Abstract

Background Obesity has become a worldwide epidemic. Recently WHO acknowledged an urgent need to examine child obesity across countries using a standardized international standard. Studies in adults find obesity and socioeconomic factors (SES) factors are correlated, but results are inconsistent for children. Using an international standard, we examined the prevalence of obesity and compared the associations between SES factors and obesity across countries.

Methods Data for children aged 6–18 from nationwide surveys in the US (NHANES III, 1988–1994), China (1993), and Russia (1992) were used. We used the recently updated US NCHS body mass index (BMI = wt/ht2) reference to define obesity (BMI ≥95th percentile) and overweight (85th≤BMI<95th percentile). The WHO recommends an early version of the NCHS reference for international use. We conducted logistic analyses to examine the relationship between SES and obesity.

Results The prevalence of obesity and overweight was 11.1% and 14.3%, respectively, in the US, 6.0% and 10.0% in Russia, and 3.6% and 3.4% in China. The relationship between obesity and SES varied across countries. Higher SES subjects were more likely to be obese in China and Russia, but in the US low-SES groups were at a higher risk. Obesity was more prevalent in urban areas in China but in rural areas in Russia.

Conclusions Child obesity is becoming a public health problem worldwide, but the prevalence of obesity varies remarkably across countries with different socioeconomic development levels. Different SES groups are at different risks, and the relationship between obesity and SES varies across countries.

Obesity has become a global epidemic, and it is still increasing in both industrialized and developing countries.1 For example, the prevalence of child obesity and overweight has doubled in North America during the past two decades. At present, about one-quarter of children in the US are obese or overweight.2 In Thailand, a transitional society, the prevalence of obesity in schoolchildren has increased from 12% in 1991 to 16% in 1993.3 Due to the difficulty of curing obesity in adults and the many long-term adverse effects of childhood obesity, the prevention of child obesity has been recognized as a public health priority.4 Increasing evidence shows that childhood obesity has a profound influence on morbidity and mortality in adult life.4–6 However, few studies have examined the worldwide situation regarding childhood obesity, particularly due to the fact that no standard or reference is agreed upon internationally. Different definitions have been used in studies to define childhood obesity.1,4,7,8 Recently a World Health Organization (WHO) consultation on obesity concluded, ‘The current lack of consistency and agreement between different studies over the classification of obesity in children and adolescents makes it difficult to give an overview of the global prevalence of obesity', and an examination of obesity in children and adolescents across the world based on a standardized obesity classification system is urgently needed.1

The literature has another major gap because a large number of studies on adults have examined the relationship between socioeconomic status (SES) factors and obesity, but less research has used data representing large populations to examine the effects of these relationships on children.9–12 Sobal and Stunkard have provided an excellent review of this topic.11 After examining over 140 published studies, they concluded that these studies reveal a strong inverse relationship between SES and obesity among women in developed societies, but the relationship is inconsistent for men and children. In contrast, in developing countries a strong relationship exists between SES and obesity among men, women, and children. It is of concern, however, that since different obesity definitions and SES indicators were used, the findings in different studies may not be comparable.

Moreover, it is argued that how SES and obesity are related is not clear. Among adults it is likely that causality operates in either direction.13–15 We expected that it may be easier to test the direction of such a causality with studies among children because usually the SES of children is determined by their parents' characteristics instead of being influenced by their own obesity status and the related social consequences. To our knowledge, limited efforts have been made to systemically examine the relationships between childhood obesity and SES across countries using large-scale survey data.

Our main objective is to examine the current cross-continental situation of child and adolescent obesity and compare the relationship between SES factors and obesity across countries using an international obesity standard. We include industrialized and developing countries from different continents and with different socioeconomic development levels.16,17 Nationwide survey data collected in the early 1990s from the US, China, and Russia are used. The national total population was 1244 million in China, 268 million in the US, and 147 million in Russia in the early 1990s; and their total populations account for over a quarter of the global population.18

Materials and Methods

Subjects

Children and adolescents aged 6–18 years who had complete anthropometric and demographic data from national surveys in each country were included. Pregnant girls were excluded. The final sample size was 6110 for the US; 3028 for China; and 6883 for Russia.

Data sources

The Third National Health and Nutrition Examination Survey (NHANES III, 1988–1994)

The third NHANES was a cross-sectional representative sample of the US civilian, non-institutionalized population aged ≥2 months. NHANES III contains data for a sample of 33 994 people, and it oversampled blacks, Mexican Americans, children under 5 years, and the elderly (≥60 years). Detailed descriptions of the sample design and operation of the survey have been published elsewhere.19 Standardized protocols were used for all interviews and examinations. Data on weight and height were collected for each individual in the full mobile examination centre through direct physical examinations. Based on self-reported race and ethnicity, subjects were classified into non-Hispanic white, non-Hispanic black, Mexican American, and other four ethnic groups.

The China Health and Nutrition Surveys (CHNS, 1993)

The CHNS covered eight provinces that vary substantially in geography, economic development, public resources, and health indicators. Anthropometric measurements were carried out by trained health workers, who followed standard protocol similar to the US NHANES protocol developed by the US National Center for Health Statistics (NCHS). Weight was measured in light indoor clothing to the nearest tenth of a kilogram with a beam balance scale. Height was measured without shoes to the nearest tenth of a centimetre using a portable stadiometer. Detailed descriptions of the CHNS have been published elsewhere.20 The 1993 data were used. The CHNS time frame matches that of US NHANES III.

The Russian Longitudinal Monitoring Survey (RLMS, 1992)

The RLMS is the first nationally representative household survey in the Russian Federation. All members of more than 6400 households from all regions of Russia were surveyed eight times from 1992 to 1998. Weight and height were measured to follow a protocol similar to the ones used in the NHANES III and CHNS surveys. Detailed descriptions of the RLMS have been published elsewhere.21 The 1992 data were used for a better comparison (i.e. similar survey year) with data from the US and China.

Measures

Definition of obesity and overweight

Adiposity was measured by using the body mass index (BMI = weight [kg]/height [m]2). Following WHO's recommendation, we used a series of sex-age-specific BMI cut-offs to define obesity and overweight.22 It is important to choose a good obesity standard for examining the global obesity epidemic and making international comparisons.1,4,7,8 A WHO Expert Committee recommends using a set of sex- and age-specific BMI percentiles, developed by Must and others based on the US NHANES I data collected in 1971–1975, to define adolescent obesity, overweight, and underweight.22–24 Specifically the WHO recommends using the BMI 85th percentile to define overweight and the 5th percentile for underweight. This reference is called the ‘WHO/ NCHS reference',22 and it has been widely used in the US and other countries.4,8,25–29 Although the WHO Expert Committee recommends using both the BMI 85th percentile and the triceps skinfold thickness 90th percentile to define adolescent obesity,22 this has been used infrequently due to the difficulty of measuring triceps skinfold thickness in large population-based studies (e.g. our RLMS data). Instead, many have recommended the use of the BMI 95th percentile to define child and adolescent obesity.23,24,29,30

Most recently the US NCHS updated the old WHO/NCHS reference by using new data and better statistical techniques,30 although the new BMI cut-offs are similar to the previous ones. It is expected that this new reference will be accepted to replace the old WHO/NCHS reference for international research. Therefore we chose to use the new NCHS BMI cut-offs. The sex-age-specific BMI 85th, and 95th percentiles were used to define overweight, and obesity respectively. Also, because the prevalence of obesity is very low in China, in our comparison analysis we focused on overweight.

An alternative that we have considered is the recently published ‘IOTF standard' proposed by the Childhood Obesity Working Group of the International Obesity Task Force (IOTF).7,31 Our previous analysis showed that in general the WHO/NCHS and IOTF references produced similar estimates of overall overweight prevalence, although the differences are noticeable for certain ages.32 In addition, recently some concerns have been raised regarding the use of the IOTF standard, except for its strengths.32,33 Our decision to use the WHO/NCHS reference allows easier comparisons between our findings and those of others who have used this reference.

Sociodemographic characteristics

Subjects were separated into two age groups: children (6–9 years) and adolescents (10–18 years). Self-reported urban-rural residences were used. As a result the meanings of urban and rural residences may vary across countries. Per capita family income tertiles were used to indicate low, middle, and high SES. Family income data were collected at the same time children's BMI measurements were collected.

Statistical analysis

First, we examined the prevalence of overweight and underweight in each country by age group, sex, urban/rural residence, and SES. For the US, to achieve national representative prevalence estimates, sampling weights were used to adjust for sample design effects. Next, using logistic regression models, we examined the associations between obesity and SES for each country. Since the prevalence of obesity was low in China and Russia, we combined obesity and overweight (i.e. BMI ≥85th percentile). Furthermore, we examined the associations between BMI and SES. For the US, in all regression analyses, sample design effects were adjusted, and the effect of ethnicity was also examined. All analyses were performed by using SAS Version 6.12 (SAS, Cary, NC, USA) and Stata Version 6.0 (Stata Co., College Station, TX, USA).

Results

Children and adolescents' main characteristics and BMI distribution in each country

Each sample's main sociodemographic characteristics are summarized in Table 1. On average American and Russian children and adolescents were taller and heavier than their Chinese counterparts. We also examined the distribution of BMI among children and adolescents in each country (Figures 1 and 2). A clear shift of increased BMI existed across the three countries. The trends were more obvious among adolescents than among children.

The prevalence of overweight and obesity across countries

As shown by Figure 3, the overall prevalence of obesity and overweight was high in the US (combined prevalence was 25.4%), low in China (combined prevalence, 7.0%), and moderate in Russia (combined prevalence, 16.0%). These clearly suggest that national socioeconomic development levels influence the epidemic of obesity. The prevalence of obesity and overweight and the combined prevalence of obesity and overweight by age, sex, and SES groups are presented in Table 2. In the US, the prevalence of obesity and overweight among American adolescents was lowest in the high-income group. By contrast, in China the high-income groups generally were at a higher risk of obesity. The prevalence of obesity was higher in rural areas in Russia but higher in urban areas in China. Interestingly, in both Russia and China, but not in the US, the prevalence of obesity and overweight was higher among children than among adolescents. The differences were especially remarkable for obesity.

Comparison of relationships between obesity and SES across countries

Socioeconomic status was related to children's and adolescents' risks of being obese or overweight (simply termed ‘obese'), although the relationships were different across countries (Table 3). In Russia urban children and adolescents were at a lower risk of obesity, but in China urban children and adolescents were at a higher risk. Compared to the medium-income group, the low-income group was at a higher risk in both the US and Russia, and in Russia the high-income group was also at an increased risk. In China, however, high-income children and adolescents were more likely to be obese. Moreover our analysis stratified by gender indicates that in both China and Russia (but not the US), boys and girls are at different risks even if they have the same SES (Table 4). That is, gender is an effect modifier.34

Sobal and Stunkard suggested that research was needed to examine whether the relationships change with age in children and adolescents.11 We tested the relationship between SES and obesity by age group (6–9 years and 10–18 years) for each country, and then by each year of age for American children and adolescents. We found that similar significant associations between obesity and SES existed in both age groups in China and Russia. By contrast, in the US only among adolescents (age ≥10 years) did we find a reverse relationship between SES (i.e. income) and obesity. Low-income adolescents were at a higher risk for obesity (OR = 1.4, 95% CI : 1.1–1.9), and high-income adolescents were at a lower risk (OR = 0.7, 95% CI : 0.5–0.9). The SES was not significantly related to obesity among children under 10 years. The association became significant only among adolescents at age 12, and it became insignificant at age 17.

Moreover, for the US consistent with many previous studies, we found that ethnicity was a significant risk factor. Compared to whites, blacks and Mexican American children and adolescents were at a higher risk for obesity and overweight; OR were 1.2 (95% CI : 1.01–1.5) and 1.4 (95% CI : 1.1–1.9), respectively. Age and gender were adjusted. However, when income and urban-rural residence were adjusted, ethnicity became insignificant. These may suggest that the SES differences across ethnic groups are likely the main explanation for the difference in obesity prevalence across ethnic groups. We could not examine these ethnic/ subculture variations in China and Russia because minorities only count for a small proportion of their total populations. In both the Russia and China surveys, only a very small proportion of the subjects are from minority ethnic groups. The small sample sizes did not allow us to conduct meaningful comparisons.

Finally, we examined BMI as a continuous variable, stratified by sex and age groups. In general, results (Table 5) were consistent with the logistic regression analysis. High-income American girls had lower BMI. In Russia urban girls had lower BMI, but low- and high-income groups had higher BMI. In contrast, in China urban boys had higher BMI. Also SES effects seemed to vary by age groups. In the US, for example, although low-income was not a significant risk factor for elevated BMI among children under 10 years, it was a strong one among adolescents (age ≥10 years). For the US we also examined the influence of ethnicity. Interestingly, African-American girls and Mexican-American girls were more likely to have a higher BMI than white girls even when controlled for family income and urban/rural residence (P < 0.05). The regression coefficients and SES were 0.87 (0.27) and 0.84 (0.39), respectively. For males, however, ethnicity was not a significant risk factor (P > 0.05). In addition, since the distribution of BMI in each sample was slightly skewed, we repeated the regression analyses using natural logarithms transformed BMI. The results we found were similar.

Discussion

The lack of consistent classifications in various studies makes it difficult to assess the global situation of child and adolescent obesity.1–8 To our knowledge, our study is the first attempt to examine child and adolescent obesity across countries using an international standard and based on data from large nationwide surveys. Overall we found that the prevalence of obesity differs remarkably across countries with different socioeconomic development levels. The combined prevalence of obesity and overweight was high in the US (25%), moderate in Russia (16%), and low in China (7%). Even though the use of different standards makes comparisons of findings difficult, previous studies indicate that in many other developed countries child and adolescent obesity have reached levels comparable to those in the US.1,35–37 In addition evidence suggests that the prevalence of overweight has increased to relatively high levels in many developing countries.1 For example, in Brazil the prevalence of overweight among schoolchildren has tripled and increased from 4% in the 1970s to 14% in late 1990s. In 1997, 17% of children and 13% of adolescents were obese or overweight in 1997.38 In Egypt 14% of adolescents were overweight or obese in 1997.25 Worthy of mention, there are still controversies over the use of a series of universal BMI cut-offs to define obesity or overweight in different populations of either adults or children.22,32,33,39 For example, it was recently suggested that different BMI cut-offs should be used for Asian and Caucasian populations.33 Nevertheless, our results clearly show that childhood obesity is becoming a worldwide epidemic. The remarkable variation in the prevalence across populations suggests that social, economic, and environmental factors are important influences on the epidemic, although it maybe also true that genetic differences across populations also play a role.5

Our analysis shows that child and adolescent obesity is related to SES, although the relationships differ among these three populations. We used family income as a primary indicator of SES, while rural-urban residence might serve as an additional indicator. In the US low SES groups had a higher risk of obesity. By contrast, in China high SES groups were at an increased risk. In Russia, a transitional society that has experienced economic difficulties since the early 1990s,21,40 both low-income and high-income groups were at an increased risk of obesity compared to the medium-income group. One possible explanation for the different SES-obesity relationship in developed countries such as the US and developing countries such as China is that the influence of SES on people's lifestyles such as diet and physical activity may differ. Take food consumption patterns as an example. In China richer people have better access to meat and other energy-dense foods (which are much more expensive than other foods such as vegetables) than the poor.26 While in the US, higher-SES groups usually consume more vegetables and fruits, which are less energy-dense, than low-SES groups.41

Unlike China, where urban children were more likely to be obese, in Russia urban groups were less likely to be obese than were rural groups. In the US no consistent rural-urban difference emerged. Similar to the results for China, the recent Egyptian national survey found that the prevalence of overweight and obesity was 22.6% among urban adolescents versus 10.4% among rural adolescents.25 These patterns are particularly attributable to the differences in people's access to food and health services, physical activity patterns, and social norms in rural and urban areas in these countries. Compared to their rural counterparts, urban Chinese usually have higher family income, better access to food (especially meat and poultry), public services such as health care and transportation. They are also more likely to have sedentary lifestyles.26,42 In contrast, US rural and urban children and adolescents have similar access to food choices thanks to the well-established food production and distribution system. They are also likely to have similar lifestyles.43 In Russia, urban groups were less likely to be obese than rural groups. This is probably due to that in the past decade, living standards of urban groups have been seriously affected by the socioeconomic difficulties occurring since the collapse of the former Soviet Union.16,17,21,40 Furthermore, our analysis stratified by sex and age further indicate that social, economic, and environmental factors may operate through complex pathways to influence childhood obesity.

In general our findings regarding the relationships between obesity and SES are consistent with findings from many previous studies.10,11,25,44 For example, among the 32 studies conducted on girls from developed societies reviewed by Sobal and Stunkard,11 40% found a inverse relationship between SES and obesity, although 25% found a positive relationship and 35% found no relationship. The results were similar for boys. The majority of the 16 studies conducted among children from developing societies indicate a clear positive relationship.11 Most recently McMurray and others also found that low SES influenced adolescents' body weight status in the US.44 EI-Tawila and others found that in Egypt the prevalence of obesity among high SES adolescents was more than double that among low SES groups (7.0% versus 3.1%).25

Another interesting finding is that in both Russia and China but not in the US the prevalence of obesity and overweight was higher among children than among adolescents. The differences were especially remarkable regarding obesity. Further research is needed to investigate whether the gap is due to the differences in children's and adolescents' social and behavioral factors, such as diet and physical activity, or if it is because of the WHO/NCHS standard, which is based on data from the US. For example, it has been well documented that thinness is desired in many developed societies, such as the US, and dieting is popular among young American females.45,46 However, ‘the biological picture is not accurately reflected in American popular culture'.46 In other words, the growth patterns of American children and adolescents are not necessarily the optimal patterns for other populations. Furthermore, ample evidence suggests that sexual maturation is related to being fat, and adolescents in many developing countries mature later than their American counterparts.22,47 Therefore we suspect that by using the WHO/ NCHS standard we might have underestimated the obesity problem among adolescents in China and Russia.

Our study has several limitations. Although the NHANES III and the RLMS samples are nationally representative, the CHNS sample is not. Only 8 of China's 31 provinces are covered. China is a country with large heterogenities.26,42 The CHNS sample can not reflex the whole situation in China. Since the CHNS study, however, was designed to monitor nationwide trends, provinces of different geography, economic development, public resources, and health indicators have been included. And, in each province, communities of different SES levels were sampled. Thanks to the close collaboration from local health officials and nutritionist, plus the support of participants, the non-participation rate was very low whenever the subjects were identified and invited to participate in the 1993 survey. Hence, we are confident that the CHNS sample can provide good insights into the SES and obesity relationship in China. In addition, a common weakness of the data sets we used is that information about children and their care-providers' attitudes and values toward obesity and children's physical activity patterns (except for the US sample) were not collected. As a result, we could not make a further investigation about how SES may influence obesity nor to make more comprehensive comparisons among the three countries. Finally, this is a cross-sectional study. We could not prove any causal relationship.34 Nevertheless, our findings provide some important insights into the child and adolescent obesity problem.

In conclusion, obesity is becoming a public health problem influencing children and adolescents in both developed and developing countries. The prevalence of obesity varied remarkably across countries with difference socioeconomic development levels, while within a certain population different SES groups are at different risks. To effectively fight the global obesity epidemic, population-based social and environmental approaches should be considered.

KEY MESSAGES

  • Using an international standard, we studied childhood obesity and the associations between socioeconomic status (SES) factors and obesity in the US, Russia and China.

  • The prevalence of obesity (including overweight) varies remarkably across the three countries of different socioeconomic development levels: 25% in the US, 16% in Russia, and 7% in China.

  • Different SES groups are at different risks, and the relationship between obesity and SES varies across countries.

Table 1

Main characteristics of each sample

 US Russia China 
aMean (SD). 
bBody mass index. 
Years of data collection 1988–1994 1992 1993 
Sample size 6110 6883 3028 
Age (years) 12.4 (5.4)a 12.9 (3.7) 12.5 (4.0) 
Female (%) 48.6 51.5 47.6 
Urban (%) 47.6 70.9 23.5 
Height (cm) 150.6 (3.7) 149.3 (19.2) 140.8 (19.4) 
Weight (kg) 47.6 (4.8) 44.3 (15.4) 36.3 (13.9) 
BMIb (kg/m220.0 (5.6) 19.3 (3.8) 17.6 (3.4) 
 US Russia China 
aMean (SD). 
bBody mass index. 
Years of data collection 1988–1994 1992 1993 
Sample size 6110 6883 3028 
Age (years) 12.4 (5.4)a 12.9 (3.7) 12.5 (4.0) 
Female (%) 48.6 51.5 47.6 
Urban (%) 47.6 70.9 23.5 
Height (cm) 150.6 (3.7) 149.3 (19.2) 140.8 (19.4) 
Weight (kg) 47.6 (4.8) 44.3 (15.4) 36.3 (13.9) 
BMIb (kg/m220.0 (5.6) 19.3 (3.8) 17.6 (3.4) 
Table 2

Prevalence (%) of obesity and overweight among children and adolescents in the US, Russia, and Chinaa

 US Russia China 
 Obese Overweight Obese or overweight Obese Overweight Obese or overweight Obese Overweight Obese or overweight 
aObesity was defined as body mass index (BMI) ≥95th percentile, overweight, 85th≤BMI<95th percentile, Obese or overweight, BMI ≥85th percentile. 
Children (6–9 years) 
    All 12.0 12.4 24.4 13.7 16.8 30.5 7.3 4.6 11.9 
    Boys 11.4 13.7 25.1 13.9 19.5 33.4 6.8 5.9 12.7 
    Girls 12.7 11.0 23.7 13.5 14.0 27.5 7.9 3.1 11.0 
    Rural 12.1 10.7 22.8 16.4 18.0 32.4 6.1 4.4 10.5 
    Urban 11.9 14.2 26.1 12.3 17.3 29.6 11.4 5.2 16.6 
    Low-income 12.1 11.1 23.2 14.5 17.7 32.2 7.6 3.4 11.0 
    Medium-income 11.1 11.9 23.0 13.8 14.2 28.0 8.1 3.7 11.8 
    High-income 13.2 15.0 28.2 12.3 20.7 33.0 5.9 7.9 13.8 
Adolescents (10–18 years)          
    All 10.7 15.2 25.9 3.2 7.4 10.6 1.8 3.0 4.8 
    Boys 11.7 14.0 25.7 4.5 7.0 11.5 1.8 3.0 4.8 
    Girls 9.7 16.4 26.1 2.0 7.7 9.7 1.8 3.0 4.8 
    Rural 11.2 16.0 27.2 3.2 10.1 13.3 1.6 2.8 4.4 
    Urban 10.2 14.2 24.4 3.2 6.3 9.5 2.5 3.5 6.0 
    Low-income 14.0 18.7 32.7 3.1 9.4 12.5 2.1 2.7 4.8 
    Medium-income 11.9 13.6 25.5 4.0 5.1 9.1 0.9 2.5 3.4 
    High-income 5.5 13.5 19.0 2.8 8.0 10.8 2.6 4.0 6.6 
 US Russia China 
 Obese Overweight Obese or overweight Obese Overweight Obese or overweight Obese Overweight Obese or overweight 
aObesity was defined as body mass index (BMI) ≥95th percentile, overweight, 85th≤BMI<95th percentile, Obese or overweight, BMI ≥85th percentile. 
Children (6–9 years) 
    All 12.0 12.4 24.4 13.7 16.8 30.5 7.3 4.6 11.9 
    Boys 11.4 13.7 25.1 13.9 19.5 33.4 6.8 5.9 12.7 
    Girls 12.7 11.0 23.7 13.5 14.0 27.5 7.9 3.1 11.0 
    Rural 12.1 10.7 22.8 16.4 18.0 32.4 6.1 4.4 10.5 
    Urban 11.9 14.2 26.1 12.3 17.3 29.6 11.4 5.2 16.6 
    Low-income 12.1 11.1 23.2 14.5 17.7 32.2 7.6 3.4 11.0 
    Medium-income 11.1 11.9 23.0 13.8 14.2 28.0 8.1 3.7 11.8 
    High-income 13.2 15.0 28.2 12.3 20.7 33.0 5.9 7.9 13.8 
Adolescents (10–18 years)          
    All 10.7 15.2 25.9 3.2 7.4 10.6 1.8 3.0 4.8 
    Boys 11.7 14.0 25.7 4.5 7.0 11.5 1.8 3.0 4.8 
    Girls 9.7 16.4 26.1 2.0 7.7 9.7 1.8 3.0 4.8 
    Rural 11.2 16.0 27.2 3.2 10.1 13.3 1.6 2.8 4.4 
    Urban 10.2 14.2 24.4 3.2 6.3 9.5 2.5 3.5 6.0 
    Low-income 14.0 18.7 32.7 3.1 9.4 12.5 2.1 2.7 4.8 
    Medium-income 11.9 13.6 25.5 4.0 5.1 9.1 0.9 2.5 3.4 
    High-income 5.5 13.5 19.0 2.8 8.0 10.8 2.6 4.0 6.6 
Table 3

Comparison of relationship between obesity and socioeconomic status across countries: OR and 95% CIa

 US Russia China 
a Combined overweight and obesity, defined as body mass index ≥85th percentile. Logistic regression analysis. 
* P < 0.05. 
Age (≥10 years) 1.1 (0.9–1.3) 0.3 (0.2–0.3)* 0.4 (0.3–0.5)* 
Female 1.0 (0.8–1.2) 0.8 (0.7–1.0) 0.9 (0.7–1.2) 
Urban 1.0 (0.8–1.3) 0.8 (0.7–0.9)* 1.5 (1.1–2.0)* 
Low-income 1.3 (1.01–1.6)* 1.4 (1.2–1.7)* 1.2 (0.8–1.7) 
High-income 0.8 (0.7–1.1) 1.2 (1.0–1.4)* 1.5 (1.0–2.1)* 
 US Russia China 
a Combined overweight and obesity, defined as body mass index ≥85th percentile. Logistic regression analysis. 
* P < 0.05. 
Age (≥10 years) 1.1 (0.9–1.3) 0.3 (0.2–0.3)* 0.4 (0.3–0.5)* 
Female 1.0 (0.8–1.2) 0.8 (0.7–1.0) 0.9 (0.7–1.2) 
Urban 1.0 (0.8–1.3) 0.8 (0.7–0.9)* 1.5 (1.1–2.0)* 
Low-income 1.3 (1.01–1.6)* 1.4 (1.2–1.7)* 1.2 (0.8–1.7) 
High-income 0.8 (0.7–1.1) 1.2 (1.0–1.4)* 1.5 (1.0–2.1)* 
Table 4

Relationships between obesity and socioeconomic status by sex: odds ratio and 95% CIa

 Russia China 
 Male Female Male Female 
a Combined overweight and obesity, defined as body mass index ≥85th percentile. Logistic regression analysis. Age was adjusted. For the US, when stratified by gender none were statistically significant (P > 0.05). 
* P < 0.05. 
Urban 0.9 (0.7–1.1) 0.7 (0.5–0.8)* 2.1 (1.4–3.2)* 0.9 (0.5–1.4) 
Low-income 1.2 (0.9–1.5) 1.7 (1.4–2.2)* 1.0 (0.7–1.7) 1.3 (0.8–2.2) 
High-income 1.0 (0.8–1.2) 1.6 (1.2–2.0)* 1.3 (0.8–2.2) 1.6 (0.9–2.8) 
 Russia China 
 Male Female Male Female 
a Combined overweight and obesity, defined as body mass index ≥85th percentile. Logistic regression analysis. Age was adjusted. For the US, when stratified by gender none were statistically significant (P > 0.05). 
* P < 0.05. 
Urban 0.9 (0.7–1.1) 0.7 (0.5–0.8)* 2.1 (1.4–3.2)* 0.9 (0.5–1.4) 
Low-income 1.2 (0.9–1.5) 1.7 (1.4–2.2)* 1.0 (0.7–1.7) 1.3 (0.8–2.2) 
High-income 1.0 (0.8–1.2) 1.6 (1.2–2.0)* 1.3 (0.8–2.2) 1.6 (0.9–2.8) 
Table 5

Relationships between body mass index and socioeconomic status stratified by sex and age: regression coefficients (β) and SEa

 Male  Female    
 β SE P-value β SE P-value 
a Multiple linear regression analysis. Age was controlled for as a continuous variable. 
US 
    All       
        Urban –0.03 0.24 0.900 0.03 0.22 0.876 
        Low-income 0.48 0.30 0.117 0.42 0.25 0.105 
        High-income 0.01 0.37 0.976 –0.66 0.26 0.013 
    6–9 years       
        Urban 0.03 0.19 0.887 0.03 0.23 0.881 
        Low-income 0.48 0.20 0.200 0.31 0.24 0.202 
        High-income –0.39 0.21 0.073 0.05 0.36 0.886 
    10–18 years       
        Urban –0.01 0.22 0.981 –0.06 0.20 0.782 
        Low-income 0.42 0.22 0.060 0.61 0.21 0.006 
        High-income –0.44 0.28 0.117 –0.39 0.24 0.121 
    Russia       
    All       
        Urban –0.21 0.13 0.131 –1.20 0.13 0.000 
        Low-income –0.29 0.16 0.068 1.05 0.14 0.000 
        High-income –0.28 0.14 0.054 0.54 0.14 0.000 
    6–9 years       
        Urban 0.06 0.27 0.826 –0.41 0.30 0.172 
        Low-income 0.15 0.35 0.667 0.90 0.36 0.012 
        High-income –0.26 0.28 0.357 0.16 0.31 0.608 
    10–18 years       
        Urban –0.30 0.16 0.059 –1.47 0.14 0.000 
        Low-income –0.42 0.18 0.016 1.11 0.15 0.000 
        High-income –0.26 0.17 0.115 0.74 0.15 0.000 
    China       
    All       
        Urban 0.64 0.17 0.000 0.09 0.20 0.641 
        Low-income 0.31 0.17 0.066 0.30 0.19 0.119 
        High-income 0.32 0.19 0.086 0.14 0.21 0.520 
    6–9 years       
        Urban 0.68 0.35 0.049 0.69 0.40 0.081 
        Low-income 0.42 0.34 0.217 –0.29 0.37 0.445 
        High-income 0.22 0.40 0.580 –0.01 0.43 0.992 
    10–18 years       
        Urban 0.62 0.19 0.001 –0.12 0.22 0.577 
        Low-income 0.24 0.19 0.193 0.58 0.22 0.009 
        High-income 0.56 0.20 0.006 0.20 0.24 0.397 
 Male  Female    
 β SE P-value β SE P-value 
a Multiple linear regression analysis. Age was controlled for as a continuous variable. 
US 
    All       
        Urban –0.03 0.24 0.900 0.03 0.22 0.876 
        Low-income 0.48 0.30 0.117 0.42 0.25 0.105 
        High-income 0.01 0.37 0.976 –0.66 0.26 0.013 
    6–9 years       
        Urban 0.03 0.19 0.887 0.03 0.23 0.881 
        Low-income 0.48 0.20 0.200 0.31 0.24 0.202 
        High-income –0.39 0.21 0.073 0.05 0.36 0.886 
    10–18 years       
        Urban –0.01 0.22 0.981 –0.06 0.20 0.782 
        Low-income 0.42 0.22 0.060 0.61 0.21 0.006 
        High-income –0.44 0.28 0.117 –0.39 0.24 0.121 
    Russia       
    All       
        Urban –0.21 0.13 0.131 –1.20 0.13 0.000 
        Low-income –0.29 0.16 0.068 1.05 0.14 0.000 
        High-income –0.28 0.14 0.054 0.54 0.14 0.000 
    6–9 years       
        Urban 0.06 0.27 0.826 –0.41 0.30 0.172 
        Low-income 0.15 0.35 0.667 0.90 0.36 0.012 
        High-income –0.26 0.28 0.357 0.16 0.31 0.608 
    10–18 years       
        Urban –0.30 0.16 0.059 –1.47 0.14 0.000 
        Low-income –0.42 0.18 0.016 1.11 0.15 0.000 
        High-income –0.26 0.17 0.115 0.74 0.15 0.000 
    China       
    All       
        Urban 0.64 0.17 0.000 0.09 0.20 0.641 
        Low-income 0.31 0.17 0.066 0.30 0.19 0.119 
        High-income 0.32 0.19 0.086 0.14 0.21 0.520 
    6–9 years       
        Urban 0.68 0.35 0.049 0.69 0.40 0.081 
        Low-income 0.42 0.34 0.217 –0.29 0.37 0.445 
        High-income 0.22 0.40 0.580 –0.01 0.43 0.992 
    10–18 years       
        Urban 0.62 0.19 0.001 –0.12 0.22 0.577 
        Low-income 0.24 0.19 0.193 0.58 0.22 0.009 
        High-income 0.56 0.20 0.006 0.20 0.24 0.397 
Figure 1

Body mass index (BMI) distribution among children (6–9 years) in the US, Russia and China: by sex

Figure 1

Body mass index (BMI) distribution among children (6–9 years) in the US, Russia and China: by sex

Figure 2

Body mass index (BMI) distribution among adolescents (10–18 years) in the US, Russia and China: by sex

Figure 2

Body mass index (BMI) distribution among adolescents (10–18 years) in the US, Russia and China: by sex

Figure 3

Overall prevalence of obesity and overweight in the US, Russia and China

Figure 3

Overall prevalence of obesity and overweight in the US, Russia and China

The study was supported in part by grants from the National Institutes of Health (NIH) (R01-HD30880 and R01-HD38700) and by the Fogarty International Center, NIH, National Science Foundation (grant #37486). I would like to thank Ms Joanna Wang for her comments and Judy Dye and Mary Hill for their editorial assistance. I am very grateful to the two anonymous reviewers for their invaluable comments and suggestions.

References

1
World Health Organization. Obesity: Preventing and Managing the Global Epidemic. Report of a WHO consultation on obesity. Geneva: WHO, 1998.
2
Troiano RP, Flegal KM. Overweight prevalence among youth in the United States: why so many different numbers?
Int J Obes Relat Metab Disord
 
1999
;
23
(Suppl.2):
S22
–27.
3
Mo-suwan L, Junjana C, Puetpaiboon A. Increasing obesity in school children in a transitional society and the effect of the weight control program.
SE Asian J Trop Med Public Health
 
1993
;
24
:
590
–94.
4
Power C, Lake JK, Cole TJ. Measurement and long-term health risks of child and adolescent fatness.
Int J Obes Relat Metab Disord
 
1997
;
21
:
507
–26.
5
Bray GA, Bouchard C, James WPT. Handbook of Obesity. New York: Marcel Dekker, 1998.
6
Must A, Strauss RS. Risks and consequences of childhood and adolescent obesity.
Int J Obes Relat Metab Disord
 
1999
;
23
(Suppl.2):
S2
–11.
7
Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey.
Br Med J
 
2000
;
320
:
1240
–43.
8
Guillaume M. Defining obesity in childhood: current practice.
Am J Clin Nutr
 
1999
;
70
:
126S
–30S.
9
Lahmann PH, Lissner L, Gullberg B, Berglund G. Sociodemographic factors associated with long-term weight gain, current body fatness and central adiposity in Swedish women.
Int J Obes Relat Metab Disord
 
2000
;
24
:
685
–94.
10
Parsons TJ, Power C, Logan S, Summerbell CD. Childhood predictors of adult obesity: a systematic review.
Int J Obes Relat Metab Disord
 
1999
;
23
(Suppl.8):
S1
–107.
11
Sobal J, Stunkard AJ. Socioeconomic status and obesity: a review of the literature.
Psychol Bull
 
1989
;
105
:
260
–75.
12
Sundquist J, Johansson SE. The influence of socioeconomic status, ethnicity and lifestyle on body mass index in a longitudinal study.
Int J Epidemiol
 
1998
;
27
:
57
–63.
13
Sobal J. Obesity and socioeconomic status: a framework for examining relationships between physical and social variables.
Med Anthropol
 
1991
;
13
:
231
–47.
14
Sobal J. Social and economic consequences of overweight in adolescence.
N Engl J Med
 
1994
;
330
:
647
.
15
Stunkard AJ, Sorensen TIA. Obesity and socioeconomic status—a complex relation.
N Engl J Med
 
1993
;
329
:
1036
–37.
16
Economic and Social Commission for Asia and the Pacific. Statistical Yearbook for Asia and the Pacific. Bangkok: ESCAP, United Nations, 1998.
17
World Bank. World Development Indicators 1999 on CD-ROM. Washington, DC: World Bank, 1999.
18
United Nations. Demographic Yearbook 1997. New York: United Nations, 1999.
19
US DHHS. Plan and operation of the third National Health and Nutrition Examination Survey, 1988–94. Vital and Health Statistics. No. 32. Hyattsville, MD: US Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Center for Health Statistics; Washington, DC, 1994.
20
Wang Y, Popkin B, Zhai F. The nutritional status and dietary pattern of Chinese adolescents, 1991 and 1993.
Eur J Clin Nutr
 
1998
;
52
:
908
–16.
21
Zohoori N, Kozyreva P, Kosolapov M et al. Monitoring the economic transition in the Russian Federation and its implications for the demographic crisis—the Russian Longitudinal Monitoring Survey.
World Develop
 
1998
;
26
:
1977
–93.
22
World Health Organization Expert Committee. Physical status, the use and interpretation of anthropometry. WHO Technical Report Series No. 854. Geneva: WHO, 1995.
23
Must A, Dallal GE, Dietz WH. Reference data for obesity: 85th and 95th percentiles of body mass index(wt/ht2 ) and triceps skinfold thickness.
Am J Clin Nutr
 
1991
;
53
:
839
–46.
24
Must A, Dallal GE, WH Dietz. Reference data for obesity: 85th and 95th percentiles of body mass index(wt/ht2 )—a correction.
Am J Clin Nutr
 
1991
;
54
:
773
.
25
El-Tawila S, Omaima ElG, Barbara I et al. Transitions to Adulthood: A National Survey of Egyptian Adolescents. Cairo: Population Council, 1999, pp.25–30.
26
Ge K, Zhai F, Yan H. The Dietary and Nutritional Status of Chinese population: 1992. National Nutrition Survey, Vol. 2 (Children and Adolescents). Beijing: People's Medical Publishing House, 1999.
27
Monteiro CA, Conde WL. Time trends in overweight prevalence in children, adolescents and adults from less and developed regions of Brazil. In: Aihaud G, Guy-Grand B (eds). Progress in Obesity Research. John Libbey & Company Ltd./8th International Congress on Obesity, 1999, pp.665–71.
28
Troiano RP, Flegal KM, Kuczmarski RJ, Campbell SM, Johnson CL. Overweight prevalence and trends for children and adolescents. The National Health and Nutrition Examination Surveys, 1963 to 1991.
Arch Pediatr Adolesc Med
 
1995
;
149
:
1085
–91.
29
Himes JH, Dietz WH. Guidelines for overweight in adolescent preventive services: recommendations from an expert committee. The Expert Committee on Clinical Guidelines for Overweight in Adolescent Preventive Services.
Am J Clin Nutr
 
1994
;
59
:
307
–16.
30
Kuczmarski RJ, Ogden CL, Grummer-Strawn LM et al. CDC Growth Charts: United States. Advance Data. No. 314, 2000. (Website: http://www.cdc.gov/growthcharts)
31
Dietz WH, Robinson TN. Use of the body mass index (BMI) as a measure of overweight in children and adolescents.
J Pediatr
 
1998
;
132
:
191
–93.
32
Wang Y, Wang JQ, Hesketh T, Ding QJ, Mulligan J, Kinra S. Standard definition of child overweight and obesity worldwide [letters].
Br Med J
 
2000
;
321
:
1158
–59.
33
Inoue S, Zimmet P, Caterson I et al. The Asia-Pacific Perspective: Redefining Obesity and its Treatment. 2000 http://iotf.org/ (The document has been co-ordinated by the International Diabetes Institute, a World Health Organization Collaborating Centre; co-sponsored jointly by the Regional Office for the Western Pacific, World Health Organization, the International Association for the Study of Obesity and the International Obesity Task Force.)
34
Rothman KJ, Greenland S. Modern Epidemiology. Philadelphia, PA: Lippincott-Raven, 1998.
35
Kotani K, Nishida M, Yamashita S et al. Two decades of annual medical examinations in Japanese obese children: do obese children grow into obese adults?
Int J Obes Relat Metab Disord
 
1997
;
21
:
912
–21.
36
Reilly JJ, Dorosty AR. Epidemic of obesity in UK children.
Lancet
 
1999
;
354
:
1874
–75.
37
Seidell JC. Obesity: a growing problem.
Acta Paediatr
 
1999
;
88
(Suppl.):
46
–50.
38
Wang Y, Monteiro C, Popkin BM. Child obesity trends in the US are not unique! A comparative 4-country analysis. FASEB J 2000;14:A500 (abstract).
39
Deurenberg P, Yap M, van Staveren WA. Body mass index and percent body fat: a meta analysis among different ethnic groups.
Int J Obes Relat Metab Disord
 
1998
;
22
:
1164
–71.
40
Popkin BM, Baturin A, Kohlmeier L, Zohoori N. Russia: monitoring nutritional change during the Reform Period. In: Wheelock V (ed.). Implementing Dietary Guidelines for Healthy Eating. London: Chapman and Hall, 1997.
41
Neumark-Sztainer D, Story M, Resnick MD, Blum RW. Correlates of inadequate fruit and vegetable consumption among adolescents.
Prev Med
 
1996
;
25
:
497
–505.
42
China National Bureau of Statistics (CNBS). China Statistical Yearbook 1999. Beijing: Publishing House, 1999.
43
Littman MS. A Statistical Portrait of the United States: Social Conditions and Trends. Lanham, MD: Bernan Press, 1998.
44
McMurray RG, Harrell JS, Deng S, Bradley CB, Cox LM, Bangdiwala SI. The influence of physical activity, socioeconomic status, and ethnicity on the weight status of adolescents.
Obes Res
 
2000
;
8
:
130
–39.
45
Bull NL. Studies of the dietary habits, food consumption and nutrient intakes of adolescents and young adults.
World Rev Nutr Diet
 
1988
;
57
:
24
–74.
46
Ritenbaugh C. Body size and shape: a dialogue of culture and biology.
Med Anthropol
 
1991
;
13
:
173
–80.
47
Wang Y, Adair LS. How does maturity adjustment influence the estimates of obesity prevalence in adolescents from different countries using an international reference? Int J Obes Relat Metab Disord 2001; 25:1–9 (in press).