Association between Weight Changes over a 4-Year Period and Health-Related Quality of Life in Middle-Aged and Older Adults in Korea: The Korean Genome and Epidemiology Study Cohort

Article information

J Korean Acad Fam Med. 2024;.kjfm.23.0152
Publication date (electronic) : 2024 June 14
doi : https://doi.org/10.4082/kjfm.23.0152
1Department of Family Medicine, Seoul National University Hospital, Seoul, Korea
2Department of Preventive Medicine, Seoul National University College of Medicine, Seoul, Korea
3Department of Family Medicine, Seoul National University Bundang Hospital, Seongnam, Korea
*Corresponding Author: Woo Kyung Bae Tel: +82-31-787-7805, Fax: +82-31-787-4014, E-mail: naslig@gmail.com, 65597@snubh.org
Received 2023 August 28; Revised 2024 February 12; Accepted 2024 March 16.

Abstract

Background

The relationship between weight change and quality of life remains controversial. This study aimed to investigate whether changes in body weight among participants in different baseline body mass index categories are associated with physical and mental health functioning.

Methods

We conducted an analysis involving 5,106 adults who participated in the Korean Genome and Epidemiology Study, a cohort comprising Korean adults aged 40 to 69 years. We categorized participants into three groups based on body weight change, and physical and mental health were assessed using the 12-Item Short-Form Health Survey in year 4. We employed logistic regression analysis to assess the association between body weight change and poor functioning at year 4. We also utilized a generalized estimating equation to determine the relationship between weight changes and mental component summary (MCS) scores over the study period for each weight group.

Results

Weight gain in both the normal weight (odds ratio [OR], 1.88; 95% confidence interval [CI], 1.13–3.11; P=0.01) and overweight groups (OR, 1.75; 95% CI, 1.05–2.91; P=0.03) was associated with poor MCS. Normal weight weight-losers were associated with a greater increase (2.69 points; 95% CI, 0.50–4.88) in MCS compared to weightmaintainers. Significant differences in mean MCS were observed for overweight weight-losers, obese weight-gainers, and underweight weight-gainers when compared to weight maintainers in each respective weight group.

Conclusion

Different patterns of relationships between weight change and mental health-related quality of life were observed. Hence, it is crucial to focus on the mental health of middle-aged and older adults when assessing body weight changes.

INTRODUCTION

Health-related quality of life (HRQoL) is a multidimensional concept encompassing the subjective perception of physical functioning and mental health. Globally, obesity poses a critical threat to public health, with its prevalence rapidly increasing and the associated burden growing [1,2]. Obesity is a factor linked to quality of life through various pathways that encompass both physical and mental health functioning [3-5]. However, the relationship between obesity and quality of life is inconsistent, with several studies demonstrating negative, positive, or no specific associations [4,6,7]. In addition, weight change, including both weight gain and loss, has been recognized as an important factor influencing quality of life [8,9]. In Asian populations, longitudinal studies investigating the relationship between weight change and quality of life have been relatively scarce compared to cross-sectional designs. Thus, the present study aimed to investigate how changes in body weight influence the quality of life, as measured by the 12-Item Short-Form Health Survey (SF-12), and determine whether these effects depend on baseline body mass index (BMI) among middle-aged and older adults in South Korea.

METHODS

1. Data and Study Participants

Data were retrieved from the population-based longitudinal study, which was part of the Korean Genome Epidemiology Study. The cohort initially included 10,030 participants aged 40 to 69 years, who underwent baseline health examinations and surveys in 2001–2002. These participants were then followed up biennially until 2020. For this study, we focused on 5,106 participants who had completed questionnaires regarding their HRQoL in both 2009–2010 and 2013–2014 (Figure 1). The study protocol was approved by the Institutional Review Board (IRB) of Bundang Seoul National University Hospital (IRB no., X-2308-846-905). The requirement for informed consent from individual patients was omitted because of the retrospective design of this study.

Figure. 1.

Selection of study subjects. SF-12, Short Form 12 Health Survey.

2. Assessment of Exposure Variables

Anthropometric measurements, including height and weight, were conducted by skilled study workers following the standard protocols. BMI was calculated as weight divided by height squared (kg/m2). We used the Asia-Pacific-specific guidelines, with BMI cut-off points of <18.5 kg/m2 for underweight, 18.5 to <23.0 kg/m2 for normal weight, 23.0 to <25.0 kg/m2 for overweight, and ≥25.0 kg/m2 for obesity [10,11]. Participants were categorized into one of three weight-change groups based on their patterns of weight change over the 4-year period: those who lost 3 kg or more, those who gained 3 kg or more, and participants whose weight change remained within 3 kg.

3. Assessment of Outcomes

The SF-12, a self-report questionnaire developed based on the 36-Item Short Form Survey (SF-36), was used to evaluate HRQoL [12]. The SF-12 comprises 12 questions, each rated on a 5-point Likert scale. The sum of the scores was transformed using a standardized scoring algorithm to fit within the 0 to 100 range. The total score encompasses eight subscales, namely physical functioning, role limitations due to physical problems, bodily pain, general health, vitality, social functioning, role limitations due to emotional problems, and mental health. Higher scores on these subscales indicate a better condition. These subscales were further divided into two components: the physical component summary (PCS) and the mental component summary (MCS).

4. Assessment of Potential Mediators

Information on socioeconomic status was collected via a questionnaire, which included details about educational attainment, monthly household income, and employment status. Educational attainment was categorized as either lower (high school or less) or higher attainment (college or above), considering that more than 90% of the population in Korea has been enrolled in high school since 1994 [13]. Monthly household income was divided into two groups: less than 3 and more than 3 million Korean won. In addition, employment status was classified as either employed or unemployed.

5. Assessment of Covariates

Participants provided information about their drinking, smoking, and exercise habits through a health-related behavior questionnaire. A current drinker was defined as drinking alcohol at least once a month. Smoking status was divided into two categories: never or former, and current smokers. A current smoker was defined as someone who has smoked at least 100 cigarettes in their lifetime and does so currently. Physical activity was divided into two categories: regular exercise (defined as engaging in physical activity for at least three episodes per week, with at least moderate intensity) and the inactive group.

We assessed medical history using a self-administered questionnaire, anthropometric measurements, and laboratory biochemical measurements. For defining blood pressure, we calculated the average of the last two of three measured values, considering both systolic and diastolic blood pressure. After at least 12 hours of fasting, blood tests were performed using a Hitachi 700-110 Chemistry Analyzer (Hitachi Co., Tokyo, Japan), and the plasma concentrations of fasting blood sugar, glycated hemoglobin, and postprandial 2-hour glucose level were evaluated. Participants whose laboratory biochemical measurements were consistent with diabetes mellitus, but who had not yet received a diagnosis, were also regarded as having diabetes mellitus [14]. Furthermore, participants diagnosed with coronary artery disease or myocardial infarction were categorized as having cardiovascular disease. Individuals were classified as having hypertension, diabetes mellitus, cardiovascular disease, or cancer (lung, stomach, liver, colon, pancreas, uterus, and breast) if they had received a diagnosis or were currently undergoing treatment for these conditions.

6. Statistical Analysis

The general characteristics of study participants were compared between the three weight-change groups according to obesity status using the analysis of variance for continuous variables and chi-square test for categorical variables. A multivariate logistic regression analysis was performed to determine the association between weight changes and the lowest quartile of PCS and MCS at year 4. Variables that showed statistical significance in the univariate analyses or those known to impact HRQoL were included in the adjusted models. The crude model was adjusted for baseline PCS and MCS, respectively. The following confounders were included in the adjusted model: age, sex, smoking, drinking habit, regular exercise, monthly household income, marital status, education attainment, employment, hypertension, diabetes mellitus, cardiovascular disease, and baseline cancer status. Furthermore, we performed generalized estimating equation analyses to assess the relationship between weight changes and changes in PCS and MCS scores over a 4-year period for each group, including underweight, normal weight, overweight, and obese participants. All statistical analyses were conducted using IBM SPSS Statistics ver. 26.0 (IBM Corp., Armonk, NY, USA), with a significance level set at P<0.05 for all tests.

RESULTS

1. Baseline Characteristics of Study Participants

Table 1 displays the baseline characteristics of participants in relation to changes in body weight over the 4-year follow-up period. Participants shared similar gender distributions; however, they exhibited differences in smoking and drinking habits and the presence of comorbidities such as hypertension and diabetes. Socioeconomic factors, including income and employment status, also varied among the groups. Notably, participants in the weight loss group tended to be older and had lower incomes. Weight gain was associated with a higher prevalence of smoking and a higher employment rate. In comparison to the stable weight group, weight loss was linked to a greater prevalence of hypertension and diabetes. Furthermore, participants who experienced weight loss of 3 kg or more reported considerably worse baseline physical functioning when compared to those with weight gain of 3 kg or more.

Characteristics of the study participants according to body weight change over the past 4 years (N=5,106)

2. Characteristics of Study Participants across BMI Categories

Table 2 presents the characteristics of participants based on their baseline BMI. The mean age was 59.4±8.7 years in the normal BMI group, 58.8±8.1 years in the overweight group, 58.7±8.0 years in the obese group, and 64.2±9.4 years in the underweight group. Participants in the obese group exhibited a higher prevalence of comorbidities, including hypertension and diabetes. Regarding SF-12 components, participants in the obese group reported lower baseline PCS scores.

Characteristics of study participants by baseline BMI categories (N=5,106)

3. Relationship between Body Weight Changes and HRQoL with/without Covariate Adjustment

In the analysis presented in Table 3, weight gain of 3 kg or more was associated with poor MCS scores in the second round after adjusting for baseline mental functioning, lifestyle factors, socioeconomic status, and baseline comorbidities. When we stratified the analysis by baseline BMI, weight gain was associated with poor MCS scores in normal weight and overweight groups, but not in the obese and underweight groups.

Weight change status in relation to the lowest quartile of mental component summary over 4 years according to BMI categories at baseline

When we examined the mean MCS scores using a generalized estimating equation, we found that normal-weight individuals who lost weight were associated with a greater increase (2.69 points; 95% confidence interval, 0.50–4.88) compared to normal-weight individuals who maintained their weight (Table 4). Weight loss was found to result in a similar increase in MCS scores regardless of baseline BMI, although these changes did not reach statistical significance. In contrast, weight gain appeared to be associated with lower MCS scores, except for underweight individuals who reported better MCS scores. Once again, the relationship between weight change and PCS was not statistically significant. As shown in Table 5, estimated differences in mean MCS scores were significant for specific groups: overweight weightlosers (-0.93 points; standard error [SE], 0.33; P<0.001), obese weightgainers (-0.81 points; SE, 0.30; P=0.01), and underweight weight-gainers (2.36 points; SE, 1.13; P=0.04), compared to weight maintainers in each respective weight group (Figure 2).

Weight change patterns in relation to MCS-change over 4 years (from 2009/2010 until 2013/2014) according to BMI category at baseline

Estimated differences in mean scores of MCS by body weight change over the past 4 years according to BMI categories at baseline

Figure. 2.

Mean differences in the 12-Item Short-Form Health Survey mental component summary (MCS) scores for individuals who experienced weight gain or weight loss compared to those with no weight change within each respective baseline weight group. Generalized estimating equation analyses were performed after adjusting for age, sex, smoking, drinking, regular exercise, monthly household income, marital status, education attainment, employment, hypertension, diabetes, cardiovascular diseases, and cancer at baseline. BMI, body mass index. *P-value <0.05.

DISCUSSION

1. Main Findings

In this population-based study using SF-12 to assess HRQoL in community-dwelling middle-aged and older adults in South Korea, we found that those with a weight gain of 3 kg or more had a higher risk of poor mental functioning, especially in normal weight and overweight groups. In addition, weight loss in normal weight groups was associated with improved mental health function. Estimated differences in mean MCS scores between weight gainers/losers and maintainers in every weight group showed distinct patterns: weight gainers in obese groups showed worse functioning, whereas those in underweight groups reported better mental functioning. Interestingly, weight losers in the overweight group reported worse MCS scores.

Our results demonstrated that weight gain increased the risk of poor mental functioning. Furthermore, more importantly, the negative effect of weight gain on mental health remained the same only among normal weight and overweight groups. Our findings are consistent with those of several published studies. For instance, a cross-sectional study conducted among women aged 40–59 years in South Korea found that both non-obese and obese women who experienced weight gain reported higher levels of perceived stress than women who experienced weight loss [8]. Furthermore, weight gain among nonobese women and weight loss among obese women were associated with higher scores on the Patient Health Questionnaire-9. A previous longitudinal study conducted on a random sample of Stockholm County residents investigated the association between 8-year weight change and HRQoL, as measured by the EuroQol-5 Dimension (EQ-5D), based on baseline BMI. The study emphasized that heavy weight gain of 5% of body weight or more increased the risk of reporting impairment in HRQoL regardless of the baseline BMI category, whereas weight reduction had no significant impact [15]. Previous studies also found that compared to those with no weight change, the weight change regardless of obesity was associated with poorer mental health [8,16,17].

Previous studies have also reported inconsistent results regarding the relationship between obesity and the mental components of HRQoL [18,19]. In two large prospective cohorts of US women, the Nurses’ Health Study (in 1992–2000) and the Nurses’ Health Study (in 1992–2000), a modest relation was found between weight change over a 4-year period and HRQoL in the domains of mental functioning, generally remaining below clinically significant levels.4) In the Helsinki Health Study cohort, no statistically significant differences were observed in the changes in MCS scores between weight maintainers and weight gainers [6].

Weight loss is associated with improved mental health functioning in the normal weight group, whereas worse mean MCS scores were reported in the overweight group. Weight change was associated with worse HRQoL in a national sample of older adults in Spain [17]. Specifically, weight loss was linked to lower scores on several physical and mental scales of the SF-36 and a reduction in HRQoL among nonobese women who lost weight and among obese women who gained weight. The association between weight gain and reduction in HRQoL is consistent with our results, whereas the relationship between weight loss and worse mental functioning in non-obese women is inconsistent with our results regarding the relationship in normal weight groups. The explanation could be that we further divided the nonobese group into underweight, normal weight, and overweight groups, which could have led to discrepancies.

Some studies have reported that a weight change is associated with poor physical functioning, whereas our study found no association or clinically insignificant impacts on physical components. These discrepancies may arise from variations in ethnicity, BMI cutoff, statistical methodologies, and the inclusion of different confounding factors in the studies.

We do not have a clear biological explanation for the observation of worse mental HRQoL in weight gainers. Possibly, besides obesity, weight gain also contributes to the genesis of inflammation-mediated mental conditions such as depression and anxiety [20].

We presented further evidence for poor mental health functioning caused by weight changes using a longitudinal dataset from South Korea and provided specific insights into the critical role of weight changes on the mental components of HRQoL. To our knowledge, no previous study has assessed the association between body weight change and the physical and mental components of HRQoL while considering baseline BMI in a prospective design within a large community-dwelling sample in South Korea.

2. Strengths and Limitations of the Study

There are several strengths of our study. First, the longitudinal data were derived from a well-organized, nationally representative cohort study. The follow-up SF-12 results, which were not previously investigated cross-sectionally, were analyzed, allowing us to identify the temporal sequence. Second, weight and height were measured by trained personnel rather than self-reported. Self-reported anthropometric data were associated with misclassification, potentially underestimating the association of weight change with HRQoL. Third, this study is one of the few to compare underweight, overweight, and obese participants with the normal-weight population within the same study. Fourth, blood tests and physical examinations were used to define certain chronic diseases (e.g., hypertension and diabetes mellitus), reducing the potential for recall bias in identifying underlying comorbidities. Finally, our study assessed both PCS and MCS scores because body weight changes may have different impacts on physical functioning and mental health.

However, there are a few limitations to this study, including the following: first, HRQoL was determined based on self-reported questionnaires and subjective reports. Using generic HRQoL measurements rather than obesity-specific HRQoL measurements may not be sufficient to distinguish the association between weight change and other chronic diseases’ impact on HRQoL. Further studies are needed to investigate the underlying mechanisms of the complex relationship between body weight change and HRQoL. Finally, the retrospective design of our study presented limitations regarding the availability of data about intentional versus unintentional weight loss throughout the study periods. A cross-sectional study conducted among the community-dwelling older population aged 75 years and above in Korea indicated that both intentional and unintentional weight loss, regardless of initial weight status, were associated with diminished HRQoL as measured by the EQ-5D instrument in comparison to individuals with no change in weight [21]. Specifically, intentional weight loss contributed to increased levels of anxiety and depression, whereas unintentional weight loss was also related to a similar trend in anxiety and depression, constituting a notable portion of the mental component of HRQoL, without statistical significance. Furthermore, older adults experiencing unintentional weight loss are susceptible to being underweight in response to diseases and age-related changes. Several other variables known to influence unintentional weight loss, including socioeconomic status (including monthly household income, marital status, educational attainment, and employment status), health-related behaviors (smoking, alcohol consumption, and regular exercise), and medical histories (hypertension, diabetes, cardiovascular disease, and cancer), were considered [22]. Our analysis was conducted following adjustments for these factors to mitigate the potential impacts associated with the intentional or unintentional nature of weight loss.

3. Conclusion

Different patterns of relationships between weight changes and mental components of HRQoL were observed within each weight group. Hence, it is crucial to focus on the mental health of middle-aged and older adults when assessing changes in their body weight. Particularly, baseline obesity status may guide target-oriented approaches to support individuals’ mental health affected by weight changes.

Notes

CONFLICT OF INTEREST

No potential conflict of interest relevant to this article was reported.

References

1. GBD 2015 Obesity Collaborators, Afshin A, Forouzanfar MH, Reitsma MB, Sur P, Estep K, et al. Health effects of overweight and obesity in 195 countries over 25 years. N Engl J Med 2017;377:13–27.
2. Swinburn BA, Kraak VI, Allender S, Atkins VJ, Baker PI, Bogard JR, et al. The global syndemic of obesity, undernutrition, and climate change: the Lancet Commission report. Lancet 2019;393:791–846.
3. Wee HL, Wu Y, Thumboo J, Lee J, Tai ES. Association of body mass index with Short-Form 36 physical and mental component summary scores in a multiethnic Asian population. Int J Obes (Lond) 2010;34:1034–43.
4. Pan A, Kawachi I, Luo N, Manson JE, Willett WC, Hu FB, et al. Changes in body weight and health-related quality of life: 2 cohorts of US women. Am J Epidemiol 2014;180:254–62.
5. Ul-Haq Z, Mackay DF, Fenwick E, Pell JP. Association between body mass index and mental health among Scottish adult population: a cross-sectional study of 37,272 participants. Psychol Med 2014;44:2231–40.
6. Svard A, Lahti J, Roos E, Rahkonen O, Lahelma E, Lallukka T, et al. Obesity, change of body mass index and subsequent physical and mental health functioning: a 12-year follow-up study among ageing employees. BMC Public Health 2017;17:744.
7. Damush TM, Stump TE, Clark DO. Body-mass index and 4-year change in health-related quality of life. J Aging Health 2002;14:195–210.
8. Jung KI, Ock SM, Song CH. Weight change and mental health according to obesity in relatively healthy middle-aged Korean women: the Korea National Health and Nutrition Examination Survey 2014, 2016, 2018. J Affect Disord 2021;287:174–81.
9. Zhu J, Xiang YB, Cai H, Li H, Gao YT, Zheng W, et al. Associations of obesity and weight change with physical and mental impairments in elderly Chinese people. Maturitas 2018;108:77–83.
10. WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 2004;363:157–63.
11. World Health Organization. The Asia-Pacific perspective: redefining obesity and its treatment Geneva: World Health Organization; 2000.
12. Ware J Jr, Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996;34:220–33.
13. Lee GB, Kim HC, Jeon YJ, Jung SJ. Association between socioeconomic status and longitudinal sleep quality patterns mediated by depressive symptoms. Sleep 2021;44:zsab044.
14. Jensen MD, Ryan DH, Apovian CM, Ard JD, Comuzzie AG, Donato KA, et al. 2013 AHA/ACC/TOS guideline for the management of overweight and obesity in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and The Obesity Society. Circulation 2014;129(25 Suppl 2):S102–38.
15. Doring N, de Munter J, Rasmussen F. The associations between overweight, weight change and health related quality of life: longitudinal data from the Stockholm Public Health Cohort 2002-2010. Prev Med 2015;75:12–7.
16. Laxy M, Holle R, Doring A, Peters A, Hunger M. The longitudinal association between weight change and health-related quality of life: the KORA S4/F4 cohort study. Int J Public Health 2014;59:279–88.
17. Leon-Munoz LM, Guallar-Castillon P, Banegas JR, Gutierrez-Fisac JL, Lopez-Garcia E, Jimenez FJ, et al. Changes in body weight and healthrelated quality-of-life in the older adult population. Int J Obes (Lond) 2005;29:1385–91.
18. Ul-Haq Z, Mackay DF, Fenwick E, Pell JP. Meta-analysis of the association between body mass index and health-related quality of life among adults, assessed by the SF-36. Obesity (Silver Spring) 2013;21:E322–7.
19. Fine JT, Colditz GA, Coakley EH, Moseley G, Manson JE, Willett WC, et al. A prospective study of weight change and health-related quality of life in women. JAMA 1999;282:2136–42.
20. Hamer M, Batty GD, Kivimaki M. Risk of future depression in people who are obese but metabolically healthy: the English longitudinal study of ageing. Mol Psychiatry 2012;17:940–5.
21. So ES. The impacts of weight change and weight change intention on health-related quality of life in the Korean elderly. J Aging Health 2019;31:1106–20.
22. Wannamethee SG, Shaper AG, Lennon L. Reasons for intentional weight loss, unintentional weight loss, and mortality in older men. Arch Intern Med 2005;165:1035–40.

Article information Continued

Figure. 1.

Selection of study subjects. SF-12, Short Form 12 Health Survey.

Figure. 2.

Mean differences in the 12-Item Short-Form Health Survey mental component summary (MCS) scores for individuals who experienced weight gain or weight loss compared to those with no weight change within each respective baseline weight group. Generalized estimating equation analyses were performed after adjusting for age, sex, smoking, drinking, regular exercise, monthly household income, marital status, education attainment, employment, hypertension, diabetes, cardiovascular diseases, and cancer at baseline. BMI, body mass index. *P-value <0.05.

Table 1.

Characteristics of the study participants according to body weight change over the past 4 years (N=5,106)

Characteristic Weight-Change Pattern
P-value
No change <±3 kg Weight loss ≥-3 kg Weight gain ≥+3 kg
No. of participants 3,710 834 562
Age (y) 58.8±8.2 60.6±8.7 58.0±8.0 <0.001*
Male 1,700 (45.8) 401 (48.1) 279 (49.6) 0.155
Current smoker 504 (13.6) 118 (14.1) 118 (21.0) <0.001*
Current drinker 1,673 (45.1) 337 (40.4) 246 (43.8) 0.047*
Regular exercise 2,177 (58.7) 513 (61.5) 345 (61.4) 0.196
Employed 2,562 (69.1) 553 (66.3) 425 (75.6) 0.001*
Education attainment
 Lower (high school or less) 3,315 (89.4) 749 (89.8) 499 (88.8) 0.832
 Higher (college or above) 395 (10.6) 85 (10.2) 63 (11.2)
Hypertension 1,198 (32.3) 353 (42.3) 161 (28.6) <0.001*
Cancer 19 (0.5) 7 (0.8) 4 (0.7) 0.493
Diabetes mellitus 497 (13.4) 234 (28.1) 50 (8.9) <0.001*
Cardiovascular disease 82 (2.2) 28 (3.4) 11 (2.0) 0.114
Monthly household income (million Korean won)
 <3 2,459 (66.3) 597 (71.6) 380 (67.6) 0.013*
 ≥3 1,251 (33.7) 237 (28.4) 182 (32.4)
Currently married 3,264 (88.0) 731 (87.6) 489 (87.0) 0.797
Short Form 12 Health Survey
 Baseline physical component summary 43.2±3.8 42.7±4.3 43.4±3.9 0.002*
 Baseline mental component summary 47.0±4.7 47.1±5.5 46.7±4.7 0.216

Values are presented as number, mean±standard deviation, or number (%). Analysis of variance for continuous variables and chi-square tests for categorical variables were used to analyze the data.

*

P<0.05.

A history of cancer diagnosis includes liver, colorectal, breast, endometrial, lung, and pancreatic cancer.

Table 2.

Characteristics of study participants by baseline BMI categories (N=5,106)

Characteristic BMI
P-value
Normal weight Overweight Obese Underweight
No. of participants 1,474 1,391 2,155 86
Age (y) 59.4±8.7 58.8±8.1 58.7±8.0 64.2±9.4 <0.001*
Male 661 (44.8) 686 (49.3) 984 (45.7) 49 (57.0) 0.015*
Current smoker 236 (16.0) 222 (16.0) 256 (11.9) 26 (30.2) <0.001*
Current drinker 613 (41.6) 630 (45.3) 973 (45.2) 40 (46.5) 0.126
Regular exercise 939 (63.7) 801 (57.6) 1226 (56.9) 69 (80.2) <0.001*
Employed 1,040 (70.6) 968 (69.6) 1,467 (68.1) 65 (75.6) 0.234
Education attainment
 Lower (high school or less) 1,327 (90.0) 1,237 (88.9) 1,922 (89.2) 77 (89.5) 0.793
 Higher (college or above) 147 (10.0) 154 (11.1) 233 (10.8) 9 (10.5)
Hypertension 341 (23.1) 422 (30.3) 935 (43.4) 14 (16.3) <0.001*
Cancer 6 (0.4) 8 (0.6) 15 (0.7) 1 (1.2) 0.626
Diabetes mellitus 148 (10.0) 185 (13.3) 447 (20.7) 1 (1.2) <0.001*
Cardiovascular disease 29 (2.0) 36 (2.6) 56 (2.6) 0 0.273
Monthly household income (million Korean won)
 <3 1,011 (68.6) 905 (65.1) 1,445 (67.1) 75 (87.2) <0.001*
 ≥3 463 (31.4) 486 (34.9) 710 (32.9) 11 (12.8)
Currently married 1,311 (88.9) 1,225 (88.1) 1,870 (86.8) 78 (90.7) 0.197
Short Form 12 Health Survey
 Baseline physical component summary 43.5±3.6 43.2±3.8 42.9±4.1 43.3±4.5 <0.001*
 Baseline mental component summary 47.1±4.9 46.9±4.7 47.0±4.8 47.1±5.5 0.507

Values are presented as number, mean±standard deviation, or number (%). Analysis of variance for continuous variables and chi-square tests for categorical variables were used to analyze the data.

BMI, body mass index.

*

P<0.05.

Underweight: BMI <18.5 kg/m2; normal weight: BMI 18.5–23.0 kg/m2; overweight: BMI 23.0–25.0 kg/m2; obese: BMI ≥25.0 kg/m2, based on the Asia-Pacificspecific cut-off point for obesity.

A history of cancer diagnosis includes liver, colorectal, breast, endometrial, lung, and pancreatic cancer.

Table 3.

Weight change status in relation to the lowest quartile of mental component summary over 4 years according to BMI categories at baseline

Variable Crude
Adjusted
OR (95% CI) P-value OR (95% CI) P-value
Weight loss ≥-3 kg
 Total 1.06 (0.86–1.31) 0.60 1.04 (0.84–1.29) 0.72
 BMI stratification
  Normal weight 1.29 (0.87–1.91) 0.21 1.28 (0.85–1.92) 0.24
  Overweight 1.07 (0.70–1.64) 0.76 1.02 (0.66–1.57) 0.94
  Obese 0.89 (0.65–1.22) 0.46 0.86 (0.63–1.19) 0.37
  Underweight 2.23 (0.56–8.82) 0.25 3.47 (0.68–17.66) 0.13
Weight gain ≥+3 kg
 Total 1.39 (1.09–1.78) 0.01* 1.40 (1.08–1.80) 0.01*
 BMI stratification
  Normal weight 1.97 (1.21–3.21) 0.01* 1.88 (1.13–3.11) 0.01*
  Overweight 1.72 (1.05–2.83) 0.03* 1.75 (1.05–2.91) 0.03*
  Obese 1.01 (0.70–1.44) 0.98 0.98 (0.67–1.41) 0.90
  Underweight 2.42 (0.27–21.82) 0.43 3.83 (0.22–66.41) 0.36

A multivariate logistic regression analysis was performed after adjusting for age, sex, smoking, drinking habit, regular exercise, monthly household income, marital status, education attainment, employment, hypertension, diabetes mellitus, cardiovascular disease, and cancer (liver, colorectal, breast, endometrial, lung, and pancreatic cancer) at baseline.

BMI, body mass index; OR, odds ratio; CI, confidence interval.

*

P<0.05.

Underweight: BMI <18.5 kg/m2; normal weight: BMI 18.5–23.0 kg/m2; overweight: BMI 23.0–25.0 kg/m2; obese: BMI ≥25.0 kg/m2, based on the Asia-Pacificspecific cut-off point for obesity.

Table 4.

Weight change patterns in relation to MCS-change over 4 years (from 2009/2010 until 2013/2014) according to BMI category at baseline

Variable MCS
PCS
Beta (95% CI) P-value Beta (95% CI) P-value
No change <±3 kg Reference Reference
Weight loss ≥-3 kg
 Normal weight 2.69 (0.50 to 4.88) 0.02* 2.07 (-0.04 to 4.17) 0.05
 Overweight 1.80 (-0.14 to 3.74) 0.07 -0.28 (-2.02 to 1.46) 0.75
 Obese 0.18 (-1.09 to 1.45) 0.78 0.94 (-0.25 to 2.14) 0.12
 Underweight 0.95 (-12.67 to 14.58) 0.89 9.81 (-1.31 to 20.93) 0.08
Weight gain ≥+3 kg
 Normal weight -0.02 (-1.78 to 1.73) 0.98 0.83 (-0.77 to 2.44) 0.31
 Overweight -1.59 (-3.52 to 0.33) 0.11 0.43 (-1.35 to 2.22) 0.63
 Obese -0.41 (-2.14 to 1.33) 0.65 0.27 (-1.26 to 1.79) 0.73
 Underweight 1.98 (-6.76 to 10.73) 0.66 2.81 (-4.75 to 10.37) 0.47

Generalized estimating equation analyses were performed after adjusting for age, sex, smoking, drinking, regular exercise, monthly household income, marital status, education attainment, employment, hypertension, diabetes, cardiovascular diseases, and cancer at baseline.

MCS, mental component summary; BMI, body mass index; PCS, physical component summary; CI, confidence interval.

*

P<0.05.

Table 5.

Estimated differences in mean scores of MCS by body weight change over the past 4 years according to BMI categories at baseline

Variable Normal weight
Overweight
Obese
Underweight
Estimation SE P-value Estimation SE P-value Estimation SE P-value Estimation SE P-value
Weight gain 0.17 0.35 0.63 -0.17 0.33 0.62 -0.81 0.30 0.01* 2.36 1.13 0.04*
Weight loss -0.70 0.40 0.08 -0.93 0.33 <0.001* -0.22 0.24 0.37 1.78 1.57 0.26
No change Ref Ref Ref Ref

Generalized estimating equation analyses were performed after adjusting for age, sex, smoking, drinking, regular exercise, monthly household income, marital status, education attainment, employment, hypertension, diabetes, cardiovascular diseases, and cancer at baseline.

MCS, mental component summary; BMI, body mass index; SE, standard error; Ref, reference.

*

P<0.05.

Underweight: BMI <18.5 kg/m2; normal weight: BMI 18.5–23.0 kg/m2; overweight: BMI 23.0–25.0 kg/m2; obese: BMI ≥25.0 kg/m2, based on the Asia-Pacificspecific cut-off point for obesity.