44 (30.78–40.10) 24.68 34.95 36.60 42.59 42.37
Obese 163 14.30 (8.93–19.67) 3.90 17.20 17.53 21.76 16.10
Ukraine
Females
Normal 529 48.18 (43.92–52.43) 83.33 64.20 40.32 25.62 28.31 <0.001
Overweight 347 31.60 (26.71–36.49) 11.63 25.31 36.02 41.88 42.77
Obese 222 20.22 (14.94–25.50) 5.04 10.49 23.66 32.50 28.92
Males
Normal 393 49.50 (44.56–54.44) 69.66 41.61 37.30 37.61 45.56
Overweight 296 37.28 (31.77–42.79) 26.07 41.61 46.03 45.3 37.22
Obese 105 13.22 (6.74–19.70) 4.27 16.79 16.67 17.09 17.22
All ages combined

Age group (%)

Gender differencea (P, chi-squared test)
N % (95% CI) 18–29 30–39 40–49 50–59 ≥60
Armenia
Females
Normalb 550 58.70 (54.59–62.81) 82.50 63.68 44.00 44.44 35.09 0.17
Overweightb 293 31.27 (25.96–36.58) 16.43 28.86 41.33 37.61 45.61
Obeseb 94 10.03 (3.96–16.10) 1.07 7.46 14.67 17.95 19.30
Males
Normal 409 51.84 (47.00–56.68) 70.45 50.00 39.24 43.56 40.16
Overweight 316 40.05 (34.65–45.45) 25.51 42.95 51.27 41.58 49.61
Obese 64 8.11 (1.42–14.80) 4.05 7.05 9.49 14.85 10.24
Azerbaijan
Females
Normal 460 56.03 (51.49–60.57) 88.26 59.04 40.83 31.82 37.33
Overweight 258 31.43 (25.77–37.09) 9.13 33.13 40.83 46.21 42.67 0.06
Obese 103 12.55 (6.15–18.95) 2.61 7.83 18.35 21.97 20.00
Males
Normal 499 63.24 (59.01–67.47) 77.46 64.06 51.45 38.78 50.63
Overweight 214 27.12 (21.16–33.08) 19.94 25.00 30.43 44.9 34.18
Obese 76 9.63 (3.00–16.26) 2.60 10.94 18.12 16.33 15.19
Belarus
Females
Normal 534 53.88 (49.65–58.11) 88.06 67.58 41.04 29.61 27.31 <0.001
Overweight 278 28.05 (22.77–33.33) 10.07 21.98 34.10 39.47 42.59
Obese 179 18.06 (12.42–23.70) 1.87 10.44 24.86 30.92 30.09
Males
Normal 406 53.13 (48.28–57.98) 75.11 51.63 42.03 37.11 40.56
Overweight 275 35.99 (30.32–41.66) 21.89 39.22 42.75 41.24 45.45
Obese 83 10.86 (4.17–17.55) 3.00 9.15 15.22 21.65 13.99
Georgia
Females
Normal 569 50.18 (46.07–54.29) 85.33 55.26 43.46 28.08 37.76 0.03
Overweight 347 30.6 (25.75–35.45) 12.00 32.46 36.29 36.95 35.27
Obese 218 19.22 (13.99–24.45) 2.67 12.28 20.25 34.98 26.97
Males
Normal 280 38.46 (32.76–44.16) 57.06 35.46 30.50 38.64 28.48
Overweight 278 38.19 (32.48–43.90) 33.13 41.13 36.88 33.33 46.36
Obese 170 23.35 (16.99–29.71) 9.82 23.40 32.62 28.03 25.17
Kazakhstan
Females
Normal 525 55.91 (51.66–60.16) 84.03 63.89 45.70 33.80 29.34 <0.001
Overweight 246 26.20 (20.71–31.69) 12.17 21.76 33.11 35.92 39.52
Obese 168 17.89 (12.09–23.69) 3.80 14.35 21.19 30.28 31.14
Males
Normal 511 60.05 (55.80–64.30) 82.04 55.61 47.90 40.74 46.88
Overweight 260 30.55 (24.95–36.15) 14.79 36.22 38.92 43.52 36.46
Obese 80 9.40 (3.01–15.79) 3.17 8.16 13.17 15.74 16.67
Kyrgyzstan
Females
Normal 598 64.30 (60.46–68.14) 89.00 68.20 47.56 43.18 41.03 0.09
Overweight 239 25.70 (20.16–31.24) 10.00 24.42 35.37 42.42 35.90
Obese 93 10.00 (3.90–16.10) 1.00 7.37 17.07 14.39 23.08
Males
Normal 558 64.14 (60.16–68.12) 83.02 58.59 54.12 46.08 47.56
Overweight 245 28.16 (22.53–33.79) 14.47 33.33 35.88 38.24 40.24
Obese 67 7.70 (1.32–14.08) 2.52 8.08 10.00 15.69 12.20
Moldova
Females
Normal 515 53.26 (48.95–57.57) 84.40 65.43 40.48 31.47 35.79 0.003
Overweight 281 29.06 (23.75–34.37) 12.40 24.69 39.29 36.04 38.42
Obese 171 17.68 (11.96–23.40) 3.20 9.88 20.24 32.49 25.79
Males
Normal 367 49.33 (44.21–54.45) 69.33 45.05 41.79 35.48 37.96
Overweight 284 38.17 (32.52–43.82) 25.21 43.24 42.54 47.58 43.8
Obese 93 12.50 (5.78–19.22) 5.46 11.71 15.67 16.94 18.25
Russia
Females
Normal 796 48.01 (43.92–52.44) 82.05 60.40 42.67 24.06 25.00 <0.001
Overweight 457 27.56 (23.46–31.66) 12.56 23.10 32.57 35.71 36.48
Obese 405 24.43 (20.25–28.61) 5.38 16.50 24.76 40.23 38.52
Males
Normal 573 50.26 (46.17–54.35) 71.43 47.85 45.88 35.65 41.53
Overweight 404 35.44 (30.78–40.10) 24.68 34.95 36.60 42.59 42.37
Obese 163 14.30 (8.93–19.67) 3.90 17.20 17.53 21.76 16.10
Ukraine
Females
Normal 529 48.18 (43.92–52.43) 83.33 64.20 40.32 25.62 28.31 <0.001
Overweight 347 31.60 (26.71–36.49) 11.63 25.31 36.02 41.88 42.77
Obese 222 20.22 (14.94–25.50) 5.04 10.49 23.66 32.50 28.92
Males
Normal 393 49.50 (44.56–54.44) 69.66 41.61 37.30 37.61 45.56
Overweight 296 37.28 (31.77–42.79) 26.07 41.61 46.03 45.3 37.22
Obese 105 13.22 (6.74–19.70) 4.27 16.79 16.67 17.09 17.22

a: Gender difference in odds of obesity using chi-squared test

b: BMI: normal <25 kg/m2; overweight ≥25 kg/m2; obese ≥30 kg/m2

### Micro-level factors

The mean age and BMI were 41.2 years (SD 16.9) and 25.2 kg/m2 (SD 4.1) for all male participants and 43.4 years (SD 17.1) and 25.3 kg/m2 (SD 5.2) for all female participants, and related baseline characteristics are presented in table 2. For men, after adjustment in the multivariate analysis, factors associated with obesity were older age, increasing educational achievement, worse self-reported health, greater frequency of alcohol consumption and automobile ownership. Males who were current smokers, not married and perceived physical activity to be important were less likely to be obese. A priori, it was decided to retain household economic situation in the final model although it was no longer significantly associated with obesity (P = 0.48) (table 2). For women, the characteristics independently associated with obesity in the multivariate analysis included greater age, worse health, reporting an average household economic situation, being married and completing secondary/some higher education. Age negatively confounded the relationship between education and obesity in women with some post-secondary or vocational training, showing for a positive association in the final model (table 2).

Table 2

Micro-level factors associated with male and female obesity, all countries combined

Variable Totala, N (%) Obeseb, N (%) Univariablec

Multivariablec

OR (95% CI) P OR (95% CI) P
Male obesity (N = 7469)
Age (years)
18–29 2371 (31.74) 94 (3.96)
30–39 1406 (18.82) 172 (12.23) 3.38 (2.60–4.38)  2.60 (1.95–3.46)
40–49 1366 (18.29) 222 (16.25) 4.70 (3.66–6.04) <0.001 3.16 (2.38–4.20) <0.001
50–59 1095 (14.66) 210 (19.18) 5.75 (4.25–7.42)  3.58 (2.66–4.80)
≥60 1231 (16.48) 203 (16.49) 4.78 (3.71–6.18)  2.91 (2.15–3.92)
Marital status
Married 4841 (64.81) 744 (15.37)
Not married 2606 (34.89) 155 (5.95) 0.35 (0.29–0.42) <0.001 0.62 (0.50–0.75) 0.01
Education
Secondary or less 924 (12.37) 82 (8.87)
Vocational/some higher 4861 (65.08) 568 (11.68) 1.36 (1.07–1.73) <0.001 1.48 (1.15–1.92) <0.001
Full higher education 1661 (22.24) 251 (15.11) 1.83 (1.40–2.38)  1.89 (1.42–2.50)
Self-reported health
Good/very good 3612 (48.36) 279 (7.72)
Fair 2842 (38.05) 439 (15.45) 2.18 (1.86–2.56) <0.001 1.56 (1.31–1.86) <0.001
Poor/very poor 993 (13.29) 182 (18.33) 2.68 (2.19–3.28)  1.86 (1.46–2.38)
Smoking status
Non-smoker 3862 (51.71) 519 (13.44)
Smoker 3602 (48.23) 379 (10.52) 0.76 (0.66–0.87) <0.001 0.69 (0.59–0.80) <0.001
Alcohol consumption
Rare/never 2548 (34.11) 238 (11.66)
Occasional/monthly 2612 (34.97) 286 (12.33) 1.05 (0.89–1.25) 0.01 1.02 (0.85–1.23) 0.01
Weekly/frequent 2239 (29.98) 299 (13.80) 1.28 (1.08–1.52)  1.26 (1.05–1.52)
Importance of healthy diet
Not important 342 (4.58) 47 (13.74)
Important 7074 (94.71) 847 (11.97) 0.85 (0.62–1.17) 0.34
Importance of physical activity
Not important 517 (6.92) 76 (16.45)
Important 6881 (92.13) 750 (12.36) 0.69 (0.54–0.88) 0.004 0.73 (0.56–0.95) 0.02
Household financial situation
Good/very good 1872 (25.06) 181 (9.67)
Average 4222 (56.53) 519 (12.29) 1.31 (1.10–1.57)  1.08 (0.89–1.31)
Bad/very bad 1297 (17.37) 190 (14.65) 1.60 (1.29–1.99) 0.001 1.24 (0.97–1.60) 0.48
Living location
Urban 4431 (59.33) 524 (11.83)
Rural 3038 (40.67) 377 (12.41) 1.06 (0.92–1.22) 0.45
Automobile ownership
No 4596 (61.53) 501 (10.90)
Yes 2873 (38.47) 400 (13.92) 1.32 (1.15–1.52) 0.001 1.34 (1.15–1.57) <0.001
Fear of harassment/threatened on street
Not worried 4677 (62.62) 569 (12.17)
Worried 2753 (36.86) 326 (11.84) 0.97 (0.84–1.12) 0.68
Female obesity (N = 9475)
Age (years)
18–29 2464 (26.01) 75 (3.04)
30–39 1837 (19.39) 205 (11.16) 4.07 (3.09–5.35)  3.28 (2.47–4.34)
40–49 1829 (19.30) 378 (20.67) 8.62 (6.65–11.18) <0.001 6.04 (4.61–7.90) <0.001
50–59 1501 (15.84) 453 (30.18) 14.42 (11.13–18.69)  8.96 (6.83–11.73)
≥60 1844 (19.46) 542 (29.39) 13.96 (10.81–18.02)  8.48 (6.45–11.14)
Marital status
Married 5666 (59.80) 1035 (18.27)
Not married 3775(39.84) 609 (16.13) 0.85 (0.76–0.95) 0.01 0.81 (0.71–0.92) 0.004
Education
Secondary or less 1251 (13.20) 253 (20.22)
Vocational/some higher 5952 (62.82) 1095 (18.40) 0.88 (0.75–1.03)  1.24 (1.04–1.48)
Full higher education 2255 (23.80) 304 (13.48) 0.61 (0.50–0.74) <0.001 0.96 (0.78–1.19) <0.001
Self-reported health
Good/very good 3365 (35.51) 213 (6.33)
Fair 4140 (43.69) 844 (20.39) 3.86 (3.29–4.54)  2.23 (1.87–2.66)
Poor/very poor 1953 (20.61) 596 (30.52) 6.85 (5.75–8.17) <0.001 3.11 (2.55–3.81) <0.001
Smoking status
Non-smoker 8631 (91.09) 1537 (17.81)
Smoker 838 (9.16) 114 (13.60) 0.70 (0.56–0.86) <0.001
Alcohol consumption
Rare/never 5785 (61.06) 1049 (18.13)
Occasional/monthly 2724 (28.75) 448 (16.45) 0.86 (0.76–0.98)
Weekly/frequent 744 (7.85) 125 (16.80) 0.87 (0.71–1.08) 0.03
Importance of healthy diet
Not important 320 (3.38) 66 (20.62)
Important 9073 (95.76) 1573 (17.34) 0.81 (0.61–1.08) 0.15
Importance of physical activity
Not important 734 (7.75) 157 (21.39)
Important 8631 (91.09) 1473 (17.07) 0.76 (0.62–0.92) 0.01
Household financial situation
Good/very good 1978 (20.88) 216 (10.92)
Average 5436 (57.37) 996 (18.32) 1.84 (1.57–2.17)  1.21 (1.01–1.44)
Bad/very bad 1953 (20.61) 431 (22.07) 2.39 (1.99–2.87) <0.001 1.04 (0.85–1.28) 0.01
Location of residence
Urban 5884 (62.10) 990 (16.83)
Rural 3591 (37.90) 663 (18.46) 1.14 (1.01–1.29) 0.03
Automobile ownership
No 6637 (70.05) 1179 (17.76)
Yes 2838 (29.95) 474 (16.70) 0.91 (0.81–1.04) 0.16
Fear of harassment/threatened on street
Not worried 4868 (51.38) 894 (18.36)
Worried 4537 (47.88) 746 (16.44) 0.84 (0.75–0.94) 0.003
Variable Totala, N (%) Obeseb, N (%) Univariablec

Multivariablec

OR (95% CI) P OR (95% CI) P
Male obesity (N = 7469)
Age (years)
18–29 2371 (31.74) 94 (3.96)
30–39 1406 (18.82) 172 (12.23) 3.38 (2.60–4.38)  2.60 (1.95–3.46)
40–49 1366 (18.29) 222 (16.25) 4.70 (3.66–6.04) <0.001 3.16 (2.38–4.20) <0.001
50–59 1095 (14.66) 210 (19.18) 5.75 (4.25–7.42)  3.58 (2.66–4.80)
≥60 1231 (16.48) 203 (16.49) 4.78 (3.71–6.18)  2.91 (2.15–3.92)
Marital status
Married 4841 (64.81) 744 (15.37)
Not married 2606 (34.89) 155 (5.95) 0.35 (0.29–0.42) <0.001 0.62 (0.50–0.75) 0.01
Education
Secondary or less 924 (12.37) 82 (8.87)
Vocational/some higher 4861 (65.08) 568 (11.68) 1.36 (1.07–1.73) <0.001 1.48 (1.15–1.92) <0.001
Full higher education 1661 (22.24) 251 (15.11) 1.83 (1.40–2.38)  1.89 (1.42–2.50)
Self-reported health
Good/very good 3612 (48.36) 279 (7.72)
Fair 2842 (38.05) 439 (15.45) 2.18 (1.86–2.56) <0.001 1.56 (1.31–1.86) <0.001
Poor/very poor 993 (13.29) 182 (18.33) 2.68 (2.19–3.28)  1.86 (1.46–2.38)
Smoking status
Non-smoker 3862 (51.71) 519 (13.44)
Smoker 3602 (48.23) 379 (10.52) 0.76 (0.66–0.87) <0.001 0.69 (0.59–0.80) <0.001
Alcohol consumption
Rare/never 2548 (34.11) 238 (11.66)
Occasional/monthly 2612 (34.97) 286 (12.33) 1.05 (0.89–1.25) 0.01 1.02 (0.85–1.23) 0.01
Weekly/frequent 2239 (29.98) 299 (13.80) 1.28 (1.08–1.52)  1.26 (1.05–1.52)
Importance of healthy diet
Not important 342 (4.58) 47 (13.74)
Important 7074 (94.71) 847 (11.97) 0.85 (0.62–1.17) 0.34
Importance of physical activity
Not important 517 (6.92) 76 (16.45)
Important 6881 (92.13) 750 (12.36) 0.69 (0.54–0.88) 0.004 0.73 (0.56–0.95) 0.02
Household financial situation
Good/very good 1872 (25.06) 181 (9.67)
Average 4222 (56.53) 519 (12.29) 1.31 (1.10–1.57)  1.08 (0.89–1.31)
Bad/very bad 1297 (17.37) 190 (14.65) 1.60 (1.29–1.99) 0.001 1.24 (0.97–1.60) 0.48
Living location
Urban 4431 (59.33) 524 (11.83)
Rural 3038 (40.67) 377 (12.41) 1.06 (0.92–1.22) 0.45
Automobile ownership
No 4596 (61.53) 501 (10.90)
Yes 2873 (38.47) 400 (13.92) 1.32 (1.15–1.52) 0.001 1.34 (1.15–1.57) <0.001
Fear of harassment/threatened on street
Not worried 4677 (62.62) 569 (12.17)
Worried 2753 (36.86) 326 (11.84) 0.97 (0.84–1.12) 0.68
Female obesity (N = 9475)
Age (years)
18–29 2464 (26.01) 75 (3.04)
30–39 1837 (19.39) 205 (11.16) 4.07 (3.09–5.35)  3.28 (2.47–4.34)
40–49 1829 (19.30) 378 (20.67) 8.62 (6.65–11.18) <0.001 6.04 (4.61–7.90) <0.001
50–59 1501 (15.84) 453 (30.18) 14.42 (11.13–18.69)  8.96 (6.83–11.73)
≥60 1844 (19.46) 542 (29.39) 13.96 (10.81–18.02)  8.48 (6.45–11.14)
Marital status
Married 5666 (59.80) 1035 (18.27)
Not married 3775(39.84) 609 (16.13) 0.85 (0.76–0.95) 0.01 0.81 (0.71–0.92) 0.004
Education
Secondary or less 1251 (13.20) 253 (20.22)
Vocational/some higher 5952 (62.82) 1095 (18.40) 0.88 (0.75–1.03)  1.24 (1.04–1.48)
Full higher education 2255 (23.80) 304 (13.48) 0.61 (0.50–0.74) <0.001 0.96 (0.78–1.19) <0.001
Self-reported health
Good/very good 3365 (35.51) 213 (6.33)
Fair 4140 (43.69) 844 (20.39) 3.86 (3.29–4.54)  2.23 (1.87–2.66)
Poor/very poor 1953 (20.61) 596 (30.52) 6.85 (5.75–8.17) <0.001 3.11 (2.55–3.81) <0.001
Smoking status
Non-smoker 8631 (91.09) 1537 (17.81)
Smoker 838 (9.16) 114 (13.60) 0.70 (0.56–0.86) <0.001
Alcohol consumption
Rare/never 5785 (61.06) 1049 (18.13)
Occasional/monthly 2724 (28.75) 448 (16.45) 0.86 (0.76–0.98)
Weekly/frequent 744 (7.85) 125 (16.80) 0.87 (0.71–1.08) 0.03
Importance of healthy diet
Not important 320 (3.38) 66 (20.62)
Important 9073 (95.76) 1573 (17.34) 0.81 (0.61–1.08) 0.15
Importance of physical activity
Not important 734 (7.75) 157 (21.39)
Important 8631 (91.09) 1473 (17.07) 0.76 (0.62–0.92) 0.01
Household financial situation
Good/very good 1978 (20.88) 216 (10.92)
Average 5436 (57.37) 996 (18.32) 1.84 (1.57–2.17)  1.21 (1.01–1.44)
Bad/very bad 1953 (20.61) 431 (22.07) 2.39 (1.99–2.87) <0.001 1.04 (0.85–1.28) 0.01
Location of residence
Urban 5884 (62.10) 990 (16.83)
Rural 3591 (37.90) 663 (18.46) 1.14 (1.01–1.29) 0.03
Automobile ownership
No 6637 (70.05) 1179 (17.76)
Yes 2838 (29.95) 474 (16.70) 0.91 (0.81–1.04) 0.16
Fear of harassment/threatened on street
Not worried 4868 (51.38) 894 (18.36)
Worried 4537 (47.88) 746 (16.44) 0.84 (0.75–0.94) 0.003

a: Where percentage of individuals <100% due to incomplete responses

b: Obese ≥30 kg/m2

c: Men: Univariable and multivariable estimates obtained from standard logistic regression model: likelihood ratio test (σ2u0 = 0): P = 0.32. Women: Univariable and multivariable estimates obtained from two-level random intercept logistic model adjusted for clustering of obesity within communities

*P > 0.05 in final multivariable model

Table 3

Characteristics of nutritional environments and their association with obesity, men and women combined (communities N = 333; individuals N = 2899)

Variable Variable indicator Number of communities (%) Univariablea

Multivariableb

OR (95% CI) P OR (95% CI) P
Not in walking distance 175 (52.55) 0.90 (0.71–1.14) 0.39 — —
Not in walking distance 283 (84.98) 1.03 (0.75–1.42) 0.86 — —
Number of healthy food stores 0–4 188 (56.46)
5–8 88 (26.43) 1.13 (0.85–1.50) 0.40 — —
≥9 57 (17.12) 1.12 (0.81–1.54)
Number of restaurants 82 (24.62)
1–4 176 (52.85) 1.20 (0.90–1.60) 0.48 — —
5+ 75 (22.52) 1.12 (0.78–1.59)
>2 114 (34.23) 0.83 (0.64–1.07) 0.16 — —
Cheapest price: 1 kg of apples (US$) ≤$0.94 160 (48.05)
>$0.94 173 (51.95) 1.25 (0.99–1.59) 0.06 1.25 (0.99–1.59) 0.06 Cheapest price: 1 L of fizzy sweetened drinks (US$) ≤$0.80 179 (53.75) >$0.80 154 (46.25) 1.03 (0.82–1.31) 0.78 — —

Two-level random intercept model Variance estimate Variance partition coefficient (%) P-value for between community variation: likelihood ratio test (σ2u0 = 0)

Unconditional model 0.12 3.64 0.02
Micro-level variables 0.14 4.10 0.02
Micro- and meso-level variables—final model 0.13 3.80 0.03
Variable Variable indicator Number of communities (%) Univariablea

Multivariableb

OR (95% CI) P OR (95% CI) P
Not in walking distance 175 (52.55) 0.90 (0.71–1.14) 0.39 — —
Not in walking distance 283 (84.98) 1.03 (0.75–1.42) 0.86 — —
Number of healthy food stores 0–4 188 (56.46)
5–8 88 (26.43) 1.13 (0.85–1.50) 0.40 — —
≥9 57 (17.12) 1.12 (0.81–1.54)
Number of restaurants 82 (24.62)
1–4 176 (52.85) 1.20 (0.90–1.60) 0.48 — —
5+ 75 (22.52) 1.12 (0.78–1.59)
>2 114 (34.23) 0.83 (0.64–1.07) 0.16 — —
Cheapest price: 1 kg of apples (US$) ≤$0.94 160 (48.05)
>$0.94 173 (51.95) 1.25 (0.99–1.59) 0.06 1.25 (0.99–1.59) 0.06 Cheapest price: 1 L of fizzy sweetened drinks (US$) ≤$0.80 179 (53.75) >$0.80 154 (46.25) 1.03 (0.82–1.31) 0.78 — —

Two-level random intercept model Variance estimate Variance partition coefficient (%) P-value for between community variation: likelihood ratio test (σ2u0 = 0)

Unconditional model 0.12 3.64 0.02
Micro-level variables 0.14 4.10 0.02
Micro- and meso-level variables—final model 0.13 3.80 0.03

a: Univariable estimate obtained from two-level random intercept model adjusted for micro-level variables

b: Multivariable estimates obtained from two-level random intercept logistic model

Table 4

Characteristics of physical environments and their association with obesity, men and women combined (communities N = 333; individuals N = 2899)

Variable Variable indicator Number of communities (%)a Univariableb

Multivariablec

OR (95% CI) P OR (95% CI) P
Not in walking distance 97 (32.01) 1.09 (0.86–1.40) 0.47 — —
Not in walking distance 104 (34.32) 1.14 (0.90–1.45) 0.28 — —
Not in walking distance 79 (26.07) 1.02 (0.79–1.31) 0.87 — —
Not in walking distance 146 (48.18) 0.84 (0.67–1.07) 0.15 — —
Not in walking distance 50 (16.50) 1.11 (0.83–1.48) 0.50 — —
Community aesthetics Aesthetically appealing 169 (55.78)
Not aesthetically appealing 134 (44.22) 1.00 (0.79–1.26) 0.97 — —
Safety/pedestrian friendly Yes 174 (57.43)
No 129 (42.57) 1.09 (0.86–1.39) 0.46 — —
Garbage present Little/none 214 (70.63)
Some/many 89(29.37) 1.22 (0.95–1.58) 0.13 1.30 (1.00–1.68) 0.05
Graffiti None/little 284 (93.73)
Some/many 19 (6.27) 0.54 (0.30–0.96) 0.04 0.49 (0.27–0.88) 0.04
Sidewalk completeness Complete 138 (45.54)
Partial 102 (33.66) 0.93 (0.70–1.22) 0.85 — —
None 61 (20.13) 1.05 (0.77–1.43)
Streetlights on walking route Complete 109 (35.97)
Some 146 (48.18) 0.90 (0.69–1.16) 0.54 — —
None 46 (15.18) 0.92 (0.64–1.32)
Pedestrian density None 30 (9.90)
Low 179 (59.08) 1.01 (0.67–1.53) 0.72 — —
Moderate/high 94 (31.02) 0.95 (0.61–1.49)
Crosswalks Yes 153 (50.50)
No 150 (49.50) 1.11 (0.87–1.41) 0.39 — —
Some paved 112 (36.96) 0.86 (0.66–1.13) 0.82 — —
Dirt/gravel 60 (19.80) 1.10 (0.79–1.53)

Variance estimate Variance partition coefficient (%) P-value for between community variation: likelihood ratio test (σ2u0 = 0)

Unconditional model 0.12 3.64 0.02
Micro-level variables 0.14 4.10 0.02
Micro- and meso-level variables—final model 0.12 3.64 0.04
Variable Variable indicator Number of communities (%)a Univariableb

Multivariablec

OR (95% CI) P OR (95% CI) P
Not in walking distance 97 (32.01) 1.09 (0.86–1.40) 0.47 — —
Not in walking distance 104 (34.32) 1.14 (0.90–1.45) 0.28 — —
Not in walking distance 79 (26.07) 1.02 (0.79–1.31) 0.87 — —
Not in walking distance 146 (48.18) 0.84 (0.67–1.07) 0.15 — —
Not in walking distance 50 (16.50) 1.11 (0.83–1.48) 0.50 — —
Community aesthetics Aesthetically appealing 169 (55.78)
Not aesthetically appealing 134 (44.22) 1.00 (0.79–1.26) 0.97 — —
Safety/pedestrian friendly Yes 174 (57.43)
No 129 (42.57) 1.09 (0.86–1.39) 0.46 — —
Garbage present Little/none 214 (70.63)
Some/many 89(29.37) 1.22 (0.95–1.58) 0.13 1.30 (1.00–1.68) 0.05
Graffiti None/little 284 (93.73)
Some/many 19 (6.27) 0.54 (0.30–0.96) 0.04 0.49 (0.27–0.88) 0.04
Sidewalk completeness Complete 138 (45.54)
Partial 102 (33.66) 0.93 (0.70–1.22) 0.85 — —
None 61 (20.13) 1.05 (0.77–1.43)
Streetlights on walking route Complete 109 (35.97)
Some 146 (48.18) 0.90 (0.69–1.16) 0.54 — —
None 46 (15.18) 0.92 (0.64–1.32)
Pedestrian density None 30 (9.90)
Low 179 (59.08) 1.01 (0.67–1.53) 0.72 — —
Moderate/high 94 (31.02) 0.95 (0.61–1.49)
Crosswalks Yes 153 (50.50)
No 150 (49.50) 1.11 (0.87–1.41) 0.39 — —
Some paved 112 (36.96) 0.86 (0.66–1.13) 0.82 — —
Dirt/gravel 60 (19.80) 1.10 (0.79–1.53)

Variance estimate Variance partition coefficient (%) P-value for between community variation: likelihood ratio test (σ2u0 = 0)

Unconditional model 0.12 3.64 0.02
Micro-level variables 0.14 4.10 0.02
Micro- and meso-level variables—final model 0.12 3.64 0.04

a: Where percent of communities total <100% due to incomplete data

b: Univariable estimate obtained from two-level random intercept model adjusted for micro-level variables

c: Multivariable estimates obtained from two-level random intercept model adjusted for clustering of obesity within communities

### Meso-level factors

Of the 3082 respondents in the 333 sampling units included in the community profiles, 124 (4.0%) were missing height and 133 (4.3%) missing weight. Analysis of the meso-level factors was therefore limited to 2899 (94.1%) individuals living in 333 PSUs. The mean age and BMI were 42.0 years (SD 17.1) and 25.4 kg/m2 (SD 4.3) for male participants (N = 1296) and 43.7 years (SD 17.1) and 25.3 kg/m2 (SD 5.3) for female participants (N = 1603).

There was strong evidence for variability in the prevalence of obesity between communities. After accounting for micro-level variability in the odds of obesity (gender/age/marital status/household financial situation/frequency of alcohol consumption/smoking status/self-reported health/education), 4.1% of the residual variation in the propensity to be obese was attributable to unobserved community characteristics (σ2u0 = 0.14, LRT σ2u0 = 0; P = 0.02) (tables 3 and 4).

Baseline characteristics of the nutritional and physical environments and their association with obesity are provided in tables 3 and 4, respectively. After accounting for micro-level variables, obesity was inversely associated with presence of graffiti [OR 0.54 (95% CI 0.30–0.96)]. In the multivariate analysis, individuals living in communities with increased presence of garbage were 30% more likely to be obese [OR 1.30 (95% CI 1.00–1.68)]. Presence of graffiti was associated with a halving of the probability of being obese [OR 0.49 (95% CI 0.27–0.88)].

## DISCUSSION

To the best of our knowledge, this is the first study of the combined effect of micro- and meso-level influences on obesity in the fSU using standardized data collection methods. This study confirms that the obesity epidemic has spread throughout the fSU. Regional estimates for excess weight (BMI ≥25 kg/m2) are comparable with WHO 2010 estimates for countries such as Denmark (48.6%) and Switzerland (46.3%), yet still below the UK (66.3%) and USA (77.1%).32 The observation that males are more likely to be overweight yet females are more likely to be obese is in line with what is seen in other developed countries and is consistent with other findings from Russia.9,24 Although the study did not document trends over time, previous analysis in Russia has noted an increasing prevalence of obesity (from 20% in 1994 to 28% in 2004).24 Given Russia’s status as a trend setter in the fSU, similar trends within other fSU countries seem likely.

Of the micro-level factors, the findings that age and gender are strong predictors of obesity are consistent with other results from developed and transition economies.1,9,29,32,33 For both genders, the positive relationship with marital status24,25,32,34,35 and inverse relationship with self-reported health have been observed previously with obesity.3638 However, self-reported health had a significantly greater influence for females than males in our study.

The variation in the association between socio-economic indices (education and household financial situation) and obesity reflects Monteiro’s observation of the complex social determination of obesity in transitioning countries. For males, income is more likely to be positively associated with obesity, whereas education usually shows an inverse association.39 However, our findings suggested the opposite, as increasing education in males was a strong predictor of obesity. This could potentially reflect the shift from skill to knowledge-based economies in the fSU and the sedentary nature of occupations requiring higher level educational qualifications.24,27 Moreover, household financial situation was not significantly associated with obesity in males, which is consistent with the weaker associations observed between economic status and obesity in males in higher income countries (where instead education and occupation appear to matter more).40 The link between the economic shift and rising prevalence of overweight and obesity experienced in the fSU has also been observed in other transitional nations including China41 and Brazil39 but less so in India.42

The clustering of alcohol intake with obesity observed in males is consistent with the Western world picture,6 but is particularly important in terms of population attributable risk given the high rates of male alcohol consumption in the study countries and regions.43,44

For females, the observed associations with socio-economic status and education may reflect a levelling off in the rapid increase of obesity experienced in the initial stages of economic transition—some individuals reach a level of economic and educational success that enables them to adopt alternative, health-seeking lifestyles.39 Women who were situated at the highest level of education and household finances were less likely to be obese than those at lower levels, providing a contrast to male counterparts. This gender difference also reflects previous findings that females tend to shift lifestyle patterns faster due to stronger cultural and societal pressures on females than on males.39,40,45

Variation in obesity in this study appears to be largely driven by micro-level factors. There are three main potential explanations. One is that the instrument was failing to capture relevant community-level variables. This is an area that is still, relatively speaking, in its infancy, and there are still many challenges to be overcome. A second may be the limited power of the multi-level analysis as we only had community-level data from a subset of PSUs. However, a third may be the relative homogeneity of communities in many parts of the fSU, reflecting the centrally planned urban housing projects and infrastructure during Soviet times.

The finding that increased presence of garbage increases the probability of obesity is consistent with the hypothesis that perceived aesthetics can influence engagement in physical activity and consequent risk of obesity. Ellaway46 found a negative association between presence of neighborhood garbage and adult obesity in the UK. In contrast to our findings for graffiti, Ellaway46 observed a negative association between graffiti and adult obesity in the UK.

### Study limitations

The study relied on self-reported height and weight, and reporting bias with self-reported BMI is well documented; individuals generally underreport weight, overreport height and subsequently underestimate BMI and so potentially attenuate the strength of observed associations with obesity.47 However, it should be noted that our estimates of BMI are similar to WHO’s estimates for 201029 [the largest difference among males are in Georgia (26.9% in HITT vs. 24.3% from WHO) and among females in Azerbaijan (24.9% vs. 26.4%) and in Belarus (25.2% vs. 27.7%)].

The cross-sectional design means that it is not possible to determine causality, and obesity may shape individual attitudes and behaviours or healthier individuals may be attracted to live in certain communities. The study also does not take into account obesogenic work and social environments if outside the selected communities.

The veracity of self-reported educational status and the subjectivity of the household economic status may potentially bias results. For the latter, we preferred it over more objective measures, such as household asset scores as they may not capture the current economic situation—especially in a region that has been subject to sharp economic fluctuations in recent years. Asset scores can also be problematic in comparative multi-country studies.

While the community observation tool has shown high reliability and feasibility in a range of countries,12 it has not yet been validated in the HITT study countries and future work should take place to validate it in the fSU countries. The limited evidence in this study for meso-level associations with obesity suggests there may be additional characteristics not captured in the community observations. Residual confounding due to the inability to account adequately for multiple meso-level dimensions is also well recognized.2,4,68,14 The role of community deprivation and interaction with micro-level SES was also not explored.

## Conclusions

This study records the scale of overweight and obesity throughout much of the fSU and sheds light on both micro- and meso-level influences on these outcomes. The findings suggest a similar obesity profile to countries in Western Europe and North America, and associations between economic status and obesity also reflect those in other transitional nations, such as China and Brazil. The strong association for men of alcohol consumption with obesity is particularly important given the high rates of male alcohol consumption in the study countries. Further research is required to better understand how such factors influence health behaviours so as to inform effective public health interventions.

## Supplementary Data

Supplementary Data are available at EURPUB online.

## Funding

The HITT Project was funded by the European Union’s 7th Framework Program; project HEALTH-F2-2009-223344. The European Commission cannot accept any responsibility for any information provided or views expressed.

Conflicts of interest: None declared.

Key points

• The prevalence of obesity appears to be spreading throughout the fSU and is now comparable to countries in Western Europe (but not yet at the levels of the UK or USA).

• The findings suggest the drivers of obesity may be shifting from those observed in developing and transitional economies to those observed in more developed economies.

• Variation in obesity appears to be largely driven by micro-level factors rather than meso-level factors.

## Acknowledgements

We are grateful to all members of the HITT project study teams who participated in the coordination and organization of data collection for this article.

## References

1
Kelly
T
Yang
W
Chen
CS
, et al.  .
Global burden of obesity in 2005 and projections to 2030
Int J Obes
,
2008
, vol.
32
(pg.
1431
-
7
)
2
Butland
B
Jebb
S
Kopelman
P
, et al.  .
Tackling Obesities: Future Choices: Project Report
,
2007
London
UK Governnent Office for Science
3
Cecchini
M
Sassi
F
Lauer
JA
, et al.  .
Tackling of unhealthy diets, physical inactivity, and obesity: health effects and cost-effectiveness
Lancet
,
2010
, vol.
376
(pg.
1775
-
84
)
4
Chow
CK
Lock
K
Teo
K
, et al.  .
Environmental and societal influences acting on cardiovascular risk factors and disease at a population level: a review
Int J Epidemiol
,
2009
, vol.
38
(pg.
1580
-
94
)
5
Egger
G
Swinburn
B
An “ecological” approach to the obesity pandemic
BMJ
,
1997
, vol.
315
(pg.
477
-
80
)
6
Swinburn
BA
Sacks
G
Hall
KD
, et al.  .
The global obesity pandemic: shaped by global drivers and local environments
Lancet
,
2011
, vol.
378
(pg.
804
-
14
)
7
Jones
A
Bentham
G
Foster
C
, et al.  .
Foresight: Tackling Obesities: Future Choices—Obesogenic Environments—Evidence Review
,
2007
London
UK Government Office for Science
8
Kirk
SFL
Penney
TL
McHugh
TLF
Characterizing the obesogenic environment: the state of the evidence with directions for future research
Obes Rev
,
2010
, vol.
11
(pg.
109
-
17
)
9
Sassi
F
Devaux
M
Cecchini
M
Rusticelli
E
The Obesity Epidemic: Analysis of Past and Projected Future Trends in Selected OECD Countries

OECD Health Working Papers, 2009:1–82
10
Swinburn
B
Egger
G
Raza
F
Dissecting obesogenic environments: the development and application of a framework for identifying and prioritizing environmental interventions for obesity
Prev Med
,
1999
, vol.
29
(pg.
563
-
70
)
11
Cerin
E
Frank
LD
Sallis
JF
, et al.  .
From neighborhood design and food options to residents’ weight status
Appetite
,
2011
, vol.
56
(pg.
693
-
703
)
12
Chow
CK
Lock
K
M
, et al.  .
Environmental Profile of a Community's Health (EPOCH): an instrument to measure environmental determinants of cardiovascular health in five countries
PLoS One
,
2010
, vol.
5
pg.
e14294

13
Giskes
K
van Lenthe
F
Avendano-Pabon
M
Brug
J
A systematic review of environmental factors and obesogenic dietary intakes among adults: are we getting closer to understanding obesogenic environments?
Obes Rev
,
2011
, vol.
12
(pg.
e95
-
e106
)
14
Papas
MA
Alberg
AJ
Ewing
R
, et al.  .
The built environment and obesity
Epidemiol Rev
,
2007
, vol.
29
(pg.
129
-
43
)
15
Schmidhauser
S
Wieser
S
Brugger
U
Environmental Determinants of Overweight and Obesity: Extended International Literature Review. Final Report
,
2009
Zurich
Federal Office of Public Health
16
Zick
CD
Smith
KR
Fan
JX
, et al.  .
Running to the store? The relationship between neighborhood environments and the risk of obesity
Soc Sci Med
,
2009
, vol.
69
(pg.
1493
-
500
)
17
Rose
D
Hutchinson
P
Bodor
N
, et al.  .
Neighborhood food environments and body mass index. The importance of in-store contents
Am J Prev Med
,
2009
, vol.
37
(pg.
214
-
9
)
18
Lachowycz
K
Jones
AP
Greenspace and obesity: a systematic review of the evidence
Obes Rev
,
2011
, vol.
12
(pg.
e183
-
e189
)
19
Giles-Corti
B
Donovan
RJ
Socioeconomic status differences in recreational physical activity levels and real and perceived access to a supportive physical environment
Prev Med
,
2002
, vol.
35
(pg.
601
-
11
)
20
Catlin
TK
Simoes
EJ
Brownson
RC
Environmental and policy factors associated with overweight among adults in Missouri
Am J Health Promot
,
2003
, vol.
17
(pg.
249
-
58
)
21
Poortinga
W
Perceptions of the environment, physical activity, and obesity
Soc Sci Med
,
2006
, vol.
63
(pg.
2835
-
46
)
22
Branca
F
Nikogosian
H
Lobstein
T
The Challenge of Obesity in the WHO European Region and the Strategies for Response: Summary
,
2007
Geneva
World Health Organization
23
Gortmaker
SL
Swinburn
BA
Levy
D
, et al.  .
Changing the future of obesity: science, policy, and action
Lancet
,
2011
, vol.
378
(pg.
838
-
47
)
24
Huffman
SK
Rizov
M
Determinants of obesity in transition economies: the case of Russia
Econ Hum Biol
,
2007
, vol.
5
(pg.
379
-
91
)
25
Stillman
S
Health and nutrition in Eastern Europe and the former Soviet Union during the decade of transition: a review of the literature
Econ Hum Biol
,
2006
, vol.
4
(pg.
104
-
46
)
26
Hinote
BP
Cockerham
WC
Abbott
P
Psychological distress and dietary patterns in eight post-Soviet republics
Appetite
,
2009
, vol.
53
(pg.
24
-
33
)
27
Jahns
L
Baturin
A
Popkin
BM
Obesity, diet, and poverty: trends in the Russian transition to market economy
Eur J Clin Nutr
,
2003
, vol.
57
(pg.
1295
-
302
)
28
Pomerleau
J
Pudule
I
Grinberga
D
, et al.  .
Patterns of body weight in the Baltic Republics
Public Health Nutr
,
2007
, vol.
3
(pg.
3
-
10
)
29
WHO Global InfoBase team
The SuRF Report 2. Surveillance of Chronic Disease Risk Factors: Country-Level Data and Comparable Estimates
,
2005
Geneva
World Health Organization
30
Leckie
G
Module 7: Multi-Level Models for Binary Response. Stata Practical

LEMMA VLE. Bristol: Centre for Multilevel Modelling, 2010
31
d'Hombres
B
Rocco
L
Suhrcke
M
, et al.  .
The influence of social capital on health in eight former Soviet countries: why does it differ?
J Epidemiol Community Health
,
2009
, vol.
65
(pg.
44
-
50
)
32
WHO
Global Status Report on Noncommunicable Diseases 2010: Description of the Global Burden of NCDs, Their Risk Factors and Determinants
,
2011
Geneva
World Health Organization
33
Berghöfer
A
Pischon
T
Reinhold
T
, et al.  .
Obesity prevalence from a European perspective: a systematic review
BMC Public Health
,
2008
, vol.
8
(pg.
1
-
10
)
34
Selassie
M
Sinha
AC
The epidemiology and aetiology of obesity: a global challenge
Best Pract Res Clin Anaesthesiol
,
2011
, vol.
25
(pg.
1
-
9
)
35
WHO
Obesity: Preventing and Managing the Global Epidemic: Report of a WHO Consultation
,
2000
Geneva
World Health Organization
36
Ford
ES
Moriarty
DG
Zack
MM
, et al.  .
Self-reported body mass index and health-related quality of life: findings from the Behavioral Risk Factor Surveillance System
Obesity
,
2001
, vol.
9
(pg.
21
-
31
)
37
Jia
H
Lubetkin
EI
The impact of obesity on health-related quality-of-life in the general adult US population
J Public Health
,
2005
, vol.
27
(pg.
156
-
64
)
38
Martínez
J
Kearney
J
Kafatos
A
, et al.  .
Variables independently associated with self-reported obesity in the European Union
Public Health Nutr
,
2007
, vol.
2
(pg.
125
-
33
)
39
Monteiro
CA
Conde
WL
Popkin
BM
Independent effects of income and education on the risk of obesity in the Brazilian adult population
J Nutr
,
2001
, vol.
131
(pg.
881S
-
886S
)
40
McLaren
L
Socioeconomic status and obesity
Epidemiol Rev
,
2007
, vol.
29
(pg.
29
-
48
)
41
Wang
Y
Mi
J
Shan
XY
, et al.  .
Is China facing an obesity epidemic and the consequences? The trends in obesity and chronic disease in China
Int J Obes
,
2007
, vol.
31
(pg.
177
-
88
)
42
Wang
Y
Chen
HJ
Shaikh
S
Mathur
P
Is obesity becoming a public health problem in India? Examine the shift from under- to overnutrition problems over time
Obes Rev
,
2009
, vol.
10
(pg.
456
-
74
)
43
Pomerleau
J
Gilmore
A
McKee
M
, et al.  .
Determinants of smoking in eight countries of the former Soviet Union: results from the Living Conditions, Lifestyles and Health Study
,
2004
, vol.
99
(pg.
1577
-
85
)
44
Pomerleau
J
McKee
M
Rose
R
, et al.  .
Hazardous alcohol drinking in the former Soviet Union: a cross-sectional study of eight countries
Alcohol Alcohol
,
2008
, vol.
43
(pg.
351
-
9
)
45
Moore
S
Hall
JN
Harper
S
Lynch
JW
Global and national socioeconomic disparities in obesity, overweight, and underweight status
J Obes
,
2010
, vol.
2010
(pg.
1
-
11
)
46
Ellaway
A
Graffiti, greenery, and obesity in adults: secondary analysis of European cross sectional survey
BMJ
,
2005
, vol.
331
(pg.
611
-
2
)
47
Gorber
SC
Tremblay
M
Moher
D
Gorber
B
A comparison of direct vs. self-report measures for assessing height, weight and body mass index: a systematic review
Obes Rev
,
2007
, vol.
8
(pg.
307
-
26
)