Abstract
The prevalence of obesity increases with age, and obesity increases the risks for a variety of morbidities including hypertension, diabetes, heart disease, stroke, arthritis, and cancer, as well as mortality (McTigue, Hess, & Ziouras, 2006; Osher & Stern, 2009; Salihu, Bonnema, & Alio, 2009). However, unlike the obesity of younger adults, that of the elderly is characterized by increased intra-abdominal fat accumulation despite a somewhat decreased body weight and Body Mass Index (BMI; Perissinotto, Pisent, Sergi, Grigoletto, & Enzi, 2002; Stevens, Katz, & Huxley, 2010; Villareal, Apovian, Kushner, & Klein, 2005), suggesting that waist circumference (WC) is a better measurement for obesity in the elderly (Moon & Kim, 2005; Pouliot et al., 1994). Also, by sex, increasing intra-abdominal fat accumulation generally occurs in middle age in men and in the post-menopausal period in women, reaching maximal fat mass at around 70 years of age, and thereafter fat measures decrease with aging in both sexes (Perissinotto et al., 2002; Stevens et al., 2010; Villareal et al., 2005). Furthermore, obesity increases the risks of morbidity and mortality with age, up to around 75 years, and subsequently declines them (Osher & Stern, 2009; Villareal et al., 2005).
Health-related quality of life (HRQL) is regarded as an important health indicator with regard to obesity in the elderly, in that it measures various aspects of subjective health status and functional status leading to disability (Heim et al., 2011; Lopez-Garcia et al., 2003; Yan et al., 2004). Disability is a predictor for not only diminishing autonomy but also increasing the demand for health services in the elderly (Corona et al., 2013; Osher & Stern, 2009). Prior studies on the relationships between obesity and HRQL in the elderly demonstrate that obesity generally impairs morbidity, a physical component of HRQL, but shows inconsistent results for the other physical components and mental component (Heim et al., 2011; Lopez-Garcia et al., 2003; Yan et al., 2004). The inconsistent findings have been contributed by (a) including mixed race, sex, and age populations without considering their individual effects on obesity and (b) measuring the elderly’s obesity by the use of BMI rather than WC, following the criteria for younger adults. Studies of anthropometric changes demonstrate that the characteristics of increased visceral fat are different depending on ethnicity, sex, and age (Perissinotto et al., 2002; Stevens et al., 2010). Specifically, Asian, women, and older adults have more visceral fat tissue at a specific WC compared with Caucasians or Blacks, men, and younger adults. However, as for the impact of obesity, a meta-analysis study demonstrates that overweight elderly people are not at increased risk with regard to health and mortality (de Hollander et al., 2012). Therefore, a few recent studies have suggested that the criteria for the obesity of older adults should be set at a higher threshold than those of younger adults and should be similar between sexes regardless of race (Heim et al., 2011; So & Yoo, 2012).
Therefore, this study attempts to examine the impact of differential degrees of obesity (measured by WC quintiles) on HRQL by sex among the representative Korean elderly population aged between 65 and 74 years, the most obese age group in this population. The present study’s findings will enhance health providers’ understanding of the elderly’s obesity with regard to HRQL and provide guidelines for the improvement of HRQL in the elderly population.
Method
Study Data and Study Participants
This study extracted the Korean National Health and Nutrition Examination Survey (KNHANES) data between 2008 and 2011 conducted by the Korean Centers for Disease Control and Prevention (KCDC; 2012). This survey used a stratified, multistage probability sampling design based on selection of geographical region, sex, and age. The respondents’ data were assigned weights to ensure the equal probability of being sampled and of covering missing data to select a representative sample of the noninstitutionalized Korean population.
This particular study was limited to elderly respondents aged between 65 and 74 years, with 4,122 individuals in total. The participants were analyzed by sex, 1,786 males and 2,336 females, and according to WC quintiles.
Measures
WC, HRQL, and other demographic, lifestyle, and chronic disease variables were identified on the basis of the results of the health interview and the health examination. The health interview was conducted by trained personnel based on a structured questionnaire. After completion of the health interview, the health examination was conducted in local community health centers and clinics by trained medical staff following standardized procedures.
WC measures
WC was measured to the nearest 0.1 cm with a nonelastic tape at the end of normal expiration at the midpoint between the lower border of the rib cage and the iliac crest while the participants were wearing light clothes. The WC values were categorized by their quintiles.
HRQL measures
HRQL was measured by the EuroQoL-5D (EQ-5D) questionnaire. The EQ-5D questionnaire consists of five dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension score ranges from 1 to 3, with a higher number for each dimension indicating greater limitation for the corresponding dimension. The five dimension scores can also be converted into a single weighted index score using the Korea valuation set developed by the KCDC (Nam, Kim, Kwon, Koh, & Poul, 2007). The single weighted index score may range from −0.171 to 1.00 as a continuous outcome, with 1.00 indicating “full health,” 0 representing “dead,” and the negative scores representing certain health states valued as “worse than dead.” A higher EQ-5D score represents a better state of health.
Covariate measures
Demographic, lifestyle, and chronic disease variables including education, smoking, alcohol consumption, and chronic diseases were collected from the questionnaire. The education level was categorized into one of four groups: completion of elementary, middle, and high school, and university. Smoking was assessed as current smoking (Yes/No). Alcohol consumption was assessed by the typical amount of alcoholic beverage intake at one sitting, categorized into three groups: heavy (measured by more than five cans of beer or seven glasses of any other alcoholic beverage for men and more than three cans or five glasses for women), moderate (measured by less than the amount above), and none (measured as never drank in their lives). Chronic diseases were determined by a doctor’ diagnosis or treatment for the following diseases: hypertension, diabetes mellitus, cardiovascular diseases (including angina pectoris, myocardial infarction, and stroke), musculoskeletal illnesses (including osteoarthritis and rheumatoid arthritis), lung diseases (including pulmonary tuberculosis and asthma), and cancer.
Data Analysis
Data were weighted by the proportion of the geographic area, sex, and age of the total Korean population. All analyses were performed by sex and according to WC quintiles (independent variables). Each mean WC quintile was presented by raising decimals to the next whole number. Demographic, lifestyle, and chronic disease variables (covariates) were displayed by their actual frequencies and weighted percentages. HRQL (dependent variables) was displayed by its weighted means and standard deviations. The χ2 test, and t test and ANOVA were used to examine group differences by sex and according to WC quintiles. Multiple regression analysis was performed to examine the HRQL differences between the reference group (the first WC quintile) and other groups before and after adjustment for covariates. All statistical analyses were performed by the use of SPSS 19.0, which includes procedures for the analysis of complex survey data.
Results
General Characteristics by Sex and According to WC Category
Table 1 represents the general characteristics of the participants by sex and according to WC quintiles. Of a total of 4,122 individuals, men (n = 1,786) accounted for 44.6% of the participants and women (n = 2,336) accounted for 55.4%. The WC quintile values for both men and women were similar: The mean WC values for each quintile were 72, 81, 85, 90, and 97 for men and 71, 79, 84, 89, and 97 for women. There were significant differences between the sexes with regard to education, smoking, alcohol drinking, and presence of hypertension, musculoskeletal illnesses, and lung diseases: Women were found to be less educated, to smoke and drink less, and have a higher prevalence of hypertension and musculoskeletal illnesses but lower prevalence of lung diseases.
General Characteristics of the Study Subjects by Sex and According to Waist Circumference Category (Total N = 4,122, Unit: n [%]).
Note. WC = waist circumference.
Marked for the statistical differences between sexes.
Marked for the statistical differences according to WC quintiles.
p < .05. **p < .01. ***p < .001.
There were also significant differences among the WC quintiles with regard to smoking and presence of hypertension, diabetes, and cardiovascular diseases in men, and education and presence of hypertension, diabetes, cardiovascular diseases, and musculoskeletal illnesses in women: Those in higher WC quintiles were likely to smoke less and have a higher prevalence of hypertension, diabetes, and cardiovascular diseases in men, and be less educated and have a higher prevalence of hypertension, diabetes, cardiovascular diseases, and musculoskeletal illnesses in women.
HRQL by Sex, and According to WC Category
Table 2 presents significant HRQL differences by sex and Table 3 presents them according to the WC quintiles based on the first quintile without and with adjustments for general characteristics. There were significant HRQL differences between the sexes: Women reported a poorer total HRQL score as well as its five dimensions. Whereas men showed no significant HRQL differences according to the WC quintiles, women in the fourth and fifth quintiles reported significantly lower unadjusted HRQL than those in the first quintile by 0.04 and 0.06 points, respectively. Also, with regard to the individual HRQL dimensions, whereas the fourth WC quintile improved usual activities in men, higher WC quintiles impaired mobility, usual activities, and pain/discomfort in women. After adjusting for general characteristics, there were no longer significant total HRQL differences found in either men or women. However, when individual HRQL dimensions were considered, only in women in the fifth WC quintile was mobility significantly worse and in women in the third WC quintile self-care was significantly better.
Mean Scores of EQ-5D Scales by Sex and According to WC Category (Unit: M [SD]).
Note. EQ-5D = EuroQoL-5D; WC = waist circumference.
Marked for the statistical differences between sexes.
p < .05. **p < .01. ***p < .001.
Effects on EQ-5D Scales According to WC Quintiles by Sex.
Note. EQ-5D = EuroQoL-5D; WC = waist circumference.
Negative values for the total index indicate worse health-related quality of life (HRQL), whereas negative values for the components mean better HRQL.
Marked for the statistical differences shown based on the first quintile.
Multiple linear regression, adjusted for influence of education, smoking, drinking alcohol, presence of any chronic diseases.
p < .05. **p < .01. ***p < .001.
Discussion
The divergent findings between the sexes with regard to the impact of obesity on HRQL are confirmatory in other ethnicities: Elderly women show a lower HRQL and impact of obesity measured by BMI on HRQL after eliminating the effects of other confounders than do elderly men (Davison, Ford, Cogswell, & Dietz, 2002; Lopez-Garcia et al., 2003; Yan et al., 2004). However, these studies show inconsistent findings as to HRQL components with regard to weight range impacting HRQL with either underweight or overweight being related or not related to lowered HRQL depending on sex. The present study thus attempts to cast light on these issues using a differentiating range of WC in Korean elderly people in the most obese age group.
In this study, we found that increased WC worsened the HRQL of elderly women, but not that of elderly men. Moreover, when sex differences were taken into account in the relationship between HRQL and WC, it was found that increased WC impaired some HRQL elements (mobility, usual activities, and pain/discomfort) in women, but men with increased WC fared better on one HRQL element (usual activities). We also found that after adjusting for general characteristics, increased WC no longer worsened the HRQL of elderly females. However, when the dimensions of HRQL were analyzed individually, whereas the highest WC quintile of elderly women had impaired mobility, the average WC quintile had improved self-care compared with the lowest WC quintile. Whether adjusting for general characteristics or not, in both sexes, increased WC did not influence psychological health.
As for the findings of lower HRQL associated with WC in Korean elderly women in this study, socioeconomic status (SES) and comorbidity could be explainable obesity-related factors. The majority of Korean elderly women were found to be in a lower SES than the men, and Korean elderly women in a low SES were tended to be more obese due to lifestyle factors including the tendency not to control their diet, activity, and body weight. Furthermore, obese Korean elderly women had a higher prevalence of comorbid chronic diseases other than metabolic syndrome (Kim, Chun, & Kwon, 2011; Park, Park, Oh, & Yoo, 2007; Song et al., 2010). Comorbid chronic diseases and their symptoms and disabilities lower the HRQL of elderly women (Orfila et al., 2006). However, unlike previous studies, the impacts of obesity-related comorbidity are more profound than those of obesity itself in Korean elderly women. The present study supported this finding after adjusting for covariates: whereas increased WC no longer influenced HRQL in elderly women, education and diseases comorbid to cardiovascular diseases, musculoskeletal illnesses, and lung diseases (using covariates) remained significant (data not shown). Based on the general characteristics of the study participants according to WC quintiles by sex (Table 1), elderly women with higher WC are characterized by a higher prevalence of musculoskeletal illnesses, particularly compared with elderly men with higher WC (Kim et al., 2011). That is, osteoarthritis has been known to be prevalent in females, those 55 years and above in age, and those with obesity (Blagojevic, Jinks, Jeffery, & Jordan, 2010; Srikanth et al., 2005). Likewise, osteoarthritis increases the levels of pain and disability (Blagojevic et al., 2010; Srikanth et al., 2005), which worsens the physical dimensions of HRQL such as mobility, usual activities, and pain/discomfort in elderly women in this study. Regardless of the effect of comorbidity, obesity aggregates mobility in Korean elderly women, consistently with previous studies (Lopez-Garcia et al., 2003; McTigue et al., 2006; Osher & Stern, 2009; Villareal et al., 2005). That is, obesity alone and in combination with aging cause a progressive decline in physical function, particularly mobility, due to a continued decrease in muscle mass and strength and an increase in joint dysfunction (Osher & Stern, 2009; Salihu et al., 2009; Villareal et al., 2005).
The range of WC is worth noting when examining sex differences in health outcomes according to the degree of obesity. A previous review study (Stevens et al., 2010) shows that as women age, differences in WC means between sexes are smaller but the WC of men still remains larger than that of women regardless of ethnicity. However, the ranges of WC of both sexes were similar in the age group (between 65 and 74 years) in this study: The mean WC values for each quintile were 72, 81, 85, 90, and 97 for men and 71, 79, 84, 89, and 97 for women. The WC mean value that increased the risk of lowered HRQL was in the fourth quintile, 89 cm in the elderly women, which provides the WC criteria deserving attention. Thus, the standard for central obesity of 85 cm for women (vs. 90 cm for men), drawn by calculating statistically significant or prominent WC increases in absolute risks for the diagnosis for the metabolic syndrome and recommended for Korean adults regardless of age (Lee et al., 2007) cannot be applied to Korean older adults. This signifies that the intra-abdominal fat accumulation is immense in the Korean elderly women aged between 65 and 74 years and mild central obesity is not associated with lower HRQL in the population of aged women.
Unlike that of Korean elderly women, increased WC in Korean elderly men did not worsen HRQL, but rather improved HRQL in the dimension of usual activities, while the dimension of anxiety/depression had a similar ratio but was not statistically significant (in the fourth quintile). Whether obesity (measured by BMI) of elderly men lowers HRQL varies by race (Davison et al., 2002; Lopez-Garcia et al., 2003; Yan et al., 2004). As for the psychological dimension, a study of Schieman, McMullen, and Swan (2007) on the association between obesity (measured by BMI) and various kinds of psychological distress by race and sex demonstrates that the obesity of White women is associated positively with depression/anger, but that of White men is associated negatively, while that of both Black men and women was not associated. Korean men tend to be obese with higher SES, and obese Korean men show better scores on the dimensions of psychosocial and psychological health than the same aged men of normal weight due to lifestyle factors causing obesity and their public values and perceptions on obesity (Park et al., 2007; Song et al., 2010). However, excessively increased WC (the fifth quintile) and the inverse, decreased WC (reference), still should be noted as the worsening cause of HRQL, as supported in a few previous studies of other races in which obesity and underweight (measured by BMI or WC) was related to increased risks of lowered HRQL including mobility limitation and pain in older adults, leading to a U-shaped relationship (Bannerman et al., 2002; Heim et al., 2010, 2011; Yan et al., 2004). Further studies examining the effects of obesity and underweight in the Korean elderly men are required.
Thus, this study has found that extreme obesity worsens mobility, and comorbidity in combination with obesity worsens HRQL including some HRQL dimensions such as mobility, usual activities, and pain/discomfort in Korean elderly women. However, mild obesity does not increase risk for lowered HRQL in Korean elderly women. Monitoring and controlling comorbid chronic diseases and within adequate WC, set by a higher threshold than that of younger adults, will decrease the risks of lowered HRQL in Korean elderly women.
Footnotes
Acknowledgements
The data for this work were provided by the Korean Centers for Disease Control and Prevention (KCDC).
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article was supported by research funds of Chonbuk National University in 2013.