Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Gender Differences and Socioeconomic Status in Relation to Overweight among Older Korean People

  • Jin-Won Noh,

    Affiliation Department of Healthcare Management, Eulji University, Seongnam, Korea

  • Minkyung Jo,

    Affiliation Graduate School of Public Health, Korea University, Seoul, Korea

  • Taewook Huh,

    Affiliation Strategy Department for Women-friendly Policy, Korean Women’s Development Institute, Seoul, Korea

  • Jooyoung Cheon,

    Affiliation School of Nursing, University of Maryland, Baltimore, Maryland, United States of America

  • Young Dae Kwon

    Affiliation Department of Humanities and Social Medicine, College of Medicine and Catholic Institute for Healthcare Management, the Catholic University of Korea, Seoul, Korea

Gender Differences and Socioeconomic Status in Relation to Overweight among Older Korean People

  • Jin-Won Noh, 
  • Minkyung Jo, 
  • Taewook Huh, 
  • Jooyoung Cheon, 
  • Young Dae Kwon



The ever-increasing older population and its association with serious overweight problems have garnered much attention. The correlation between being overweight and socioeconomic status factors could be helpful for understanding the inequalities among the overweight population. We examined the correlation between being overweight and some key variables, such as demographics, socioeconomic status, general health status, and health behavior in a large sample of older individuals, by each gender.


We used data from the 2008 Korean Longitudinal Study of Aging and it included 8,157 participants who were 45 years or older. To understand the relationship between the overweight participants in accordance to demographic and socioeconomic characteristics, health status, and health behaviors, a weighted chi-square test and logistic regression analysis were conducted by separating variables related to overweight, according to the genders.


The number of people in the normal group was 6,347 (77.8%), while the people who were considered overweight were 1,810 (22.2%). Women (n = 4,583) constituted 52.7% of the subject, 24.9% of whom were classified as overweight. Meanwhile, 20.6% of the 47.3% (n = 3,574) of the sample who were men were classified as overweight. Participants between the ages of 45 and 64 with chronic diseases were more likely to be overweight. Men in the 4th quartile of household income were more likely to be overweight than those who were in the 1st quartile, in contrast, while unemployed women with lower education levels and urban residents were at greater risk for being overweight.


Among the men, health status and health behavior appeared to show a correlation with being overweight; however, among women, socioeconomic status factors were strongly related to being overweight. These findings appear to support the association of gender-specifics with the prevalence of being overweight.


The increasing trend of overweight individuals has become a major health issue over the past several decades [1]. According to the World Health Organization (WHO), it was estimated that more than 1.4 billion adults were overweight in 2008, and among them, at least 500 million were obese. This was an increase of over 200% since 1980 [1].

South Korea is one of the fastest industrialized countries. The prevalence of overweight among Koreans has increased during the past several decades due to socioeconomic development and the increase in life expectancy [2], [3]. According to the Korea National Health and Nutrition Examination Survey data, the prevalence of overweight [body mass index (BMI), 25.0–29.9 kg/m2] increased from 26.7% in 1998 to 30.9% in 2009.

Since Sobal and Stunkard's exhaustive review comparing the association between socioeconomic status (SES) and overweight [4], it was found that there may be a correlation between being overweight and economic development. There is an inverse association between overweight and SES among women in developed societies, whereas the association is direct in developing countries [5][7]. These findings stress significant gender differences in the association of overweight prevalence with SES factors [7].

SES factors, such as education level, income, and employment status, have a differential effect on body weight and fat distribution [8], [9]. Generally, people who were considered to have lower education were more highly associated with being overweight [10][12]. Some studies report that women with a lower education level are more likely to have a significantly higher BMI than compared to those with a higher education levels, even though there are no significant differences that were found in men [8], [13], [14]. Higher income was also found to have an association with a higher risk of being overweight [10]. Men with higher income typically have a greater BMI and a higher risk of being overweight than those who earn less income [8], [15], [16]. In contrast, for women, the prevalence of being overweight increased with lower income [15], [16]. Ball, Mishra, and Crawford reported that low-status employed women showed the higher risk of being overweight than high-status employed women [17]. However this finding was not observed among men. A study reported that SES, represented by income and education levels, was positively associated with overweight in Korean men, whereas a negative association between SES and overweight was found in Korean women [18]. A study of people in a rural community in Korea reported that the prevalence of overweight was higher among men with higher income as well as less education [19].

Interestingly, marital status is also an important factor that affects overweight. Married individuals are more likely to have higher BMI or have a higher risk of being overweight, particularly men [12], [13], [20], [21]. Also, location of residence was considered to be a factor because of the differential associated diet, physical activity, and environment (urban vs. rural) [14]. However, the relationship between location and being overweight was not evident. Padez reported that adult males who lived in rural areas had a lower risk of being overweight than those who lived in urban area [22], while a study by Tchicaya and Lorentz showed that adults who lived outside of urban areas were at a higher risk of being overweight [14].

Overall, gender differences in relation to the association between overweight prevalence and SES factors would be helpful for understanding the inequalities in being overweight; however, there is a lack of information about what actual risk factors affect the overweight occurrence among older people by gender in Korea [23]. Therefore, the aim of this study is to examine gender differences in the risk factors of being overweight, focusing on the SES factors among Koreans over 45 years of age. It is hypothesized that being overweight is associated with the SES, but the pattern of association would be different by gender.


1. Data and Subjects

We used data from the 2008 Korean Longitudinal Study of Aging (KLoSA), which was obtained from a public repository ( The Korea Labor Institute conducted KLoSA, funded by the Korean Ministry of Labor, to identify the aging progress of Koreans and to utilize its findings in academic research and policy-making. KLoSA has gathered data using computer assisted personal interviewing. The response rate is 86.6% and hot deck imputation method has been used to deal with missing data. The population of KLoSA participants includes adults aged 45 and over. The current study was carried out using data from 8,157 of the original 8,688 KLoSA subjects, after excluding subjects who were underweight or were missing BMI data. Since the database contains only de-identified data, this study does not pose any risks to study subjects. This study was also approved by the Institutional Review Board of the Catholic University of Korea with a waiver for informed consent because the data obtained from a public database.

2. Variables

Dependent variables.

Based on WHO criteria, people with BMI of less than 18.5 kg/m2 (underweight) were excluded from the analysis because this research focused only on the overweight. People with a BMI from 18.5 to 25 kg/m2 were classified as ‘normal’ and those with a BMI of 25 kg/m2 or greater was considered as ‘overweight’. Other independent variables were defined using a questionnaire survey of the KLoSA.

Independent variables.

Demographic variables included: age and gender. Socioeconomic variables included: marital status, level of education, residence, general quality of life (QOL), employment, household income, and medical insurance type. Marital status was classified into two groups: married and other status (divorced, parted by death, separated, single). Level of education was divided into four levels: college or higher than high school, middle school diploma, elementary school and less than elementary school. Relative QOL was assessed through peer comparison using a 10-point interval scale from 0 to 100 points. A higher score meant a higher QOL and the QOL range was divided into three categories at the less than 60, 60–80, and more than 80 point levels. Economic status was assessed based on annual household income in the previous year (2007). Medical insurance type was divided into National Health Insurance and Medical Aid.

Health status variables included: self-rated health, whether they have depressive symptom, whether they have disabilities, and whether they have chronic disease. The presence of depressive symptoms was assessed by the Korean version of the Center for Epidemiologic Studies Depression Scale (CES-D) survey. A simplified form of the CES-D, the CES-D10 is a questionnaire composed of 10 questions and was developed as a depressive symptom screening scale for epidemiological investigations. In the Korean version of CES-D10, answers to the question on the frequency of experiencing depressive symptoms during the past week were composed of four choices: ‘for a while’ (less than one day), ‘sometimes’ (from one to two days), ‘often’ (from three to four days), and ‘at all times’ (from five to seven days). We assigned answers of ‘for a while’ and ‘sometimes’ 0 points, and ‘often’ and ‘at all times’ 1 point, and set a cutoff of more than 4 points in 10 questions as having depressive symptoms. According to medical certification of their disability, the respondents were divided into two groups: impaired and non-impaired. We determined that a patient has chronic disease if he/she has at least one of the following items related to chronic illness: high blood pressure, diabetes, cancer, chronic lung disease, liver disease, heart disease, cerebrovascular disease, arthritis/rheumatism, prostate disease, and urinary incontinence.

Health behavior included: whether the respondents exercised, smoked, and/or drank. Respondents who answered yes to the question, “Do you usually exercise more than once a week?” were classified as the ‘regular exercise group,’ while those who said no to the same question were classified as the ‘no exercise group.’ Regarding smoking, the respondents were classified into current smokers, former smokers, and non-smokers. The same standard was applied to drinking.

3. Statistical Analysis

The chi square test and simple logistic regression analysis were conducted in order to determine the relationship among socio-demographic characteristics, health status and health behaviors based on gender and to identify the variables that show differences in normal weight and overweight by gender. Interaction effect analysis was examined between gender and the rest of the variables, and stratified analysis was conducted for each gender. In order to identify the factors that influence overweight based on gender, two models of multivariable logistic regression analysis was conducted for each gender group. Model 1 showed the association of age and socioeconomic position with overweight varied by gender. Model 2 included the addition of variables of health status and health behavior. All of the statistical analyses were two sided, and P<0.05 was considered to be significant. In order to generalize the analysis results through a sample survey, the KLoSA analyzed the survey results by subject, applying the weighted value which was calculated by taking into account the benchmarking adjustment values using extraction rate, response rate, and other external information. For all statistical analyses in this study, SAS ver. 9.2 (SAS Institute, Cary, NC, USA) was used, and all analyses were tested at a 95% confidence level.


1. General Characteristics of Subjects

Out of the total 8,157 subjects, the number of people in normal group was 6,347 (77.8%), while that of people who were overweight was 1,810 (22.2%). In the case of marital status, men were higher than women in percentage. Significantly more men (19.7%) than women (5.3%) had an education level of college or higher, and, naturally, fewer men (23.5%) than women (50.3%) had less than an elementary school education. With employment, men recorded a much higher employment rate of 69.2% than women’s 34.1%. Likewise, in self-rated health, 50.2% of men reported excellent or good health in contrast to 33.7% of women who reported the same. For such health status variables as depression, disability and chronic diseases, women were disproportionately associated with negative health factors, and men were more likely to be smokers and drinkers. In other areas such as residence, general QOL, household income, medical insurance, and exercise, men and women showed similar rates of distribution to each other (Table 1).

Table 1. Characteristics of the study population (n = 8,157), Korea, 2008*.

2. Relationships between Demographic and Social Characteristics and Health Behaviors with Overweight

Table 2 displays the results from chi square test which indicates which variables among socio-demographic characteristics, health status, and health behaviors show differences in normal weight and overweight by gender. Results from examining the interaction effect between gender and the rest of the variables showed that most of the variables had significant interaction effect except for marital status, residence, and drinking. Residence, employment, and chronic diseases were found to have a significant relationship with being overweight in both men and women. As for age, QOL, household income, insurance, depression (CES-D10), and exercise, a significant relationship was identified in men, but not women. Level of education, self-rated health, and presence of a disability showed significance only among women (P<0.05) (Table 2).

Table 2. Chi square test on demographics, socioeconomic characteristics, and health behaviors with overweight.

3. Factors Affecting Overweight among Demographic and Social Characteristics and Health Behaviors

In the case of men, the overweight OR did not significantly decrease for those in the 55–64 age group compared to the 45–54 age group, but did for the 65–74 age group and the over 75 age group. Thus, the OR of overweight decreased with older age. The findings showed that the overweight OR was higher for those in the 4th quartile of household income than the 1st quartile in Model 1 and Model 2. The result of Model 2 for men showed that age, level of education, household income, chronic diseases, and exercise were significant variables. It was found that men with less education than elementary school level were more likely to belong to the overweight group compared to men from college or higher (OR, 1.51; 95% CI, 1.08–2.13). The probability of being overweight among men increased significantly with chronic diseases (OR, 1.94; 95% CI, 1.56–2.42) and exercise (OR, 1.06; 95% CI, 1.01–1.12) (Table 3).

Table 3. Multivariable logistic regression analysis on BMI [for 25 kg/m2≤ BMI (compared to 18.5 kg/m2≤ BMI <25 kg/m2)].

On the other hand, when compared to the 45–54 age group women, the OR of overweight was significantly lower only in the over 75 age group. Women with high school, middle school, and less than elementary school level education were more likely to be overweight than those who had a college or higher degree and the proportion of overweight was higher among women with living urban in both Model 1 and Model 2. Unemployed women showed a higher risk of being overweight (OR, 1.26; 95% CI, 1.05–1.51) in Model 1. Age, level of education, residence, disability, and chronic diseases were significant variables among women in Model 2. The proportion of being overweight was higher women with a disability (OR, 1.25; 95% CI, 1.02–1.52) and women with chronic diseases (OR, 2.16; 95% CI, 1.82–2.56) (Table 3).


In this study we examined the correlation between overweight and some key variables, such as demographics, SES, health status, and health behaviors, in a large sample of older individuals (over 45 years of age). This research showed first that health status (chronic diseases) and age had a higher correlation with the risk of being overweight – particularly, younger men and women. Among men, those who exercise were more likely to be overweight, whereas among women, SES factors were strongly related to being overweight. In addition, among men, household income was strongly associated with being overweight. Among women, the level of education (high school, middle school, less than elementary school) was significantly related to being overweight in both Models 1 and 2. In short, these findings may support gender-specific associations with the prevalence of overweight.

Regarding the demographics, we found that age may have a relationship with the increased risk of being overweight – being in inverse proportion. While the prevalence of being overweight in the 65 years and older decreased relatively, the prevalence of overweight in the 45–64 age group increased. This result was not consistent with previous research that supports the association that the prevalence of overweight increases with age – directly proportional [10], [11], [13], [24][26]. One possible explanation for our results is that environmental factors, such as cultural background and dietary habits, may be contributing factors [27]. For example, Koreans in the 45–64 age group may appear to more frequently engage in lifestyles involving heavy dinners and alcohol consumption. In addition, office workers in Korea usually drink to increase sociability and reduce tension [28]. According to a survey conducted in 2005, 53.3% of white-collar employees in Korea go for a drink with colleagues more than twice a week, and 36.2% of employees have an incidence of heavy drinking once in a week [29].

In relation to SES factors, the following important findings were revealed. The prevalence of being overweight appears to be significantly associated with a lower level of education than college or higher, particularly among women. We found that women with a lower level of education had higher significant relationship with overweight than their more educated counterparts (‘high school diploma’). This is consistent with the results of a large number of studies that demonstrate that being female with a high SES in a developed country is inversely related to being overweight [17], [18]. This result is in line with the idea that females with upper SES (including the employed with higher level of education) may be more interested in maintaining a good body figure for achieving greater confidence and work efficiency as well. In addition, it may reflect another fact that women in Korea used to be discriminated against in terms of education. In the past, the disparities in education background between men and women were huge, and only a few women could attain a high level of education. Therefore, levels of education and social activities among women (rather than men) have a more positive association with a physical and psychological well-being lifestyle. This also demonstrates that the Korean society places importance on physical appearances of women.

Also, we found that people living in urban areas were more likely to be obese, and females living in urban areas had an even stronger relationship with an increased risk of being overweight. As argued in a previous study [30], this relationship may be because urban dwellers tend to do less leisure-time physical activity and have less desirable eating habits. This relationship also shows that characteristics of the social environment may relate to trends in weight gain [12]. However, among men, there was no relationship with living in urban areas when we did multiple logistic regression analysis. It could be explained that the associations with other variables was stronger than that of the location of their home.

Meanwhile, in relation to the health status, although generally assumed, we particularly identified again that the critical factor associated with the risk of overweight, regardless of gender, appears to be chronic diseases. The presence of chronic diseases showed that those who have chronic diseases, regardless of age, are more obese than those without. Many studies have shown that overweight is strongly associated with an increased risk for many chronic diseases, including osteoarthritis, diabetes, hypertension, and hypercholesterolemia [27], [31].

However, there are several limitations in this study. Although household income only had no for female in this study, the association between level of income and the prevalence of overweight has been identified in previous research [8], [32]. In the current study we did not include income from property, which can be an important household income source. With more exact sources of household income (economic determinant) and social status factors that include various social activities and involvement in social and health groups, the current study could have even better defined the relationship between SES and the risks of overweight. Another limitation is that this study does not show time-series analysis since it is a cross sectional study. Therefore, through this research, any time series changes and findings are not analyzed and/or explained. Therefore, a longitudinal study that could further determine cause-outcome relationships is necessary. It can also be argued that the limitations of sampling survey are overcome from applying the weighted value to solve these problems of the sampling survey. Also, in terms of generalization of the above findings, it is critically required to look at different contexts of SES factors in different countries and societies as well. These limitations are areas for meaningful further study and the need for careful interpretation of the findings in this study.


This study looked at the gender difference in the risk of being overweight in relation to SES among older Koreans. It reveals that there are some important gender-specific associations with the risk of overweight. While males were found to have a significantly more associations with health status and health behavior, such as chronic diseases and exercise, female were found to be strongly related to socioeconomic factors, such as education, residence, and employment status. Culturally, it is possible that female with upper SES may be more strongly required to manage her body shape than male. This study shows that socioeconomic factors contributing to being overweight are different by gender. Therefore, it is necessary to consider the different risk of SES by gender in implementing a comprehensive national policy or initiative program to reduce and prevent the increasing problem of being overweight.

Author Contributions

Conceived and designed the experiments: JWN YDK. Analyzed the data: MKJ JYC. Wrote the paper: JWN TWH YDK.


  1. 1. WHO Media Centre. Obesity and overweight. Fact sheet N°311. Available: Accessed 2013 Mar 5.
  2. 2. Oh SW (2011) Obesity and metabolic syndrome in Korea. Diabetes Metab J 35: 561–566.
  3. 3. Korean National Statistical Office (2012) Population projections for Korea. Available: Accessed 2013 Feb 10.
  4. 4. Sobal J, Stunkard AJ (1989) Socioeconomic status and obesity: a review of the literature. Psychol Bull 105: 260–275.
  5. 5. Ball K, Crawford D (2005) Socioeconomic status and weight change in adults: a review. Soc Sci Med 60: 1987–2010.
  6. 6. McLaren L (2007) Socioeconomic status and obesity. Epidemiol Rev 29: 29–48.
  7. 7. Wells JC, Marphatia AA, Cole TJ, McCoy D (2012) Associations of economic and gender inequality with global obesity prevalence: understanding the female excess. Soc Sci Med 75: 482–490.
  8. 8. Reas DL, Nygård JF, Svensson E, Sørensen T, Sandanger I (2007) Changes in body mass index by age, gender, and socio-economic status among a cohort of Norwegian men and women (1990–2001). BMC Public Health 7: 269.
  9. 9. Vernay M, Malon A, Oleko A, Salanave B, Roudier C, et al. (2009) Association of socioeconomic status with overall overweight and central obesity in men and women: the French nutrition and health survey 2006. BMC Public Health 9: 215.
  10. 10. Grujić V, Martinov-Cvejin M, Ač-Nikolić E, Dragnić N, Mijatović-Jovanović V, et al. (2009) Association between obesity and socioeconomic factors and lifestyle. Vojnosanit Pregl 66: 705–710.
  11. 11. Rodríguez-Martín A, Novalbos Ruiz JP, Martinez Nieto JM, Escobar Jimenez YL (2009) Life-style factors associated with overweight and obesity among Spanish adults. Nutr Hosp 24: 144–151.
  12. 12. Sund ER, Jones A, Midthjell K (2010) Individual, family, and area predictors of BMI and BMI change in an adult Norwegian population: findings from the HUNT study. Soc Sci Med 70: 1194–1202.
  13. 13. Martin AR, Nieto JM, Ruiz JP, Jimenez LE (2008) Overweight and obesity: the role of education, employment and income in Spanish adults. Appetite 51: 266–272.
  14. 14. Tchicaya A, Lorentz N (2012) Socioeconomic inequality and obesity prevalence trends in Luxembourg, 1995–2007. BMC Res Notes 5: 467.
  15. 15. Roskam AJ, Kunst AE (2008) The predictive value of different socio-economic indicators for overweight in nine European countries. Public Health Nutr 11: 1256–1266.
  16. 16. Sabanayagam C, Shankar A, Saw SM, Tai ES, Wong TY (2009) The association between socioeconomic status and overweight/obesity in a Malay population in Singapore. Asia Pac J Public Health 21: 487–496.
  17. 17. Ball K, Mishra G, Crawford D (2003) Social factors and obesity: an investigation of the role of health behaviours. Int J Obes Relat Metab Disord 27: 394–403.
  18. 18. Park HS, Park CY, Oh SW, Yoo HJ (2008) Prevalence of obesity and metabolic syndrome in Korean adults. Obes Rev 9: 104–107.
  19. 19. Kim MY, Oh JK, Lim MK, Yun EH, Kang Y (2012) The association of socioeconomic and psychosocial factors with obesity in a rural community. Korean J Obes 21: 18–28.
  20. 20. Cho J, Juon HS (2006) Assessing overweight and obesity risk among Korean Americans in California using World Health Organization body mass index criteria for Asians. Prev Chronic Dis 3: A79.
  21. 21. Sobal J, Hanson KL, Frongillo EA (2009) Gender, ethnicity, marital status, and body weight in the United States. Obesity (Silver Spring) 17: 2223–2231.
  22. 22. Padez C (2006) Trends in overweight and obesity in Portuguese conscripts from 1986 to 2000 in relation to place of residence and educational level. Public Health 120: 946–952.
  23. 23. Kim IH, Chun H, Kwon JW (2011) Gender differences in the effect of obesity on chronic diseases among the elderly Koreans. J Korean Med Sci 26: 250–257.
  24. 24. Santos R, Aires L, Santos P, Ribeiro JC, Mota J (2008) Prevalence of overweight and obesity in a Portuguese sample of adults: results from the Azorean Physical Activity and Health Study. Am J Hum Biol 20: 78–85.
  25. 25. Hajian-Tilaki KO, Heidari B (2007) Prevalence of obesity, central obesity and the associated factors in urban population aged 20–70 years, in the north of Iran: a population-based study and regression approach. Obes Rev 8: 3–10.
  26. 26. Jitnarin N, Kosulwat V, Rojroongwasinkul N, Boonpraderm A, Haddock CK, et al. (2010) Risk factors for overweight and obesity among Thai adults: results of the National Thai Food Consumption Survey. Nutrients 2: 60–74.
  27. 27. Kim JY, Chang HM, Cho JJ, Yoo SH, Kim SY (2010) Relationship between obesity and depression in the Korean working population. J Korean Med Sci 25: 1560–1567.
  28. 28. Kim O, Jeon HO (2012) Relations of drinking motives and alcohol consumption in Korean male office workers. Psychol Rep 111: 963–970.
  29. 29. Kwon GY (2005) Factors influencing drinking of employees: focus on the white collar employees. Korean J Soc Welfare 57: 93–118.
  30. 30. Gutiérrez-Fisac JL, Guallar-Castillón P, Díez-Gañán L, López García E, Banegas BanegasJR, et al. (2002) Work-related physical activity is not associated with body mass index and obesity. Obes Res 10: 270–276.
  31. 31. Park HJ, Lee SK (2011) Association of obesity with osteoarthritis in elderly Korean women. Maturitas 70: 65–68.
  32. 32. Ball K, Mishra G, Crawford D (2002) Which aspects of socioeconomic status are related to obesity among men and women? Intl J Obes Relat Metab Disord 26: 559–565.