In recent decades, cardiovascular disease (CVD) has been the leading cause of death worldwide.^1,2 Among the risk factors for CVD, dyslipidemia is a preventable contributor.³ In a young adult group consisting of individuals in their 20s and 30s, dyslipidemia has been found to contribute to CVD.⁴ However, awareness of dyslipidemia is low among young adults compared with those in older age groups.⁵

Previous studies have reported a relationship between weight changes and lipid profiles. Weight loss has been associated with an increase in high-density lipoprotein (HDL) cholesterol and a decrease in total cholesterol, low-density lipoprotein (LDL) cholesterol, and triglycerides.^6-10 Conversely, weight gain has been associated with a decrease in HDL cholesterol and an increase in total cholesterol, LDL cholesterol, and triglycerides.^6,8,9,11 This has been observed not only in Western populations but also in Asian populations.^6,8 Furthermore, it has been observed in both sexes.^9,12 However, there is a lack of research investigating the relationship between weight changes and dyslipidemia in young adults.

Due to changing lifestyle patterns, the incidence of dyslipidemia in Korea has increased significantly.¹³ In addition, a recent study found a higher prevalence of hypercholesterolemia and hypertriglyceridemia in Korean men than in women, and they showed less interest in weight control.¹⁴ Weight gain has been more prominent among young men compared with older age groups or women.¹² However, as lipid profiles are not typically assessed for young men, there is a lack of research on the effect of weight changes on dyslipidemia in this population. As the majority of young Korean men are required to serve in the military and undergo health screenings during their service, we used military health screening data to investigate the effect of weight changes on dyslipidemia in this population.

Study population

The health screening data of soldiers were collected through the Armed Forces Medical Command Defense Medical Information System database. Soldiers who had received both private health screenings (conducted immediately after transferring to a military unit) and corporal health screenings (conducted either before or after promotion to the corporal rank) were enrolled. As military service obligations apply exclusively to males in South Korea, the study participants consisted entirely of males. To ensure sufficient time for detecting differences in dyslipidemia associated with changes in body weight, individuals who had received two health screenings within a 90-day period were excluded.¹⁵ Individuals with missing data on anthropometric measurements, health questionnaires, or blood tests were also excluded. Finally, individuals who were already taking medications for dyslipidemia were excluded. This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and Good Clinical Practice guidelines. This study was approved by the Institutional Review Board of the Korean Armed Forces Medical Command (AFMC-202211-HR-041-01). The Korean Armed Forces Medical Command’s Institutional Review Board waived the requirement for patient consent due to the study’s retrospective nature and the use of anonymized clinical data.

Primary measure

Weight was measured with the participants wearing lightweight clothing without shoes and was recorded to the nearest 0.1 kg. Changes in weight were calculated as the difference between weight at the corporal health screening and weight at the private health screening. As previous studies investigating the association between weight change and dyslipidemia have used a 5% or 10% weight change to quantify changes,^6,11 we categorized weight change into five groups as follows: ≥10% weight loss, 5%–10% weight loss, <5% weight change (considered as no weight change), 5%–10% weight gain, and ≥10% weight gain.

Baseline covariates

Covariates that potentially affect dyslipidemia, such as age, baseline body mass index (BMI), smoking, regular exercise, and hypertension, were collected from private health screenings.^6,16-19 Anthropometric measurements included height, weight, and blood pressure. Height was measured while wearing lightweight clothing without shoes (to the nearest 0.1 cm). BMI was calculated by dividing weight in kg by height in m². BMI was classified into four categories: underweight (BMI <18.5 kg/m²), normal (BMI 18.5 to 23 kg/m²), overweight (BMI 23 to 25 kg/m²), and obesity (BMI ≥25 kg/m²).²⁰ Blood pressure was measured after 5 minutes of rest while wearing lightweight clothing. Hypertension was defined as either systolic blood pressure ≥140 mmHg, diastolic blood pressure ≥90 mmHg, or the use of antihypertensive medications during the baseline assessment.²¹ In addition, health questionnaires included questions about current medication use, smoking status, and regular exercise habits, all of which were assessed at the baseline. Current smokers were defined as individuals who currently smoke, excluding past smokers. Regular exercise was defined as sweating during exercise at least three times per week.⁵

Outcomes

Dyslipidemia was the primary outcome, which was defined as meeting any of the following criteria: hypercholesterolemia (total cholesterol ≥240 mg/dL), hyper-LDL-cholesterolemia (LDL cholesterol ≥160 mg/dL), hypo-HDL-cholesterolemia (HDL cholesterol <40 mg/dL), hypertriglyceridemia (triglycerides ≥200 mg/dL), or taking a lipid-lowering drug.²² The individual components of dyslipidemia (hypercholesterolemia, hyper-LDL-cholesterolemia, hypo-HDL-cholesterolemia, and hypertriglyceridemia) were the secondary outcomes. Blood tests were performed during corporal health screening after fasting overnight and included measurements of total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides using a TBA-2000 FR Chemistry Analyzer (Toshiba).

Statistical analysis

The data are presented as the mean±standard deviation (SD) for continuous variables, number with proportion for categorical variables, and geometric mean (95% confidence interval [CI]) for non-normally distributed variables. To compare the baseline characteristics of the study participants according to weight change and BMI category, the chi-square test and one-way analysis of variance (ANOVA) were used for categorical and continuous variables, respectively. Multivariable logistic regression analysis was performed to calculate the odds ratio (OR) and 95% CIs of weight change associated with dyslipidemia risk, adjusting for age, baseline BMI, smoking, regular exercise, and hypertension. The “<5% change” group was used as a reference to estimate the multivariable-adjusted OR and 95% CI for the incidence of dyslipidemia for each weight change group. Subgroup analyses were performed to assess the effect modification on the risk of dyslipidemia based on baseline BMI, smoking status, regular exercise habits, and hypertension status. Interaction terms were added to test for effect modification across subgroups. All data analyses were conducted using IBM SPSS Statistics version 25 (IBM Co.). P<0.05 was considered statistically significant.

Participant selection

The health screening data of 34,173 soldiers from four medical battalions and one armed forces hospital were collected from May 2020 to April 2022 through the Armed Forces Medical Command Defense Medical Information System database (Fig. 1). A total of 15,183 soldiers who received only private or corporal health screening and 890 individuals who underwent two health screenings within a 90-day period were excluded. Additionally, 3,032 individuals who had missing data on anthropometric measurements, health questionnaires, or blood tests were excluded. No individuals were found to be taking medications for dyslipidemia; thus, the data from 15,068 soldiers were analyzed. The data of the study population showed a mean±SD interval of 250.1±84.4 days and a median (interquartile range [IQR]) interval of 237 days (IQR, 186 to 308) between private and corporal health screenings. A total of 1,379 participants (9.2%) were identified as having dyslipidemia, including 474 individuals (3.1%) with hypercholesterolemia, 434 individuals (2.9%) with hyper-LDL-cholesterolemia, 526 individuals (3.5%) with hypo-HDL-cholesterolemia, and 403 individuals (2.7%) with hypertriglyceridemia.

Baseline characteristics according to the status of weight changes

Baseline characteristics according to weight change are summarized in Table 1. In comparison with weight loss, weight gain was significantly associated with a lower baseline body weight, lower proportion of individuals with a BMI of 23 kg/m² or higher, lower proportion of current smokers, and higher proportion of hypertension.

Weight changes and risk of dyslipidemia

Data on the association between weight change and dyslipidemia and its components are summarized in Table 2. In comparison with no weight change, weight gain increased the risk of dyslipidemia (adjusted OR, 1.44 [95% CI, 1.22 to 1.69] for the 5% to 10% weight gain group; and 2.08 [95% CI, 1.67 to 2.59] for the ≥10% weight gain group), whereas weight loss decreased the risk (adjusted OR, 0.82 [95% CI, 0.69 to 0.97] for the 5% to 10% weight loss group; and 0.41 [95% CI, 0.30 to 0.55] for the ≥10% weight loss group) after adjusting for relevant covariates (model 3). These findings were consistent across the secondary outcomes. Weight gain increased the risk of hypercholesterolemia (adjusted OR, 1.63 [95% CI, 1.26 to 2.07] for the 5% to 10% weight gain group; and 2.16 [95% CI, 1.54 to 2.97] for the ≥10% weight gain group), whereas weight loss decreased the risk (adjusted OR, 0.52 [95% CI, 0.30 to 0.85] for the ≥10% weight loss group). Similar patterns were observed for hyper-LDL-cholesterolemia (adjusted OR, 1.84 [95% CI, 1.26 to 2.61] for the ≥10% weight gain group; and 0.43 [95% CI, 0.23 to 0.71] for ≥ the 10% weight loss group), hypo-HDL-cholesterolemia (adjusted OR, 1.51 [95% CI, 1.01 to 2.19] for the ≥10% weight gain group; and 0.43 [95% CI, 0.23 to 0.71] for the ≥10% weight loss group), and hypertriglyceridemia (adjusted OR, 1.71 [95% CI, 1.30 to 2.22] for the 5% to 10% weight gain group; 2.57 [95% CI, 1.78 to 3.63] for the ≥10% weight gain group; 0.36 [95% CI, 0.24 to 0.53] for the 5% to 10% weight loss group; and 0.24 [95% CI, 0.12 to 0.43] for the ≥10% weight loss group).

Subgroup analysis

Subgroup analysis was conducted on the primary outcome by stratifying the participants based on baseline BMI values of 25 and 23 kg/m², smoking status, regular exercise habits, and hypertension status (Fig. 2, Supplementary Table 1). No significant differences were observed between the subgroups, as indicated by P-values for interaction of >0.05.

In this study, weight change was associated with the incidence of dyslipidemia in young male soldiers. Weight gain was associated with a higher dyslipidemia incidence, whereas weight loss was associated with a lower dyslipidemia incidence. No significant differences were observed between the subgroups.

Previous studies conducted on middle-aged individuals have shown that changes in body weight may be associated with changes in lipid levels in both Western^7,9,11 and Asian populations.^6,8 Among studies conducted on Asian populations, a study involving Japanese adults aged 18 to 59 years found that a weight gain of above 2 kg in a 4-year period was associated with a two-fold increase in the risk of hypercholesterolemia.⁸ Another study involving Japanese males aged 40 to 50 years found that in a 10-year period, those who gained 5% to 15% of their body weight had a two-fold higher chance of having dyslipidemia when compared with those who did not gain any weight, and those who gained more than 15% of their body weight had a 2.6 times higher risk.⁶ In addition, a Korean study with individuals aged 19 to 64 years found that gaining weight over a year was associated with worsening total cholesterol, LDL cholesterol, HDL cholesterol, and triglyceride levels.¹² However, this study differs from prior studies in certain aspects. First, the study population in this study was limited to young adults. Second, body weight was physically measured in this study in contrast to previous studies that relied on self-reported body weight.

Dyslipidemia increases the risk of CVD even in those who are not obese.^23,24 However, the association between weight change and dyslipidemia in non-obese young adults has not been well-studied. In a study involving non-obese Japanese males aged 40 to 50 years, approximately 1.4-fold higher risk of dyslipidemia was observed in the group with a weight gain of more than 5% compared to the group with less than 5% weight change.⁶ Similarly, our study demonstrated an association between weight gain and dyslipidemia in young individuals who were not obese. These findings suggest that it is crucial to maintain healthy lifestyle behaviors to prevent weight gain even in non-obese individuals, considering that weight gain could have a negative effect on the incidence of dyslipidemia in this subpopulation.

In this study, we investigated the effect of weight change on four lipid profile components. Among the four components, triglycerides were the most sensitive lipid marker to weight change. Our findings are in agreement with those of previous studies demonstrating that triglycerides were more responsive than other lipid components to weight changes.^6,25,26 The findings may be explained by weight gain-induced increase in the flux of free fatty acids into the liver, which causes hepatic triglyceride accumulation and the subsequent synthesis of very LDLs, the predominant type of lipoprotein that transports triglycerides in the blood.²⁷

Approximately 26.6% of Korean males in their 20s have been reported to have dyslipidemia.²² In contrast to the findings for the general population, the prevalence of dyslipidemia in this study was substantially lower at 9.2%, which is consistent with the results of recent studies showing that metabolic diseases are less common in Korean soldiers than in males in their 20s in the general population.²⁸ This observation may be explained by differences in the lifestyle of soldiers compared with that of the general population. Specifically, this study found that 59.9% of the participants exercised regularly, which is known to be good for dyslipidemia prevention.²⁹ However, 51.7% of Korean males do not engage in any physical activity.³⁰ Furthermore, alcohol intake is well known as a risk factor for hypertriglyceridemia,³¹ which is restricted while serving in the military. In contrast, the average monthly alcohol intake rate among Korean males was 70.2% in 2020.30 Therefore, a healthy lifestyle may have reduced the likelihood of dyslipidemia among study participants.

This study has several limitations. First, it exclusively focused on individuals who were under mandatory military service, which, in South Korea, applies only to males. As a result, the study’s participant pool was limited to males, and its findings may not be fully representative of the general population. Second, the baseline (private health screening) lipid profile was not assessed. Consequently, self-reported data were used to determine the presence of dyslipidemia at baseline. Third, the dietary habits of the participants, which affect dyslipidemia, were not assessed.¹⁸ Lastly, the health questionnaire only evaluated the frequency of exercise that induced sweating and did not collect information on exercise intensity or duration, which may have limited the accuracy of the data on physical activity levels.

In conclusion, we found that the incidence of dyslipidemia was associated with weight changes in young military personnel. Through direct measurements of body weight and analyses of dyslipidemia components, our study identified the association between weight change and lipid profile alteration. Notably, weight gain increased the risk of dyslipidemia even in non-obese individuals. Our findings suggest that limiting weight gain in young adults by encouraging a healthy lifestyle may help prevent dyslipidemia. However, future prospective studies are needed to validate our findings.

Supplementary materials can be found online at https://doi.org/10.7570/jomes23056.

The authors declare no conflict of interest.

This work was supported by a grant from the National Research Foundation (NRF) of Korea (NRF-2022R1C1C1006021) to Wonsuk Choi.

Study concept and design: JYY, CKP, and WC; acquisition of data: JYY, CKP, and WC; analysis and interpretation of data: JYY, WJP, CKP, and WC; drafting of the manuscript: JYY, CKP, and WC; critical revision of the manuscript: JYY, WJP, CKP, and WC; statistical analysis: JYY, WJP, CKP, and WC; obtained funding: CKP and WC; administrative, technical, or material support: JYY, WJP, HKK, HCK, CKP, and WC; and study supervision: JYY, WJP, HKK, HCK, CKP, and WC.