CC BY
37
ТС PH
ЕЕ
Tobacco Control and Public Health in Eastern Europe
Chagarna N. S. & Andreeva T. I. (2014). Lifestyle correlates of overweight and obesity among the population of Ukraine. Tobacco Control and Public Health in Eastern Europe, 4(1), 15-28. doi: 10.6084/m9.figshare. 1060345
Lifestyle correlates of overweight and obesity among the population of Ukraine
Natalia S. Chagarna, Tatiana I. Andreeva
BACKGROUND: Overweight and obesity are among the greatest health challenges nowadays. This study aimed to investigate correlates of overweight among the population of Ukraine.
METHODS: Data from the survey "Health and Well-Being in Transition Societies" (2000) were analyzed. Body mass index (BMI) based on self-reported height and weight was used to estimate the extent of overweight both in continuous and di-chotomous form. Binary dependent variable was computed by setting BMI of 25 as the borderline between normal weight and overweight. Education, demographic characteristics, field of activity, level of physical activity, behavioral and eating habits were considered as independent variables.
RESULTS: In 2000, 44% of Ukrainian population (39% of men and 49% of women) were overweight. Risk of overweight increased with age: among those below 40 years old, 16% of men and 12% of women were overweight, about 47% of men and 60% of women were overweight in 40+ groups. Among young men (under 40), those con-
suming more fruit (0R=0,31) and more physically active (0R=0,43) were less likely to be overweight. In women over 60, those overweight were more physically active (OR=2,32). In men over 40 and women below 40, prevalence of overweight was lower in smokers than in non-smokers. Eating potatoes, meat and milk was associated with increased BMI in some groups. No consistent associations were found for consumption of fish, butter, oil, sugar, and vegetables, as well as for education.
CONCLUSION: In 2000, Ukrainian population did not bear patterns of obesity epidemics found in the West: no associations with either socioeconomic status or foods usually associated with overweight were found. Overweight was more likely related to increased use of traditional foods like dairy products, meat and potatoes. Smoking was related to lower BMI in populations with high prevalence of smoking.
KEYWORDS: overweight; obesity; BMI; eating patterns; smoking; alcohol use.
Поведенческие корреляты избыточной массы тела и ожирения среди населения Украины
Чагарна Н.С., Андреева Т.И.
АКТУАЛЬНОСТЬ: Избыточный вес (ИВ) и ожирение является одной из наиболее серьезных проблем для здоровья. Цель нашего исследования заключалась в определении поведенческих коррелят ИВ среди населения Украины.
МЕТОДЫ: Проанализированы данные опроса "Здоровье и благополучие в переходных обществах" (2000). Для оценки степени избыточности массы тела использовался индекс массы
тела (ИМТ), рассчитанный с использованием сообщенных респондентами массы и роста. Данные проанализированы методами логистической регрессии и генерализованной линейной модели (GLM). Бинарная зависимая переменная сформирована путем установления ИМТ=25 в качестве границы между ИВ и нормальным весом. Независимыми переменными являлись уровень образования, демографические характеристики,
сфера деятельности, уровень физической активности, поведенческие и пищевые привычки.
РЕЗУЛЬТАТЫ: В 2000 году у 44% населения Украины (39% мужчин и 49% женщин) был ИВ. Риск ожирения увеличивается с возрастом: среди респондентов моложе 40 лет, 16% мужчин и 12% женщин имели ИВ, в то время как в группе 40 + ИВ присущ около 47% мужчин и 60% женщин. Среди молодых мужчин (до 40 лет) те, кто употребляет больше фруктов (ОК^^) и физически активны (ОК = 0,43), были менее склонны к ИВ. Среди женщин старше 60 лет имеющие ИВ чаще занимались физическими упражнениями (ОК = 2,32). Распространенность ИВ среди мужчин старше 40 лет и женщин моложе 40, которые курят, была ниже, чем среди тех, кто не употребляет табак. Для употребляющих картофель, мясо и молоко характерен повышенный ИМТ. Не было обнару-
жено устойчивых ассоциаций для потребления рыбы, сливочного и растительного масла, сахара, овощей, а также для уровня образования.
ЗАКЛЮЧЕНИЕ: В 2000 году в Украине не было обнаружено основных коррелят возникновения ИВ, характерных для стран Запада: не было найдено ассоциации с социально-экономическим статусом или употреблением продуктов, как правило, связанных с избыточным весом. ИВ был с большей вероятностью связан с частым употреблением «традиционных» продуктов питания, таких как молочные продукты, мясо и картофель. Курение связано с низким ИМТ в группах населения с высокой распространенностью курения.
КЛЮЧЕВЫЕ СЛОВА: избыточный вес; ожирение; индекс массы тела ИМТ; структура питания; курение; употребления алкоголя.
Поведiнковi кореляти надмiрноí ваги i ожиршня серед населення Украши
Чагарна Н.С., Андреева Т.1.
АКТУАЛЬН1СТЬ: Надмiрна вага (НВ) i ожиршня е одшею з найбтьш серйозних проблем для здо-ров'я. Це дослщження мало на мет визначити поведiнковi кореляти НВ серед населення УкраТни.
МЕТОДИ: Проаналiзовано дан опитування "Здо-ров'я та благополуччя в перехщних суспть-ствах" (2000). Для оцшки ступеня надлишковоТ маси тта використано шдекс маси тта (1МТ), розрахований з використанням даних респондент щодо ваги i зросту. Даш проаналiзовано методами лопстично'Т регресп та генералiзованоТ лшшноТ моделi (GLM). Бшарну залежну змшну обчислено шляхом встановлення 1МТ = 25 як межу мiж нормальною вагою та НВ. Основними незалежними змшними були рiвень осв^и респондент, демографiчнi характеристики, сфера дiяльностi, рiвень фiзичноТ активносп, поведш-ковi i харчовi звички.
РЕЗУЛЬТАТИ: У 2000 роц 44 % населення УкраТни (39 % чоловшв i 49 % жшок) мали НВ. Ризик НВ збтьшуеться з вком: серед респондент молодше 40 роюв 16% чоловшв i 12% жiнок мали НВ, у той час як у груш 40 + НВ була притаманна близько 47% чоловшв i 60% жшок. Серед молодих чоловшв (до 40 роюв) т^
хто споживае бтьше фрую1в (ОК = 0,31) i бiльш фiзично активы (ОК = 0,43), менш схильш до НВ. Жiнки старше 60 роюв з НВ бтьш фiзично активнi (ОК = 2,32). Поширешсть НВ серед чоло-вiкiв старше 40 роюв i жiнок молодше 40, яю ку-рять, була нижчою, нiж серед некурцiв. Вжи-вання картоплi, м'яса i молока було пов'язано з пщвищеним 1МТ в деяких групах. Не виявлено асощацш для споживання риби, вершкового масла, олп, цукру, овочiв, а також для рiвня освiти.
ВИСНОВОК: У 2000 роц в УкраТн не було виявлено основних ризиюв виникнення ожирiння, характерних для краТн Заходу: не було знай-дено асоцiацiТ з соцiально-економiчним статусом або вживанням продук^в, як правило, пов'яза-них з надмiрною вагою. НВ з бтьшою ймовiрнi-стю пов'язана з частим вживанням «традицш-них» продуктiв харчування, таких як молочш продукти, м'ясо i картопля. Куршня було пов'язано з низьким шдексом маси тiла в групах населення з високою поширенiстю куршня.
КЛЮЧОВ1 СЛОВА: надмiрна вага; ожиршня; ш-декс маси тiла 1МТ; структура харчування; ку-рiння; вживання алкоголю.
introduction
The attention to the worldwide epidemic of obesity, called "globe-sity", is caused by its being the fifth leading risk factor for global deaths (Haslam & James, 2005). Obesity is associated with increased risk of death from all causes as well as from cancer and cardiovascular diseases (Ma, Flanders, Ward, & Jemal, 2011) and with incidence of type II diabetes, all cancers except esophageal (in females), pancreatic and prostate cancer, all cardiovascular diseases (except congestive heart failure), asthma, gallbladder disease, os-teoarthritis and chronic back pain (Guh et al., 2009).
The obesity epidemic is recognized to be a result of changes in energy intake and/or energy expenditure that have led to energy imbalance in a large portion of the population (Crawford & Ball, 2002). Environmental changes lead to obesity through behaviors related to energy intake and expenditure; thus researchers consider determinants of eating patterns, physical activity, and sedentary behaviors instead. However, these may differ by population group, including gender and age groups.
Different dietary regimens can affect body weight, e.g. increased fruit and vegetable intake resulted in greater reduction of weight than limited intake of high-fat, low-nutrient dense foods with controlled physical activity (Epstein et al., 2001). Environmental factors such as food availability on the one hand, and psychosocial factors including eating traditions and beliefs regarding food taste and health impact, on the other, may also contribute to the country-specific eating patterns responsible for lower or higher risk of developing
obesity in a particular population. Apart from traditional patterns, healthy or unhealthy eating practices of a person as well as level of individuals' physical activity could be affected by policy-related factors, such as promotion and availability of high-energy food or implementation of population-based program aimed to change life style. Social and environmental factors considered at four levels including (1) international, (2) national/regional, (3) community/locality, (4) work/school/home (Huang, Drewnosksi, Kumanyika, & Glass, 2009) might have a more profound effect on body weight status than do individuals' characteristics (Wang & Beydoun, 2007). Identifying such factors is needed to inform healthy public policies.
With plenty of studies devoted to predictors of overweight and obesity in other countries, we could not identify such studies among adult population in Ukraine. Just one study among children indicated that excessive BMI was positively associated with the family income and the mother's weight (Friedman et al., 2009). Cross-country comparison of obesity prevalence and weight management practices among adolescents (Kalabiska et al., 2010; Page et al., 2006) classified 7% of Ukrainian compared to 10% of Hungarian adolescents as overweight. BMI was considered as a predictor of mortality in Ukrainian diabetic patients (Kha-langot, Tronko, Kravchenko, Kulchinska, & Hu, 2009). Just two studies published internationally considered eating patterns in Ukraine (Biloukha & Utermohlen, 2000; Bilukha & Utermohlen, 2002). Taking into account this lack of evidence we decided to
conduct an exploratory study with secondary analysis of existing data collected in 2000.
With this study we aimed to address several exploratory research questions:
1. How widespread was overweight in 2000 in Ukraine among different age, gender, and occupation groups?
2. Was overweight more or less widespread in groups of Ukrainians characterized by certain health-related behaviors including tobacco smoking and alcohol use?
3. Was overweight positively or negatively associated with the self-reported level of physical activities (namely, physical exercises or sports reported to be done sometimes or never)?
4. Are there any specific eating patterns/food preferences in Ukrainian population associated with higher/lower values of body mass index?
methods
The present analysis is based on the data from the survey "Health and Well-Being in Transition Societies", conducted in 2000 by the Kyiv International Institute of Sociology (Paniotto, Martsinkiv, Kharchenko, Zakhozha, & Buslayeva, 2004) and dedicated to study interactions between the level of material well-being and the health of households in transition societies. Data were collected through structured interviews with 1635 adults (720 - men, 915 -women,) in 77 settlements (39 cities, 9 towns and 29 villages) of Ukraine. Participants were chosen randomly both from rural and urban areas from all 26 regions (oblasts) of Ukraine.
Sampling
The main consideration when developing the sample was to ensure that each household had an equal probability of being selected for the interview. For this research, multistage sample random at each step of selection was used.
First stage of the sampling included the selection of settlements. The sample was drawn in such a fashion that each oblast was proportionally represented. All citizens in each oblast are officially considered to live in one of three kinds of population points: cities, towns, and villages. The sample was strati-fled by the type of settlement (Pan-iotto et al., 2004). Second stage of sampling included the selection of postal districts. Third stage included selection of households and respondents. One respondent from each of selected households was interviewed.
Measurements
The questionnaire was composed of questions concerning demographics, work sector, behavioral and eating habits, as well as socioeconomic and health status. The data concerning eating patterns, life style and socioeconomic status of respondents were self-reported, as well as their weight and height. More detailed information on sampling, assessment method, etc. can be found in the survey primary outcome paper (Paniotto et al., 2004).
The main dietary components were presented in the questionnaire not individually but grouped in wider general categories, such as meat products, fruits or milk products. Separate question indicated main drink at meals, which also included milk (whole, non-packed) as an alternative choice. Frequency of con-
sumption of each group of products was presented as "often" if consumed four or more times per week, "rare" when products were consumed one- three times per week, or "never". Some variables were described by two characteristics: "often" and "rare" (including absence of consumption during week).
Level of physical activity was defined as any type of physical exercises or sports done by respondent (sometimes) or not done (never).
Data Analysis
To address our research questions, we considered as potential determinants variables present in the dataset that characterized socio-de-mographic status (level of education, working status of respondent), unhealthy behaviors (alcohol and tobacco use), and dietary data.
Body mass index (BMI) was considered the main outcome roughly estimating total body fat mass. The BMI was calculated using formula:
BMI = Weight in Kilograms /
(Height in Meters)2
As the BMI rises above 25.0, the risk of obesity-related diseases increases significantly. Thus one of the approaches we used was to consider dichotomized BMI based on overweight status. Binary outcome variable was computed by setting BMI of 25 as the borderline between normal and excess weight. Tests of possible associations were first performed in a bivariate analysis with the Pearson Chi-square test of independence where normal/excess weight was analyzed as binary dependent variable, and all the independent variables as categorical variables. All the associations identified within the bivariate analysis with Pearson Chi-square were then
included in the multivariate logistic regression models to adjust for the potential confounding effects. As both gender and age were found to significantly modify the revealed associations, logistic regression analysis was stratified by gender and age to control for effect modification. We stratified the sample into three age groups: young - people aged up to 40, middle age group - 40-60 year old respondents, and old - those over 60. All p-values are two-sided, and the level of statistical significance was set to be 0.05.
Apart from logistic regression analysis with binary outcome, we performed General Linear Model analysis with BMI as continuous outcome. Analysis was conducted stratified by gender. Independent variables were included in the model consecutively and excluded if no association was found for either gender. After main effects of all the variables significantly associated in bivariate analysis have been tested in the multivariate model, their interactions with age were added in the model. Those interactions that did not show significant associations were excluded as well as main effects that became insignificant after interactions were taken into account. Results are presented as Intercept equal to BMI in those who belong to reference categories of all variables included in the model ± b-coefficients corresponding to those categories which showed significant difference for the reference categories.
As multiple comparisons could lead to chance findings, we considered these two regression models and performed them in stratified sample in order to observe whether certain associations are more con-
sistent and reproducible across the socio-demographic groups and types of analysis used.
All data analyses were accomplished with SPSS for Windows.
results
Respondents (720 men, 915 women) were 15-92 years old while the average age was 47. Many of young and middle-aged men and women were wage earners; smaller groups were unem-
ployed and self-employed, most of elderly people were pensioners. One in seven respondents had university education. Over half of men and less than 9% of women were smokers; two in three men and three in four women reported to never do physical exercises.
Overall, 44% of study participants were overweight or obese: 39% of men and 49% of women. Younger people were less likely overweight (16% of men and 12% of women below 40 years) than those in older
age groups. Overweight in men and women was found to be associated with different variables. While in both middle and old age men the percentage of those overweight was quite similar and below half, percentage of overweight women increased dramatically with age: 53% had excess weight in middle-age group and 65% in old age. Sample characteristics together with results of bivariate analysis are presented in Table 1.
Table 1. socio-demographic, lifestyle, and food consumption characteristics of the studied group, by gender: number of respondents, percent overweight, chi-square p-value, ukraine, 2000
Univariate distributions Bivariate distributions
Men Women All Men Women
N % N % N % % with BMI>25 Sig. % with BMI>25 Sig.
Education
Primary 359 49,9 486 53,1 845 51,7 36,2 0,41 48,5 0,94
Special technical 243 33,8 276 30,2 519 31,7 41,0 49,7
Higher 118 16,4 153 16,7 271 16,6 40,7 49,1
Main field of activity
Wages 281 39,0 289 31,6 570 34,9 38,9 <0,05 51,7 <0,05
Pensioner 243 33,8 429 46,9 672 41,1 47,8 63,8
Unemployed 113 15,7 109 11,9 222 13,6 33,6 33,3
Other 83 11,5 88 9,6 171 10,5 18,4 8,9
Age group
Young <40 310 43,1 325 35,5 635 38,8 15,7 <0,05 12,3 <0,05
Middle age 40-60 199 27,6 249 27,2 448 27,4 46,6 53,4
Old >60 211 29,3 341 37,3 552 33,8 47,7 65,3
Beer consumption
Often 101 14,0 32 3,5 133 8,1 38,4 0,3 37,1 <0,05
Rare 348 48,3 226 24,7 574 35,1 32,7 20,0
Never 271 37,6 657 71,8 928 56,8 40,6 54,9
Vodka consumption
Often 132 18,3 22 2,4 154 9,4 37,8 0,9 41,7 0,53
Rare 437 60,7 478 52,2 915 56,0 39,1 50,5
Never 151 21,0 415 45,4 566 34,6 37,2 47,8
Current smoking status
Non-smoker 329 45,7 835 91,3 1164 71,2 48,2 <0,05 51,0 <0,05
Smoker 391 54,3 80 8,7 471 28,8 30,4 29,0
Table 1. socio-demographic, lifestyle, and food consumption characteristics of the studied group, by gender: number of respondents, percent overweight, chi-square p-value, ukraine, 2000 (continued)
Univariate distributions Bivariate distributions
Men Women All Men Women
N % N % N % % with Sig. % with Sig.
BMI>25 BMI>25
Physical activity
Never 477 66,3 681 74,4 1158 70,8 41,2 0,04 53,2 <0,05
Sometimes 243 33,8 234 25,6 477 29,2 33,5 36,5
Main foods consumption:
Meat
Rare 481 66,8 694 75,8 1175 71,9 37,5 0,4 50,7 0,06
Often 239 33,2 221 24,2 460 28,1 40,7 44,1
Fish
Never 416 57,8 590 64,5 1006 61,5 36,6 0,2 49,1 0,96
Sometimes 304 42,2 325 35,5 629 38,5 41,2 48,9
Butter
Rare 539 74,9 714 78,0 1253 76,6 39,3 0,45 50 0,25
Often 181 25,1 201 22,0 382 23,4 36,2 45,7
Oil
Often 652 90,6 833 91,0 1485 90,8 39,1 0,35 49,9 0,05
Rare 68 9,4 82 9,0 150 9,2 33,3 39,6
Potato
Often 633 87,9 795 86,9 1428 87,3 37 <0,05 50,5 <0,05
Rare 87 12,1 120 13,1 207 12,7 51,1 39,7
Vegetables
Often 535 74,3 663 72,5 1198 73,3 38,4 0,87 49,9 0,32
Rare 185 25,7 252 27,5 437 26,7 39 46,4
Fruits
Rare 612 85,0 773 84,5 1385 84,7 39,4 0,33 50 0,15
Often 108 15,0 142 15,5 250 15,3 34,6 38,7
Sugar
Often 627 87,1 777 84,9 1404 85,9 36,2 0,6 55,2 0,09
Rare 93 12,9 138 15,1 231 14,1 39 47,8
Milk (dairy) products
Rare 503 69,9 638 69,7 1141 69,8 37,7 0,43 47,5 0,18
Often 217 30,1 277 30,3 494 30,2 40,7 52,0
Main drink at meals
Tea or coffee 576 80,0 741 81,0 1317 80,6 38,2 <0,05 47,5 <0,05
Water 44 6,1 61 6,7 105 6,4 21,7 38,2
Milk (whole or non-packed) 40 5,6 50 5,5 90 5,5 43,2 69,2
Other 60 8,3 63 6,9 123 7,5 50,8 59,7
Logistic regression analysis for men stratified by age groups is presented in Table 2. More associations were found in younger age group of men (aged below 40). Higher BMI was found in those
0.23-0.78), consuming more fruit (0R=0.31 95% CI 0.12-0.84) and potatoes were all associated with lower odds of being overweight. Among other age groups of men, the only factor significantly associ-
who reported high consumption of milk and dairy products (OR= 1.78 95%CI 0.94-3.38). Being a self-employed person (OR=0.28 95%CI 0.11-0.73), more physically active (OR=0.43 95% CI
Table 2. Results of multivariate binary logistic regression analysis of overweight (Body Mass Index >25) in men, stratified by age group: numbers of respondents, adjusted odds ratios and their 95% confidence intervals, ukraine, 2000
age group <40 age group 40-60 age group >60 N AOR (95% CI) N AOR (95% CI) N AOR (95% CI)
Main field of activity Wages 159 1 120 1 2
Pensioner 7 0,3 (0,0-2,6) 27 0,8 (0,3-2,2) 209
Unemployed 73 1,3 (0,7-2,5) 40 0,5 (0,2-1,2)
Self-employed, housekeeper 71 0,3 (0,1-0,7) 12 3,7 (0,9-16,0)
Main drink at meals Tea or coffee 256 1 155 1 165 1
Water 20 0,8 (0,2-3,1) 13 0,4 (0,1-1,7) 11 0,3 (0,1-1,4)
Milk (whole or non-packed) 15 1,1 (0,3-4,1) 12 0,5 (0,1-2,2) 13 1,3 (0,3-5,1)
Other 19 1,6 (0,5-5) 19 2,3 (0,8-6,8) 22 1,9 (0,7-5,4)
Education Primary 132 1 90 1 137 1
Special tech 128 1,0 (0,5-1,8) 71 1,3 (0,6-2,7) 44 1,4 (0,6-3,3)
Higher 50 0,8 (0,3-1,9) 38 1,1 (0,4-2,9) 30 1,2 (0,5-3,1)
Beer consumption Often 76 1,2 (0,4-3,1) 22 1,0 (0,3-3,6) 3 0,4 (0,0-5,9)
Rare 166 1,0 (0,4-2,5) 105 1,2 (0,5-2,6) 77 1,1 (0,6-2,3)
Never 68 1 72 1 131 1
Vodka consumption Often 60 1,2 (0,4-3,5) 48 1,5 (0,4-5,0) 24 1,3 (0,4-4,1)
Rare 195 1,0 (0,4-2,7) 119 1,7 (0,6-4,6) 123 1,3 (0,6-2,8)
Never 55 32 64
Physical activity Sometimes 159 0,4 (0,2-0,8) 47 1,7 (0,8-3,8) 37 1,4 (0,6-3,2)
Never 151 1 152 1 174 1
Smoking status Smoker 185 0,6 (0,4-1,2) 131 0,3 (0,2-0,6) 75 0,3 (0,2-0,6)
Non-smoker 125 1 68 1 136 1
Fish Never 181 0,9 (0,5-1,6) 106 1,3 (0,6-2,5) 129 1,0 (0,5-1,9)
Sometimes 129 1 93 1 82 1
Butter Rare 216 1 145 1 178 1
Often 94 1,1 (0,5-2,1) 54 1,0 (0,4-2,3) 33 1,0 (0,4-2,6)
Milk (dairy) products Rare 200 1 150 1 153 1
Often 110 1,8 (0,9-3,4) 49 1,3 (0,6-2,9) 58 0,9 (0,4-1,9)
Sugar Rare 34 0,7 (0,3-2,1) 26 0,4 (0,2-1,2) 33 1,0 (0,4-2,4)
Often 276 1 173 1 178 1
Fruit Rare 261 1 170 1 181 1
Often 49 0,3 (0,1-0,8) 29 0,8 (0,3-2,1) 30 1,0 (0,4-2,6)
Vegetables Rare 72 0,6 (0,3-1,2) 54 0,9 (0,4-2,1) 59 1,5 (0,7-3,2)
Often 238 1 145 1 152 1
Oil Rare 32 0,8 (0,3-2,1) 22 1,2 (0,4-3,7) 14 0,5 (0,1-1,7)
Often 278 1 177 1 197 1
Potato Rare 32 4,6 (1,9-11,3) 32 2,0 (0,8-5,3) 23 1,1 (0,4-2,8)
Often 278 1 167 1 188 1
Meat Rare 189 1 127 1 165 1
Often 121 1,3 (0,7-2,4) 72 0,9 (0,4-2,1) 46 1,5 (0,7-3,4)
Table 3. Results of multivariate binary logistic regression analysis of overweight (Body Mass Index >25) in women, stratified by age group: numbers of respondents, adjusted odds ratios and their 95 confidence intervals, ukraine, 2000
age group <40 age group 40-60 age group >60
N AOR (95%CI) N AOR (95%CI) N AOR (95% CI)
Main field of activity Wages 162 1 123 1 4 1
Pensioner 6 0,8 (0,1-8,4) 86 1,0 (0,5-1,9) 337 0,5 (0,0-7,2)
Unemployed 75 0,7 (0,4-1,5) 34 0,4 (0,2-0,8)
Self-employed, housekeeper 82 0,1 (0,0-0,3) 6 2,4 (0,3-23,3)
Main drink at meals Tea or coffee 281 1 199 1 261 1
Milk (whole or non-packed) 11 6,4 (1,4-30,0) 16 0,7 (0,2-2,5) 23 2,3 (0,8-7,1)
Other 12 3,2 (0,5-19,9) 24 1,0 (0,4-2,5) 27 3,0 (1,1-8,5)
Education Primary 137 1 97 1 252 1
Special technical 116 2,0 (0,9-4,4) 103 0,8 (0,4-1,5) 57 1,9 (1,0-3,7)
Higher 72 2,0 (0,8-4,9) 49 0,7 (0,3-1,6) 32 1,4 (0,6-3,6)
Beer consumption Often 23 0,1 (0,0-0,7) 9 0,4 (0,1-2,0)
Rare 123 0,4 (0,2-0,8) 77 1,4 (0,7-2,7) 26 0,3 (0,1-0,8)
Never 179 1 163 1 315 1
Vodka consumption Often 9 3,6 (0,5-24,5) 9 1,0 (0,2-5,1) 4 0,8 (0,1-8,1)
Rare 196 1,7 (0,9-3,5) 174 1,3 (0,6-2,5) 108 2,5 (1,4-4,7)
Never 120 1 66 1 229 1
Physical activity Sometimes 135 0,7 (0,4-1,5) 47 0,9 (0,4-1,8) 52 2,3 (1,1-5,0)
Never 190 1 202 1 289 1
Smoking status Smoker 57 0,4 (0,2-0,9) 20 0,6 (0,2-1,8) 3 1,3 (0,1-25,4)
Non-smoker 268 1 229 1 338 1
Main products consumed (diet)
Fish Never 205 0,9 (0,5-1,8) 148 0,8 (0,5-1,5) 237 0,9 (0,5-1,5)
Sometimes 120 1 101 1 104 1
Butter Rare 226 1 185 1 303 1
Often 99 1,0 (0,5-2,1) 64 1,5 (0,7-3,2) 38 1,6 (0,7-3,7)
Milk (dairy) products Rare 228 1 162 1 248 1
Often 97 1,0 (0,5-2,0) 87 1,9 (1,0-4,0) 93 1,1 (0,6-2,0)
Sugar Rare 37 1,8 (0,7-4,9) 35 1,4 (0,6-3,2) 66 1,3 (0,7-2,6)
Often 288 1 214 1 275 1
Fruit Rare 263 1 213 1 297 1
Often 62 1,1 (0,5-2,4) 36 0,8 (0,3-1,8) 44 0,5 (0,3-1,2)
Vegetables Rare 81 0,5 (0,2-1,1) 53 0,9 (0,4-1,8) 118 1,3 (0,7-2,2)
Often 244 1 196 1 223 1
Oil Rare 31 2,2 (0,7-6,5) 20 1,0 (0,4-2,9) 31 0,5 (0,2-1,1)
Often 294 1 229 1 310 1
Potatoes Rare 40 0,5 (0,2-1,6) 36 0,8 (0,3-1,8) 44 0,5 (0,2-0,9)
Often 285 1 213 1 297 1
Meat Rare 212 1 188 1 294 1
Often 113 1,1 (0,6-2,1) 61 0,7 (0,4-1,5) 47 0,6 (0,3-1,3)
ated with prevalence of overweight was smoking: in men aged 40-60 years prevalence of overweight in smokers was lower than in non-smokers (OR=0.31 95%CI 0.150.64), and same was true for men over 60 (0R=0.29 95%CI 0.150.56).
Logistic regression analysis for women stratified by age groups is presented in Table 3. Associations found for women were more consistent across age groups than for men. In contrast to men, significant negative association between overweight and smoking was found for women below 40 (0R=0.39 95%CI 0.17-0.92), not those above 40 as for men. In younger women, overweight was positively associated with drinking milk at meals (OR=6.44 95%CI 1.38-30.04), same relationship while non-significant was found for older women. Beer consumption was negatively associated with overweight in most age groups: 0R=0.41 (95%CI 0.21-0.79) for women below 40 and 0R=0.30 (95%CI 0.12-0.78) for older ones, on the contrary, intake of vodka was positively associated with overweight (OR=2.53 95%CI 1.37-4.66). While in younger women those who reported some physical activity were less likely overweight, in older women there was an association between overweight and physical activity (OR=2.32 95%CI 1.075.03). For potatoes, in all age groups we see either non-significant or significant association between overweight and eating potatoes (rarely - 0R=0.45 95%CI 0.22-0.94). No consistent associations were found for either age group for fish, butter, oil, sugar, vegetables, and meat.
Results of General Linear Model analysis are shown in Table 4. In-
tercepts corresponding to reference categories of all included variables were very similar for both genders and constituted 26.62 (95%CI 25.74-27.51) for men and 26.64 (25.79-27.49) for women. BMI was significantly lower by about 2 units in younger people than in older ones, for both genders. With age controlled for, association with education and occupation was found non-significant.
Physical exercises reported to be performed 'sometimes' compared to 'never' were found to correspond 1.5 units lower BMI in men, and 2,2 lower BMI in women but only of younger age. In older group no significant difference was found.
Among the foods whose frequency was collected in the survey, only more frequent eating of meat and dairy products was associated with slightly higher BMI, but only in men.
BMI was slightly higher in both men and women who reported drinking strong alcohol on a monthly-to-weekly basis. Contrary to this, drinking beer monthly-to-weekly compared to 'never' was associated with BMI lower by about 2 units, but this was found in only younger women.
Smoking on a daily basis in men was associated with lower BMI by 2 units; however, this was less typical for younger men.
discussion
Explorative analysis of data from the survey "Health and Well-being in Transition Societies", 2000 by age and gender groups showed certain associations pertinent to all and others quite specific for particular demographic groups.
Like in many other countries, women in Ukraine were found to be more likely overweight, especially in older age groups. Obesity is more prevalent in women worldwide due to biological (higher ability to deposit fat tissue than men) and social factors (James, Leach, Kalamara, & Shayeghi, 2001). However, this gender disparity could be partly accounted for by sedentary lifestyles (Blanck et al., 2007): little exercise, but much TV-watching (Rosmond & Bjorntorp, 1999).
Another finding in our study common for men and women was that proportion of people with excessive weight increased with age. This is in line with the results of other researchers (McTigue, Hess, & Ziouras, 2006; Ogden et al., 2006).
Much attention is traditionally given to socioeconomic factors including education, income, unemployment, problems at work when employed, dissatisfaction with work as determinants of excess body weight (Rosmond & Bjorn-torp, 1999; Wardle, Waller, & Jarvis, 2002). However, association with socioeconomic status is inconsistent. In traditional societies, representatives of higher SES are more likely to be overweight, while in modern societies the opposite is true for adult women but mixed patterns are typical for other age/sex groups (Sobal, 1991) or, as was seen later, an "increasing proportion of positive associations and a decreasing proportion of negative associations occurs as one moves from countries with high levels of socioeconomic development to countries with medium and low levels of development" (McLaren, 2007), that is in less developed countries more studies show that
Table 4. Results of General Linear Model analysis in different age groups, stratified by gender, Intercept equal to BMI of reference groups with values set to 0, ukraine, 2000
Interaction Variables Values terms
Intercept
Age before and after 45 15-44 45+
All age groups
Frequency of alcoholic drinks use Daily or nearly daily Once a week 2 - 3 times a month About once a month or less frequently Never Tobacco use Yes
Sometimes No
Frequency of meat consumption Rare
Frequently Frequency of milk products consumption Rare
Frequently
Age 15-44
How often are you physically active? Sometimes Never Tobacco use Yes
Sometimes No
Frequency of beer consumption Daily or nearly daily Once a week 2 - 3 times a month About once a month or less frequently Never
Men Women
N B 95% CI N B 95% CI
26,62 25,74 27,51 26,64 25,79 27,49
369 -2,20 -3,27 -1,13 471 -2,08 -2,91 -1,26
361 0,00 ■ 549 0,00
26 1,25 -0,25 2,75 6 -1,08 -5,16 2,99
109 1,08 0,14 2,01 18 1,27 -1,08 3,61
123 0,87 -0,04 1,77 70 1,98 0,74 3,22
318 0,77 0,04 1,51 478 1,35 0,69 2,00
154 0,00 448 0,00 ,
373 -1,97 -2,70 -1,23 70 -1,50 -4,19 1,20
23 0,51 -1,65 2,67 19 -2,46 -9,51 4,59
334 0,00 931 0,00 ,
488 -0,82 -1,37 -0,27 758 0,18 -0,52 0,88
242 0,00 262 0,00
510 -0,61 -1,16 -0,06 703 -0,46 -1,10 0,18
220 0,00 317 0,00
181 -1,53 -2,23 -0,82 173 -2,19 -3,11 -1,26
189 0,00 299 0,00
214 1,78 0,73 2,82 57 1,87 -1,13 4,87
15 -1,31 -4,22 1,61 18 1,14 -6,27 8,55
143 0,00 , 398 0,00
15 0,11 -1,82 2,04 4 -3,22 -8,01 1,57
69 0,04 -1,11 1,20 30 -2,36 -4,23 -0,48
75 0,94 -0,19 2,06 43 -2,26 -3,83 -0,69
126 0,04 -0,95 1,02 140 -1,85 -2,83 -0,87
87 0,00 , , 255 0,00 , ,
higher SES is associated with heavier bodies and in more developed countries more populations show inverse relationship between body mass and SES. In our study, no association between educational attainment and overweight was found in any age or gender groups. This absence of correlates found in either traditional or modern societies might mean that Ukraine in 2000 was genuinely 'in transition'
between the two. Later uptake of obesity epidemic was also seen in other Eastern European countries (Webber et al., 2012).
Smoking was found to be significantly related to lower body mass, and these associations were more prominent in those demographic groups with higher smoking prevalence - older men and younger women. This finding is in line with many earlier observations showing
that smokers are more likely to have lower body mass than non-smokers (Fehily, Phillips, & Yarnell, 1984; Jacobs & Gotten-borg, 1981; Shimokata, Muller, & Andres, 1989), which is associated with the duration of smoking more than with its intensity (Albanes, Jones, Micozzi, & Mattson, 1987); however, the new thing in our study is that the association with smoking was found predominantly
in groups with high smoking prevalence where variance of BMI created by other factors becomes less visible.
Within several demographic groups, use of milk and dairy products was found to be associated with higher body mass index. Studies considering relationship between dairy products and body mass from the point of view of calcium intake either have shown body mass and adiposity decrease (Van Loan, 2009) in adults (Mirmi-ran, Esmaillzadeh, & Azizi, 2004; Shahar et al., 2010) and children (Kelishadi et al., 2009) or no change in body mass (Gunther et al., 2005) while studies revealing inverse relationship being mostly cross-sectional or retrospective. A systematic review of randomized trials of the impact of dairy products on body weight has shown that most found no effect while some have shown weight gain, which is likely to be explained by increased energy intake (Barr, 2003). Some studies indicate that BMI change is dependent on the type of dairy products consumed (Rosell, Hakansson, & Wolk, 2006). Obviously, if respondents in our study reported consumption of whole milk at meals where others consumed water or tea, this meant higher energy intake, which is translated in weight gain. Consumption of fatty dairy products as a factor of overweight and consequent chronic diseases was addressed in neighbouring Poland (Zatonski, McMichael, & Powles, 1998). On the other hand, consumption of whole milk might be a marker of specific lifestyle, or living in a countryside, owning a cow or living in proximity to places where someone owns a cow. This might comprise a wide range of other factors affecting BMI.
The revealed associations with physical activity illustrate the pattern we hypothesized to find. In younger age groups of those more physically active have lower BMI, and in older age groups those overweight are more likely to start exercising. Besides this interpretation, many studies show that physical activity can possibly lead to weight gain (Fogelholm & Kukkonen-Har-jula, 2000), as BMI is actually not a specific measure of adiposity, and physically active people with increased muscular mass may have increased BMI as well (Schroeder, 2007).
The association between vodka consumption and BMI revealed in women might be explained either by additional alcohol-related energy intake or by certain eating patterns linked to vodka consumption. Association between alcohol use and increased BMI (Cournot et al., 2004) in women was also found in France. In contrast, BMI was found to be negatively related to wine consumption, while beer and strong liquor did not show any significant relationships to the BMI (Rosmond & Bjorntorp, 1999). In the USA (Williamson et al., 1987) alcohol use was shown to have no association with the BMI in men while the association was inverse in women.
The revealed association with eating potatoes was somewhat contradictory: for young men, the association was inverse while in most groups of women it was positive: the more potatoes consumed the higher was BMI. However, other studies find similar patterns: adults whose diets include more white potatoes have higher BMIs while overweight boys were found to consume less total vegetables (2.5 servings) as well as less white potatoes (1.2 servings) than other
boys (Lin & Morrison, 2002). Moreover, the mean annual change in BMI was found to be much higher for subjects in the so-called 'meat-and-potatoes cluster' than in 'healthy food cluster' (Newby et al., 2003), and this pattern is expected to be quite widespread in Ukraine. Nutrition of Ukrainian workers was found to be rich in calories due to bread, potato, pure carbohydrates and some products of animal origin, mostly fats (Priputina, 1975).
Some of the revealed associations point to certain lifestyle patterns, especially among women. For example, slim women are more likely to be self-employed and to drink beer, which is hardly to be causal as beer provides extra calories. Overweight women were more likely to have lower than university education and to admit drinking vodka.
Absence of associations found for most foods can be due to either non-sensitive way of measurement of their consumption, or the fact that they do not clearly belong to those lifestyle patterns associated with BMI revealed for women. Some of the typical associations like the one that risk of being overweight or obese was positively associated with soft drink and sugar consumption (Vartanian, Schwartz, & Brownell, 2007), was not found in our analysis.
Furthermore, a factor known to be a cause for obesity may reveal no association with the body mass index due to absence of sufficient variance in the studied group. It is necessary to take into account that in the Soviet times and partly thereafter nutrition of the population was rather uniform with centralized distribution of meat, sausage, and butter. Other quite
uniform pattern for Ukrainian rural population would contain much potato and some vegetables. All the abovementioned peculiarities of the data would lead statistical tests to show no association between the eating patterns and BMI.
Results of the present study should be treated with certain caution due to cross-sectional nature of the data that made impossible to establish the sequence of events and causality of certain relationships found between variables. Self-reported height and weight of the participants might be a source of bias especially in older age groups where overweight status can be underestimated (Kuczmarski, Kuczmarski, & Najjar, 2001).
Implications for further research include but are not limited to longitudinal observation of how trajectories of eating patterns specific for Ukraine and risk factors including overweight are related to each other over time. Another research question which deserves further consideration is whether the ideas of 'healthy eating' is a factor that plays a role in food choices by Ukrainian dwellers.
conclusions
1. In 2000, 44% of representative sample of Ukrainian population were overweight or obese: 39% of men and 49% of women. Younger people were less likely overweight (16% of men and 12% of women below 40 years) than those in older age groups (47% of men over 40, 53% of women over 40 and 65% of women over 60 years old). No disparity by education or occupation group was found.
2. Smokers were less likely overweight; however, this association was seen only in groups with rela-
tively high smoking prevalence -in men over 40 and women below 40 years old. Beer users, especially women, were less likely to be overweight, strong alcohol users had on average larger body mass index; however, this reflects lifestyle and eating patterns rather than causal effects.
3. Men and women who reported practicing physical exercises at least sometimes had BMI 1.5-2.0 units lower than those never doing exercises; however, this was seen only among those below 45 years old. Above that age, overweight women more likely reported doing exercises.
4. More frequent consumption of potatoes, meat and milk in some demographic groups was associated with high BMI, and eating fruit—with lower. The expectable association with using sugar or soft drinks was not found in Ukraine in 2000.
acknowledgement
Authors are grateful to the Kyiv International Institute of Sociology for the data collected in 2000.
about the authors
Natalia S. Chagarna is a graduate of the School of Public Health, National University of Kyiv-Mohyla Academy, Kiev, Ukraine.
ORCID 0000-0002-6891-7961
Email: [email protected]
Tatiana I. Andreeva was affiliated with the same School of Public Health, National University of Kyiv-Mohyla Academy, Kiev, Ukraine.
ORCID 0000-0001-5663-6225 Email: [email protected]
This paper was received: January,
2012; accepted: June 16, 2014;
published: June 27, 2014.
references
Albanes, D., Jones, D. Y., Micozzi, M. S., & Mattson, M. E. (1987). Associations between smoking and body weight in the US population: analysis of NHANES II. Am J Public Health, 77(4), 439-444. doi: 10.2105/ajph.77.4.439
Barr, S. I. (2003). Increased Dairy Product or Calcium Intake: Is Body Weight or Composition Affected in Humans? The Journal of Nutrition, 133(1), 245S-248S.
Biloukha, O. O., & Utermohlen, V. (2000). Correlates of food consumption and perceptions of foods in an educated urban population in Ukraine. Food Quality and Preference, 11(6), 475-485. doi: 10.1016/s0950-3293(00)00020-3
Bilukha, O. O., & Utermohlen, V. (2002). Internalization of Western standards of appearance, body dissatisfaction and dieting in urban educated Ukrainian females. European Eating Disorders Review, 10(2), 120-137. doi: 10.1002/erv.431
Blanck, H. M., McCullough, M. L., Patel, A. V., Gillespie, C., Calle, E. E., Cokkinides, V. E., . . . Serdula, M. K. (2007). Sedentary Behavior, Recreational Physical Activity, and 7-Year Weight Gain among Postmenopausal U.S. Women[ast]. Obesity, 15(6), 1578-1588.
Cournot, M., Ruidavets, J.-B., Marquie, J.-C., Esquirol, Y., Baracat, B., & Ferrieeres, J. (2004). Environmental factors associated with body mass index in a population of Southern France. European Journal of Cardiovascular Prevention & Rehabilitation, 11(4), 291-297. doi:
10.1097/01.hjr.0000129738.22970.62
Crawford, D., & Ball, K. (2002). Behavioural determinants of the obesity epidemic. Asia Pac J Clin Nutr, 11 Suppl 8, S718-721. doi: 374 [pii]
Epstein, L. H., Gordy, C. C., Raynor, H. A., Beddome, M., Kilanowski, C. K., & Paluch, R. (2001). Increasing
Fruit and Vegetable Intake and Decreasing Fat and Sugar Intake in Families at Risk for Childhood Obesity. Obesity, 9(3), 171-178.
Fehily, A., Phillips, K., & Yarnell, J. (1984). Diet, smoking, social class, and body mass index in the Caerphilly Heart Disease Study. The American Journal of Clinical Nutrition, 40(4), 827-833.
Fogelholm, M., & Kukkonen-Harjula, K. (2000). Does physical activity prevent weight gain - a systematic review. Obesity Reviews, 1(2), 95-111. doi: 10.1046/j. 1467-789x.2000.00016.x
Friedman, L. S., Lukyanova, E. M., Serdiuk, A., Shkiryak-Nizhnyk, Z. A., Chislovska, N. V., Zvinchuk, A. V., . . . Hryhorczuk, D. (2009). Social-environmental factors associated with elevated body mass index in a Ukrainian cohort of children. Int JPediatr Obes, 4(2), 81-90. doi: 904886991 [pii] 10.1080/17477160802449264 [doi]
Guh, D., Zhang, W., Bansback, N., Amarsi, Z., Birmingham, C. L., & Anis, A. (2009). The incidence of co-morbidities related to obesity and overweight: A systematic review and meta-analysis. BMC Public Health, 9(1), 88.
Gunther, C. W., Legowski, P. A., Lyle, R. M., McCabe, G. P., Eagan, M. S., Peacock, M., & Teegarden, D. (2005). Dairy products do not lead to alterations in body weight or fat mass in young women in a 1-y intervention. The American Journal of Clinical Nutrition, 81(4), 751-756.
Haslam, D., & James, P. (2005). Prevalence of obesity worldwide. Lancet, 366, 1197-1209.
Huang, T. T., Drewnosksi, A., Kumanyika, S., & Glass, T. A. (2009). A systems-oriented multilevel framework for addressing obesity in the 21st century. Prev Chronic Dis, 6(3), A82. doi: A82 [pii]
Jacobs, D. R., Jr, & Gottenborg, S. (1981). Smoking and weight: the Minnesota Lipid Research Clinic. Am J Public Health, 71(4), 391-396. doi: 10.2105/ajph.71.4.391
James, P. T., Leach, R., Kalamara, E., & Shayeghi, M. (2001). The Worldwide Obesity Epidemic. Obesity, 9(11S), 228S-233S.
Kalabiska, I., Uvacsek, M., Petrekanits, M., Cseprekal, O., Ihasz, F., & Frenkl, R. (2010). Comparison of running performances and prevalence of overweight and obesity in Hungarian and Ukrainian adolescents. Acta Physiol Hung, 97(4), 393-400. doi: 962G21P23331L7L1 [pii] 10.1556/APhysiol.97.2010.4.5 [doi]
Kelishadi, R., Zemel, M. B., Hashemipour, M., Hosseini, M., Mohammadifard, N., & Poursafa, P. (2009). Can a Dairy-Rich Diet Be Effective in Long-Term Weight Control of Young Children? Journal of the American College of Nutrition, 28(5), 601-610.
Khalangot, M., Tronko, M., Kravchenko, V., Kulchinska, J., & Hu, G. (2009). Body mass index and the risk of total and cardiovascular mortality among patients with type 2 diabetes: a large prospective study in Ukraine. Heart, 95(6), 454-460. doi: hrt.2008.150524 [pii] 10.1136/hrt.2008.150524 [doi]
Kuczmarski, M. F., Kuczmarski, R. J., & Najjar, M. (2001). Effects of age on validity of self-reported height, weight, and body mass index: findings from the Third National Health and Nutrition Examination Survey, 1988-1994. J Am Diet Assoc, 101(1), 28-34; quiz 35-26. doi: 10.1016/s0002-8223(01)00008-6
Lin, B. H., & Morrison, R. M. (2002). Higher fruit consumption linked with lower body mass index. Food Review, 25(3), 28-32.
Ma, J., Flanders, W. D., Ward, E. M., & Jemal, A. (2011). Body Mass Index in Young Adulthood and Premature Death: Analyses of the US National Health Interview Survey Linked Mortality Files. Am J Epidemiol, 174(8), 934-944. doi: 10.1093/aje/kwr169
McLaren, L. (2007). Socioeconomic status and obesity. Epidemiol Rev, 29, 29-48. doi: 10.1093/epirev/mxm001
McTigue, K., Hess, R., & Ziouras, J. (2006). Obesity in older adults: a systematic review of the evidence for diagnosis and treatment. Obesity, 14(9), 1485-1497.
Mirmiran, P., Esmaillzadeh, A., & Azizi, F. (2004). Dairy consumption
and body mass index: an inverse relationship. Int J Obes Relat Metab Disord, 29(1), 115-121.
Newby, P., Muller, D., Hallfrisch, J., Qiao, N., Andres, R., & Tucker, K. L. (2003). Dietary patterns and changes in body mass index and waist circumference in adults. The American Journal of Clinical Nutrition, 77(6), 1417-1425.
Ogden, C. L., Carroll, M. D., Curtin, L. R., McDowell, M. A., Tabak, C. J., & Flegal, K. M. (2006). Prevalence of Overweight and Obesity in the United States, 1999-2004. JAMA: the journal of the American Medical Association, 295(13), 1549-1555. doi: 10.1001/jama.295.13.1549
Page, R. M., Ihasz, F., Uvacsek, M., Kalabiska, I., Johnson, E., Hantiu, I., . . . Simonek, J. (2006). Self-perception of body weight, weight management practices and goals, and other weight-related factors in central and eastern European adolescents. Int Q Community Health Educ, 27(1), 39-58. doi: R803723K48318749 [pii] 10.2190/IQ.27.1.d [doi]
Paniotto, V., Martsinkiv, A., Kharchenko, N., Zakhozha, V., & Buslayeva, E. (2004). A study of health, the need for health and social services, and health-related living conditions in Ukraine. In A. N. Hannu Valtonen (Ed.), Health and Well-Being in Transition Societies (pp. 84-120). Finland: Gummerus Printing Saarijarvi.
Priputina, L. S. (1975). Nutritional characteristics and the prevalence of obesity among individual groups of the population in the Ukrainian SSR. Vopr Pitan(1), 3-6.
Rosell, M., Hakansson, N. N., & Wolk, A. (2006). Association between dairy food consumption and weight change over 9 y in 19 352 perimenopausal women. The American Journal of Clinical Nutrition, 84(6), 1481-1488.
Rosmond, R., & Bjorntorp, P. (1999). Psychosocial and socio-economic factors in women and their relationship to obesity and regional body fat distribution. International journal of obesity and related metabolic disorders : journal of the International Association for the Study of Obesity, 23(2), 138-145.
Schroeder, S. A. (2007). We Can Do Better — Improving the Health of the American People. New England Journal of Medicine, 357(12), 1221-1228. doi: doi:10.1056/NEJMsa073350
Shahar, D. R., Schwarzfuchs, D., Fraser, D., Vardi, H., Thiery, J., Fiedler, G. M., . . . Group, f. t. D. (2010). Dairy calcium intake, serum vitamin D, and successful weight loss. The American Journal of Clinical Nutrition, 92(5), 1017-1022. doi: 10.3945/ajcn.2010.29355
Shimokata, H., Muller, D. C., & Andres, R. (1989). Studies in the Distribution of Body Fat. JAMA: the journal of the American Medical Association, 261(8), 1169-1173. doi: 10.1001/jama.1989.03420080089037
Sobal, J. (1991). Obesity and socioeconomic status: a framework for examining relationships between physical and social variables. Med
Anthropol, 13(3), 231-247. doi:
10.1080/01459740.1991.9966050
[doi]
Van Loan, M. (2009). The Role of Dairy Foods and Dietary Calcium in Weight Management. Journal of the American College ofNutrition, 2#(Supplement 1), 120S-129S.
Vartanian, L. R., Schwartz, M. B., & Brownell, K. D. (2007). Effects of Soft Drink Consumption on Nutrition and Health: A Systematic Review and Meta-Analysis. Am J Public Health, 97(4), 667-675. doi: 10.2105/ajph.2005.083782
Wang, Y., & Beydoun, M. A. (2007). The Obesity Epidemic in the United States—Gender, Age, Socioeconomic, Racial/Ethnic, and Geographic Characteristics: A Systematic Review and MetaRegression Analysis. Epidemiologic Reviews, 29(1), 6-28. doi: 10.1093/epirev/mxm007
Wardle, J., Waller, J., & Jarvis, M. J. (2002). Sex Differences in the
Association of Socioeconomic Status With Obesity. Am J Public Health, 92(8), 1299-1304. doi: 10.2105/ajph.92.8.1299
Webber, L., Kilpi, F., Marsh, T., Rtveladze, K., McPherson, K., & Brown, M. (2012). Modelling obesity trends and related diseases in Eastern Europe. Obes Rev, 13(8), 744-751. doi: 10.1111/j.1467-789X.2012.00999.x
Williamson, D. F., Forman, M. R., Binkin, N. J., Gentry, E. M., Remington, P. L., & Trowbridge, F. L. (1987). Alcohol and body weight in United States adults. Am J Public Health, 77(10), 1324-1330. doi: 10.2105/ajph.77.10.1324
Zatonski, W., McMichael, A., & Powles, J. (1998). Ecological study of reasons for sharp decline in mortality from ischaemic heart disease in Poland since 1991 British Medical Journal, 316(7137), 1047-1051.
