Serum vitamin D level is inversely associated with non–high-density lipoprotein/high-density lipoprotein ratio: a cross-sectional study utilizing Korea National Health and Nutrition Examination Survey 2022–2023 data
This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Recent evidence has underscored the crucial metabolic role of vitamin D. An increased non–high-density lipoprotein (non-HDL)/HDL ratio has been identified as a meaningful marker of metabolic risk. This study aimed to examine the relationship between serum 25-hydroxyvitamin D (25(OH)D) concentration and non-HDL/HDL ratio in a cohort of Korean adults.
Methods
For analysis, 10,864 participants aged 19 years or older were selected from the 2022–2023 Korea National Health and Nutrition Examination Survey (KNHANES). Serum 25(OH)D concentrations were classified into three groups: deficient (<20 ng/mL), insufficient (20–30 ng/mL), and sufficient (≥30 ng/mL). Non-HDL/HDL ratio was derived from total cholesterol and HDL cholesterol values. One-way analysis of variance and separate multiple regression analyses were conducted to assess differences in metabolic parameters between the 25(OH)D groups in both the general and diabetes subgroups (glycated hemoglobin ≥6.5%).
Results
Triglycerides, low-density lipoprotein cholesterol, and non-HDL/HDL ratio values were significantly lower in the group with sufficient vitamin D levels. Serum 25(OH)D and non-HDL/HDL ratio showed an inverse correlation in both general (r=–0.09) and diabetes (r=–0.15) subgroups. Across all regression models, a statistically significant inverse association persisted even after comprehensive adjustments for potential confounders.
Conclusion
A significant inverse correlation exists between serum 25(OH)D concentration and non-HDL/HDL ratio in Korean adults, with a more pronounced effect observed in individuals with diabetes.
Vitamin D (VD) is a fat-soluble secosteroid essential for maintaining bone integrity, regulating immune responses, and supporting cellular processes [1,2]. The primary sources of VD are cutaneous synthesis following UVB exposure and dietary intake. In addition to bone health, VD is associated with cardiovascular diseases (CVDs), metabolic disorders, and immune modulation [3-5]. VD deficiency is a widespread public health problem affecting nearly 50% of the global population, with approximately one billion individuals classified as VD-deficient [6]. In Korea, a notable segment of the population, particularly older adults, demonstrates VD insufficiency or deficiency, emphasizing its importance as a regional health concern [7]. Low VD concentrations are linked to a range of pathological conditions, including diabetes, cancer, autoimmune disorders, and neurodegenerative diseases [8]. However, the mechanisms underlying VD metabolism and its interactions with lipid metabolism remain unclear.
Cholesterol functions as a fundamental component of cellular membranes, bile acids, and steroid hormones; however, disturbances in its regulation are major contributors to atherosclerosis and CVDs. Approximately 80% of serum cholesterol is synthesized within the body, mainly by the liver, whereas the remainder is acquired through diet. Cholesterol balance is achieved through coordinated biosynthesis, absorption, cellular uptake, efflux, and excretion. Disruptions to this equilibrium elevate cardiovascular risk. Advancing age is linked to impaired cholesterol regulation, typified by increased total cholesterol and low-density lipoprotein (LDL) levels, which increase the prevalence of CVD [9,10]. These metabolic changes are also correlated with altered VD metabolism, although detailed mechanisms are not yet fully understood [11].
VD deficiency may contribute to dyslipidemia, particularly through alterations in non-high-density lipoprotein (non-HDL) and HDL [12]. LDL-based assessments may overlook atherosclerotic risk, as recent studies have identified the total burden of apolipoprotein B (ApoB)-containing lipoproteins, including very-low-density lipoprotein, intermediate-density lipoprotein, remnant lipoproteins, and lipoprotein(a), as powerful determinants of atherosclerosis and residual cardiovascular risk [13]. Even when guideline-recommended LDL targets are achieved, elevated ApoB-containing particles may persist and contribute to a substantial residual risk [14], highlighting the limitations of LDL-based assessments. Among these indices, the non-HDL/HDL ratio integrates pro-atherogenic ApoB-containing lipoproteins relative to anti-atherogenic HDL, thus providing a more comprehensive indicator of cardiometabolic risk [15]. This ratio has demonstrated superior predictive performance for metabolic syndrome, insulin resistance, and cardiovascular outcomes, particularly in populations with metabolic abnormalities such as obesity, insulin resistance, and type 2 diabetes [16], in which small dense LDL and triglyceride-rich remnants are disproportionately elevated [17]. Despite its clinical relevance, the association between VD status and comprehensive lipid ratio has not been clearly established in population-based studies. Furthermore, whereas most previous national cohort studies relied on immunoassay-based measurements of 25-hydroxyvitamin D (25(OH)D), the recent Korea National Health and Nutrition Examination Survey (KNHANES) 2022–2023 dataset adopted liquid chromatography–tandem mass spectrometry (LC-MS/MS), an internationally recommended reference method, to improve the biochemical accuracy and global comparability of VD assessment. Using this standardized LC-MS/MS dataset, our study provides the first large-scale, population-based evaluation of the association between VD and the non-HDL/HDL cholesterol ratio in Korea, offering updated evidence that better aligns with current clinical and epidemiological standards.
Methods
Data collection
This study analyzed data from the 2022–2023 KNHANES, a nationwide cross-sectional survey conducted annually by the Korea Disease Control and Prevention Agency to evaluate the health and nutritional status of the Korean population. The KNHANES implements a complex, stratified, and multistage probability sampling design to generate an independent sample each year, ensuring representation of the Korean population during the survey period. Serum 25(OH)D concentrations were assessed using liquid chromatography-tandem mass spectrometry, with standardization initiated in the 9th cycle (2022–2023) to enhance measurement accuracy and inter-study comparability. Of the 13,194 individuals surveyed during 2022–2023, adults aged 19 years and older were eligible for inclusion. Participants lacking data on age, sex, serum 25(OH)D, total cholesterol, or HDL were excluded, resulting in a final analytical sample of 10,864 participants (Figure 1). The 2022–2023 KNHANES protocol requires all participants to provide written informed consent before enrollment. This study adhered to the ethical standards outlined in the Declaration of Helsinki. As the KNHANES datasets were publicly accessible, anonymized, and de-identified, this secondary data analysis did not require Institutional Review Board (IRB) approval (IRB no., AJOUIRB-EX-2025-106).
Definitions
Serum total 25(OH)D concentration, determined as the combined value of 25(OH)D2 and 25(OH)D3 along with serum uric acid levels, was extracted from the KNHANES dataset. The demographic variables included age (years) and sex (male or female). Anthropometric indicators included body mass index (BMI; kg/m²), waist circumference (WC; cm), and blood pressure (systolic [mm Hg] and diastolic [mm Hg]). Medical history included the presence of hypertension, diabetes, or dyslipidemia, evaluated using responses to standardized questionnaires. Laboratory assessments performed during fasting measured fasting glucose, glycated hemoglobin (HbA1c), and lipid profiles (total cholesterol, triglycerides, LDL, and HDL). Serum 25(OH)D status was classified in accordance with the Endocrine Society clinical practice guidelines as deficient (<20 ng/mL), insufficient (20–30 ng/mL), or sufficient (≥30 ng/mL) [10].
Statistical analysis
One-way analysis of variance was performed to compare the differences in the non-HDL/HDL ratio across VD categories. When statistically significant differences were identified, Tukey’s honest significant difference test was applied for post hoc pairwise comparisons. Pearson’s correlation coefficients (r) were calculated and regression lines with 95% confidence intervals were generated to visualize linear associations.
Given the markedly elevated cardiometabolic risk in diabetes and previous reports suggesting stronger VD–lipid interactions under impaired glucose homeostasis, subgroup analyses stratified by diabetes status were performed to determine whether the association between serum 25(OH)D and non-HDL/HDL ratio differed according to metabolic condition.
Multiple linear regression analyses were subsequently conducted in the general population and diabetes subgroup to evaluate the independent association between serum 25(OH)D concentration and non-HDL/HDL ratio. Model 1 was adjusted for age and sex; model 2 further included BMI; and model 3 additionally incorporated WC, HbA1c, chronic diseases (hypertension, diabetes, and dyslipidemia), smoking status, alcohol consumption, and regular aerobic physical activity. Hypertension, diabetes, and dyslipidemia were defined based on physician diagnoses reported in the KNHANES questionnaire. Smoking status, alcohol intake, and aerobic physical activity were derived from the lifestyle questionnaire, with aerobic activity categorized per WHO guidelines.
The results of each model are summarized in the tables for both groups. Boxplots were generated to illustrate the distribution of the non-HDL/HDL ratio across the VD categories in each group, highlighting the median, interquartile range, and potential outliers. All statistical analyses were performed using the R software ver. 4.2.2 (The R Foundation for Statistical Computing).
Results
VD status was defined as deficient (<20 ng/mL), insufficient (20–30 ng/mL), or sufficient (≥30 ng/mL), and metabolic parameters were compared between the groups (Table 1). Age differed significantly between the groups (P<0.001), with the sufficient group having the highest mean age. BMI and WC were significantly lower in the sufficient group (both P<0.001). While fasting glucose and total cholesterol levels did not vary significantly across the VD categories, a higher VD status was consistently associated with higher HDL concentrations (P<0.001). Triglycerides, LDL, and non-HDL/HDL ratio also progressively improved across VD categories, with the lowest levels observed in the sufficient group (Figure 2A).
In the diabetes subgroup, defined according to the American Diabetes Association criteria (HbA1c ≥6.5%, fasting glucose ≥126 mg/dL, or diagnosed diabetes/antidiabetic medication use), a total of 1,672 individuals were included (Table 2). Similar patterns were observed, with significant differences in BMI, WC, HbA1c, HDL, and lipid profiles across the VD categories. Notably, the inverse association between VD levels and non-HDL/HDL ratio was more pronounced in this subgroup than in the general population (Figure 2B).
Scatterplot analyses indicated a negative correlation between serum 25(OH)D levels and non-HDL/HDL ratio in both the general population (r=–0.09, P<0.001) (Figure 3A) and the diabetes subgroup (r=–0.15, P<0.001) (Figure 3B). Importantly, this association was more robust in the diabetes subgroup.
Multivariate regression was performed using three sequentially adjusted models: Model 1 was adjusted for age and sex; model 2 was additionally adjusted for BMI; and model 3 was further adjusted for WC, HbA1c, chronic disease status (hypertension, diabetes, and dyslipidemia), smoking, alcohol consumption, and regular aerobic physical activity. As shown in Tables 3 and 4, serum 25(OH)D concentration was consistently and inversely associated with non-HDL/HDL ratio across all sequentially adjusted models in both the general population and the diabetes subgroup (all P<0.001). In the general population (Table 3), the inverse association remained statistically significant from the minimally adjusted to the fully adjusted model, and the same pattern was observed in the diabetes subgroup (Table 4), indicating that the relationship between serum 25(OH)D concentration and non-HDL/HDL ratio was robust and independent of major demographic, anthropometric, metabolic, and lifestyle factors.
Multivariate regression was performed using three sequentially adjusted models: Model 1 was adjusted for age and sex; model 2 was additionally adjusted for BMI; and model 3 was further adjusted for WC, HbA1c, chronic disease status (hypertension, diabetes, and dyslipidemia), smoking, alcohol consumption, and regular aerobic physical activity. As shown in Tables 3 and 4, serum 25(OH)D concentration was consistently and inversely associated with non-HDL/HDL ratio across all sequentially adjusted models in both the general population and the diabetes subgroup (all P<0.001). In the general population (Table 3), the inverse association remained statistically significant from the minimally adjusted to the fully adjusted model, and the same pattern was observed in the diabetes subgroup (Table 4), indicating that the relationship between serum 25(OH)D concentration and non-HDL/HDL ratio was robust and independent of major demographic, anthropometric, metabolic, and lifestyle factors.
Discussion
In this nationally representative sample of Korean adults, lower serum 25(OH)D concentrations were significantly associated with higher non-HDL/HDL ratios, which remained robust after adjusting for multiple metabolic and lifestyle covariates. These results are consistent with previous population-based findings showing that the inverse relationship between VD and lipid status persists even when using a comprehensive atherogenic index that reflects the total ApoB-containing lipoprotein burden rather than LDL alone. This association was more pronounced in individuals with diabetes, indicating that impaired glucose metabolism may increase susceptibility to VD–related alterations in lipid profiles. The stronger association observed in the diabetic subgroup could be attributed to the greater metabolic vulnerability exhibited in diabetes, such as increased insulin resistance, persistent inflammation, and altered lipid homeostasis [18-20]. Recognizing the increased cardiovascular risk among patients with diabetes and examining the association between VD levels and lipid parameters, specifically the non-HDL/HDL ratio, is essential for improved risk assessment and clinical management. Similarly, a stronger inverse association between VD deficiency and atherogenic lipid markers in diabetic cohorts than in the general population has been previously reported [21]. Since many existing studies have predominantly addressed the general population and relatively few have focused on the distinct metabolic context of diabetes, our subgroup analysis enhances our understanding of the potential impact of VD on cardiovascular risk factors in patients with diabetes.
VD is widely recognized for its essential contribution to bone health; however, its impact extends further to metabolic regulation, immune function, and CVD prevention [22]. Reduced levels of 25(OH)D have been linked to unfavorable metabolic parameters such as increased triglycerides, lower HDL, higher LDL, and insulin resistance [8,11]. Our study builds on these findings by showing that VD deficiency is associated with an elevated non-HDL/HDL ratio—a comprehensive indicator of atherogenic lipid burden [9]. This specific lipid ratio, which accounts for all ApoB-containing lipoproteins, has been identified as more effective than LDL alone in predicting cardiovascular events, particularly in individuals with metabolic conditions. Notably, the observed inverse relationship between serum 25(OH)D concentration and non-HDL/HDL ratio was more pronounced in the diabetes subgroup (r=–0.15 vs. –0.09, respectively, in the overall population), corroborating previous results reported by Mohamad et al. [21] in patients with type 2 diabetes.
The stronger correlation observed in individuals with diabetes may be attributed to the combined effects of diminished insulin sensitivity, increased inflammatory activity, and the disruption of lipid metabolism [23]. VD may help counteract these adverse processes through its anti-inflammatory actions, enhancement of insulin signaling, and support of hepatic lipid regulation [23]. As a fat-soluble secosteroid, VD undergoes hepatic conversion to 25(OH)D, followed by renal conversion to its bioactive form 1,25-dihydroxyvitamin D (1,25(OH)D). The extensive presence of VD receptors in multiple tissues suggests the presence of several molecular pathways through which VD affects lipid homeostasis [24]. Evidence from mechanistic studies has demonstrated that VD modulates lipid metabolism by inhibiting sterol regulatory element-binding proteins (SREBPs) via accelerated degradation of SREBP cleavage-activating proteins, thereby decreasing fatty acid and triglyceride production [20]. Moreover, VD may influence lipid metabolism by regulating pro-inflammatory cytokines such as tumor necrosis factor-α and interleukin-6, which are commonly elevated in states of insulin resistance and dyslipidemia [16]. VD may also affect lipoprotein lipase activity and calcium-dependent mechanisms, collectively supporting the observed inverse association between serum VD concentration and atherogenic lipid parameters, such as the non-HDL/HDL ratio [25].
Epidemiological evidence has consistently shown that elevated serum 25(OH)D concentrations are associated with more favorable lipid profiles, characterized by lower levels of triglycerides, total cholesterol, and LDL, as well as improved lipid ratios [26]. Meta-analyses involving pediatric and adolescent populations have indicated that a higher VD status is associated with a 27% reduction in the risk of hypertriglyceridemia and a 22% lower risk of low HDL levels [27]. These results reinforce the association between VD insufficiency and dyslipidemia across various demographic groups and age ranges, with the relationship potentially influenced by sex and hormonal factors. Meta-analyses of randomized controlled trials also indicated that VD supplementation may beneficially impact lipid profiles by significantly reducing triglycerides and total cholesterol, while also producing modest increases in HDL [28]. Among individuals with type 2 diabetes, supplementation resulted in reductions in total cholesterol and LDL, with trends suggesting decreased triglycerides, but with only a minor effect on HDL [21]. Notably, these interventional data were consistent with the inverse association found in the current KNHANES 2022–2023 dataset between serum VD levels and non-HDL/HDL ratio, underscoring the role of VD status in the regulation of lipid metabolism. Nevertheless, due to the cross-sectional design and possible residual confounding factors, further longitudinal and interventional investigations are necessary to establish causality.
The strengths of our study include the use of a large nationally representative dataset (KNHANES), comprehensive subgroup analyses based on diabetes status, and rigorous adjustments for confounding variables across the three regression models. Nonetheless, this study has some limitations that must be acknowledged. The cross-sectional nature of this study prevents definitive conclusions regarding causality. The VD status was assessed using serum 25(OH)D, which is an inactive precursor rather than the metabolically active form of 1,25(OH)D. Although 25(OH) D is widely accepted as a standard clinical indicator, it does not entirely represent functional hormonal activity. For example, the bioavailability of circulating 25(OH)D is affected by VD-binding proteins. Additionally, information on seasonality, sunlight exposure, dietary factors, supplement intake, and specific diabetes medications was unavailable, which may have contributed to unmeasured confounding factors. For instance, differences in seasonal sun exposure have a significant impact on the VD status, and the absence of this information may limit the interpretation of associations between VD and lipid profiles [29]. Despite these limitations, our findings suggest that lower serum VD levels are linked to a more atherogenic lipid profile, particularly in diabetic populations. This finding supports the potential benefits of routine VD monitoring and correction in the cardiovascular risk management of metabolic disorders. Further prospective cohort and randomized intervention studies are necessary to determine whether VD supplementation can effectively reduce non-HDL/HDL ratio and improve cardiovascular outcomes.
Notes
Conflict of interest
No potential conflict of interest relevant to this article was reported.
Funding
This research was supported by a grant of the MD-Phd/Medical Scientist Training Program through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea.
Data availability
The data presented in this study are available from the Korea National Health and Nutrition Examination Survey (KNHANES) database at https://knhanes.kdca.go.kr.
Author contribution
Conceptualization: NSJ. Data curation: HRL, YC, DEC. Formal analysis: HRL. Funding acquisition: NSJ. Investigation: HRL, YC. Methodology: NSJ. Project administration: NSJ. Visualization: HRL. Writing–original draft: YC. Writing–review & editing: NSJ. Final approval of the manuscript: all authors.
Figure. 1.
Flow chart of participant selection. KNHANES, Korea National Health and Nutrition Examination Survey; 25(OH)D, 25-hydroxyvitamin D.
Figure. 2.
Distribution of non–high-density lipoprotein (HDL)/HDL ratio across serum 25(OH)D, 25-hydroxyvitamin D (25(OH)D) categories in the general population (A) and in the diabetes subgroup (B). Vitamin D status was categorized as deficient (<20 ng/mL), insufficient (20–30 ng/mL), and sufficient (≥30 ng/mL). Boxes represent the interquartile range (IQR), the horizontal line inside each box indicates the median, and whiskers extend to 1.5×IQR. Overall group differences were assessed using the Kruskal-Wallis test, and pairwise comparisons were conducted using Wilcoxon rank-sum tests with Benjamini-Hochberg correction for multiple testing. ***P<0.001 (statistical significance).
Figure. 3.
Relationship between serum 25-hydroxyvitamin D (25(OH)D) concentration and non-high-density lipoprotein (HDL)/HDL ratio in the general population (A) and the diabetes subgroup (B). Scatterplots show serum 25(OH)D concentration (ng/mL) versus the non-HDL/HDL ratio for each group. Panel (A) illustrates the general population (r=–0.09) with the regression line (red) and 95% confidence interval (shaded area). Panel (B) illustrates the diabetes group (r=–0.15), including the regression line, 95% confidence interval, correlation coefficient (r), and P-value displayed on the plot.
Table 1.
General characteristics of the overall population categorized by vitamin D group
Values are presented as mean±standard deviations for continuous variables or number (%) for categorical variables. Group comparisons utilized the chi-square test for categorical data and one-way analysis of variance for continuous data.
BMI, body mass index; WC, waist circumference; HbA1c, glycated hemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein.
a)Non-HDL/HDL ratio=total cholesterol–HDL/HDL;
b)25(OH)D (ng/mL), the total serum concentration of 25-hydroxyvitamin D, representing the combined values of 25(OH)D2 and 25(OH)D3.
Table 2.
General characteristics of the diabetes subgroup by vitamin D group
Values are presented as mean±standard deviations for continuous variables or number (%) for categorical variables. Group comparisons were performed using the chi-square test for categorical variables and one-way ANOVA for continuous variables.
BMI, body mass index; WC, waist circumference; HbA1c, glycated hemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein.
a)Non-HDL/HDL ratio=total cholesterol–HDL/HDL;
b)25(OH)D (ng/mL), the total serum concentration of 25-hydroxyvitamin D, calculated as the sum of 25(OH)D2 and 25(OH)D3;
c)Diabetes subgroup was defined according to American Diabetes Association (ADA) criteria.
Table 3.
Multivariable linear regression assessing the association between serum 25(OH)D levels and the non-HDL/HDL ratio in the general population
Model
Estimate
P-value
R2
Model 1
–0.008
<0.001
0.062
Model 2
–0.005
<0.001
0.134
Model 3
–0.003
<0.001
0.236
Multiple regression analyses of the association between serum 25(OH)D levels and the non-HDL/HDL ratio, showing regression coefficients (β), P-values for 25(OH)D, and R2 values from three models with progressively adjusted covariates. Model 1: adjusted for age, sex; Model 2: adjusted for age, sex, BMI; and Model 3: adjusted for age, sex, BMI, WC, HbA1c, chronic diseases (hypertension, diabetes mellitus, dyslipidemia), smoking, drinking, aerobic physical activity.
25(OH)D, 25-hydroxyvitamin D; HDL, high-density lipoprotein; BMI, body mass index; WC, waist circumference; HbA1c, glycated hemoglobin.
Table 4.
Multivariable linear regression assessing the association between serum 25(OH)D levels and the non-HDL/HDL ratio in individuals with diabetes
Model
Estimate
P-value
R2
Model 1
–0.008
<0.001
0.063
Model 2
–0.005
<0.001
0.127
Model 3
–0.003
<0.001
0.236
Multiple regression analyses of the association between serum 25(OH)D levels and the non-HDL/HDL ratio in the diabetes group (individuals meeting American Diabetes Association criteria for diabetes). Regression coefficients (β), P-values for 25(OH)D, and R2 values are presented for three models with progressively adjusted covariates. Model 1: adjusted for age, sex; Model 2: adjusted for age, sex, BMI; and Model 3: adjusted for age, sex, BMI, WC, HbA1c, chronic diseases (hypertension, diabetes mellitus, dyslipidemia), smoking, drinking, aerobic physical activity.
25(OH)D, 25-hydroxyvitamin D; HDL, high-density lipoprotein; BMI, body mass index; WC, waist circumference; HbA1c, glycated hemoglobin.
References
1. Bikle DD. Vitamin D metabolism, mechanism of action, and clinical applications. Chem Biol 2014;21:319-29.
3. Wang TJ, Pencina MJ, Booth SL, Jacques PF, Ingelsson E, Lanier K, et al. Vitamin D deficiency and risk of cardiovascular disease. Circulation 2008;117:503-11.
4. Al-Daghri NM, Guerini FR, Al-Attas OS, Alokail MS, Alkharfy KM, Draz HM, et al. Vitamin D receptor gene polymorphisms are associated with obesity and inflammosome activity. PLoS One 2014;9:e102141.
8. Pittas AG, Lau J, Hu FB, Dawson-Hughes B. The role of vitamin D and calcium in type 2 diabetes: a systematic review and meta-analysis. J Clin Endocrinol Metab 2007;92:2017-29.
9. Paunica I, Mihai AD, Stefan S, Pantea-Stoian A, Serafinceanu C. Comparative evaluation of LDL-CT, non-HDL/HDL ratio, and ApoB/ApoA1 in assessing CHD risk among patients with type 2 diabetes mellitus. J Diabetes Complications 2023;37:108634.
10. Holick MF, Binkley NC, Bischoff-Ferrari HA, Gordon CM, Hanley DA, Heaney RP, et al. Evaluation, treatment, and prevention of vitamin D deficiency: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab 2011;96:1911-30.
11. Warren T, McAllister R, Morgan A, Rai TS, McGilligan V, Ennis M, et al. The interdependency and co-regulation of the vitamin D and cholesterol metabolism. Cells 2021;10:2007.
12. Li B, Wang J, Xu J, Xie J, Liu Q, Yang C, et al. Association between dyslipidemia and vitamin D deficiency: a cross-sectional study in Chinese healthy population. Front Endocrinol (Lausanne) 2025;16:1450924.
15. Raja V, Aguiar C, Alsayed N, Chibber YS, ElBadawi H, Ezhov M, et al. Non-HDL-cholesterol in dyslipidemia: review of the state-of-the-art literature and outlook. Atherosclerosis 2023;383:117312.
16. Kim SW, Jee JH, Kim HJ, Jin SM, Suh S, Bae JC, et al. Non-HDL-cholesterol/HDL-cholesterol is a better predictor of metabolic syndrome and insulin resistance than apolipoprotein B/apolipoprotein A1. Int J Cardiol 2013;168:2678-83.
19. Contreras-Bolivar V, Garcia-Fontana B, Garcia-Fontana C, Munoz-Torres M. Mechanisms involved in the relationship between vitamin D and insulin resistance: impact on clinical practice. Nutrients 2021;13:3491.
21. Mohamad MI, El-Sherbeny EE, Bekhet MM. The effect of vitamin D supplementation on glycemic control and lipid profile in patients with type 2 diabetes mellitus. J Am Coll Nutr 2016;35:399-404.
22. Liu D, Meng X, Tian Q, Cao W, Fan X, Wu L, et al. Vitamin D and multiple health outcomes: an umbrella review of observational studies, randomized controlled trials, and mendelian randomization studies. Adv Nutr 2022;13:1044-62.
23. Jorde R, Figenschau Y, Hutchinson M, Emaus N, Grimnes G. High serum 25-hydroxyvitamin D concentrations are associated with a favorable serum lipid profile. Eur J Clin Nutr 2010;64:1457-64.
24. Xu SM, Lu K, Yang XF, Ye YW, Xu MZ, Shi Q, et al. Association of 25-hydroxyvitamin D levels with lipid profiles in osteoporosis patients: a retrospective cross-sectional study. J Orthop Surg Res 2023;18:597.
25. Hajhashemy Z, Tirani SA, Askari G, Saneei P. The association between serum vitamin D levels and abnormal lipid profile in pediatrics: a GRADE-assessed systematic review and dose-response meta-analysis of epidemiologic studies. Nutr Rev 2025;83:e88-105.
26. Mirhosseini N, Rainsbury J, Kimball SM. Vitamin D supplementation, serum 25(OH)D concentrations and cardiovascular disease risk factors: a systematic review and meta-analysis. Front Cardiovasc Med 2018;5:87.
27. Lu Q, Liang Q, Xi Y. The effects of vitamin D supplementation on serum lipid profiles in people with type 2 diabetes: a systematic review and meta-analysis of randomized controlled trials. Front Nutr 2024;11:1419747.
28. Meng L, Shapses SA, Bikle DD, Wang X. Use of vitamin D binding protein and 25-hydroxyvitamin D levels in clinical practice for endocrinologists. Endocr Pract 2026;32:427-41.
29. Kim JH, Moon SJ. Time spent outdoors and seasonal variation in serum concentrations of 25-hydroxyvitamin D in Korean women. Int J Food Sci Nutr 2000;51:439-51.
Serum vitamin D level is inversely associated with non–high-density lipoprotein/high-density lipoprotein ratio: a cross-sectional study utilizing Korea National Health and Nutrition Examination Survey 2022–2023 data
Figure. 1. Flow chart of participant selection. KNHANES, Korea National Health and Nutrition Examination Survey; 25(OH)D, 25-hydroxyvitamin D.
Figure. 2. Distribution of non–high-density lipoprotein (HDL)/HDL ratio across serum 25(OH)D, 25-hydroxyvitamin D (25(OH)D) categories in the general population (A) and in the diabetes subgroup (B). Vitamin D status was categorized as deficient (<20 ng/mL), insufficient (20–30 ng/mL), and sufficient (≥30 ng/mL). Boxes represent the interquartile range (IQR), the horizontal line inside each box indicates the median, and whiskers extend to 1.5×IQR. Overall group differences were assessed using the Kruskal-Wallis test, and pairwise comparisons were conducted using Wilcoxon rank-sum tests with Benjamini-Hochberg correction for multiple testing. ***P<0.001 (statistical significance).
Figure. 3. Relationship between serum 25-hydroxyvitamin D (25(OH)D) concentration and non-high-density lipoprotein (HDL)/HDL ratio in the general population (A) and the diabetes subgroup (B). Scatterplots show serum 25(OH)D concentration (ng/mL) versus the non-HDL/HDL ratio for each group. Panel (A) illustrates the general population (r=–0.09) with the regression line (red) and 95% confidence interval (shaded area). Panel (B) illustrates the diabetes group (r=–0.15), including the regression line, 95% confidence interval, correlation coefficient (r), and P-value displayed on the plot.
Graphical abstract
Figure. 1.
Figure. 2.
Figure. 3.
Graphical abstract
Serum vitamin D level is inversely associated with non–high-density lipoprotein/high-density lipoprotein ratio: a cross-sectional study utilizing Korea National Health and Nutrition Examination Survey 2022–2023 data
Characteristic
Total
Deficient (<20 ng/mL)
Insufficient (20–30 ng/mL)
Sufficient (≥30 ng/mL)
P-value
No. of patients
10,864
4,227
3,651
2,986
Age (y)
53.55±16.87
48.06±17.64
54.38±15.95
60.35±13.95
<0.001
Sex, male
4,736 (43.6)
1,992 (47.1)
1,744 (47.8)
1,000 (33.5)
<0.001
BMI (kg/m2)
24.07±3.72
24.32±4.10
24.20±3.58
23.56±3.26
<0.001
WC (cm)
84.04±10.8
84.39±11.67
84.48±10.35
83.03±9.99
<0.001
Glucose (mg/dL)
101.21±23.08
101.41±25.64
101.27±22.39
100.86±19.92
0.598
HbA1c (%)
5.63±0.79
5.59±0.86
5.63±0.75
5.68±0.72
<0.001
Cholesterol (mg/dL)
186.4±40.47
186.94±39.74
186.51±39.98
185.43±42.02
0.287
HDL (mg/dL)
57.21±15.49
56.03±15.37
56.64±15.01
59.59±15.96
<0.001
Triglycerides (mg/dL)
126.76±96.18
132.33±102.54
127.09±102.30
118.47±76.44
<0.001
LDL (mg/dL)
113.02±36.81
114.06±36.30
113.65±36.68
110.69±37.53
<0.001
Non-HDL/HDL ratioa)
2.47±1.30
2.57±1.23
2.51±1.55
2.28±1.01
<0.001
25(OH)D (ng/mL)b)
24.45±11.34
14.00±3.77
24.87±2.86
38.76±8.90
<0.001
Hypertension
2,927 (27.0)
900 (21.3)
1,000 (27.4)
1,027 (34.4)
<0.001
Diabetes
1,278 (11.8)
436 (10.3)
400 (11.0)
442 (14.8)
<0.001
Dyslipidemia
2,518 (23.2)
710 (16.8)
835 (22.9)
973 (32.6)
<0.001
Characteristic
Total
Deficient (<20 ng/mL)
Insufficient (20–30 ng/mL)
Sufficient (≥30 ng/mL)
P-value
No. of patients
1,672
616
531
525
Age (y)
64.65±11.59
62.23±0.52
64.54±0.48
67.61±0.42
<0.001
Sex, male
887 (53.1)
353 (57.3)
322 (60.6)
212 (40.4)
<0.001
BMI (kg/m2)
25.26±3.80
25.81±0.17
25.48±0.15
24.39±0.15
<0.001
WC (cm)
89.98±9.97
91.34±0.43
90.39±0.41
87.97±0.43
<0.001
Glucose (mg/dL)
134.99±40.23
141.48±1.86
135.33±1.71
127.04±1.37
<0.001
HbA1c (%)
6.88±1.23
7.03±0.06
6.85±0.05
6.71±0.05
<0.001
Cholesterol (mg/dL)
162.12±42.63
166.05±1.82
163.30±1.84
156.31±1.70
<0.001
HDL (mg/dL)
50.74±13.54
49.47±0.53
50.54±0.57
52.44±0.62
<0.001
Triglycerides (mg/dL)
148.40±106.02
166.07±5.11
143.68±3.85
132.43±4.00
<0.001
LDL (mg/dL)
91.18±37.05
93.87±1.59
93.19±1.62
86.00±1.45
<0.001
Non-HDL/HDL ratioa)
2.37±1.17
2.55±0.05
2.41±0.05
2.13±0.04
<0.001
25(OH)D (ng/mL)b)
25.39±11.64
14.16±0.15
24.96±0.13
39.00±0.37
<0.001
Hypertension
955 (57.1)
329 (53.4)
300 (56.5)
326 (62.1)
0.0495
Diabetesc)
1,278 (76.4)
436 (70.8)
400 (75.3)
442 (84.2)
<0.001
Dyslipidemia
857 (51.3)
296 (48.1)
263 (49.5)
298 (56.8)
0.0075
Model
Estimate
P-value
R2
Model 1
–0.008
<0.001
0.062
Model 2
–0.005
<0.001
0.134
Model 3
–0.003
<0.001
0.236
Model
Estimate
P-value
R2
Model 1
–0.008
<0.001
0.063
Model 2
–0.005
<0.001
0.127
Model 3
–0.003
<0.001
0.236
Table 1. General characteristics of the overall population categorized by vitamin D group
Values are presented as mean±standard deviations for continuous variables or number (%) for categorical variables. Group comparisons utilized the chi-square test for categorical data and one-way analysis of variance for continuous data.
BMI, body mass index; WC, waist circumference; HbA1c, glycated hemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein.
Non-HDL/HDL ratio=total cholesterol–HDL/HDL;
25(OH)D (ng/mL), the total serum concentration of 25-hydroxyvitamin D, representing the combined values of 25(OH)D2 and 25(OH)D3.
Table 2. General characteristics of the diabetes subgroup by vitamin D group
Values are presented as mean±standard deviations for continuous variables or number (%) for categorical variables. Group comparisons were performed using the chi-square test for categorical variables and one-way ANOVA for continuous variables.
BMI, body mass index; WC, waist circumference; HbA1c, glycated hemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein.
Non-HDL/HDL ratio=total cholesterol–HDL/HDL;
25(OH)D (ng/mL), the total serum concentration of 25-hydroxyvitamin D, calculated as the sum of 25(OH)D2 and 25(OH)D3;
Diabetes subgroup was defined according to American Diabetes Association (ADA) criteria.
Table 3. Multivariable linear regression assessing the association between serum 25(OH)D levels and the non-HDL/HDL ratio in the general population
Multiple regression analyses of the association between serum 25(OH)D levels and the non-HDL/HDL ratio, showing regression coefficients (β), P-values for 25(OH)D, and R2 values from three models with progressively adjusted covariates. Model 1: adjusted for age, sex; Model 2: adjusted for age, sex, BMI; and Model 3: adjusted for age, sex, BMI, WC, HbA1c, chronic diseases (hypertension, diabetes mellitus, dyslipidemia), smoking, drinking, aerobic physical activity.
25(OH)D, 25-hydroxyvitamin D; HDL, high-density lipoprotein; BMI, body mass index; WC, waist circumference; HbA1c, glycated hemoglobin.
Table 4. Multivariable linear regression assessing the association between serum 25(OH)D levels and the non-HDL/HDL ratio in individuals with diabetes
Multiple regression analyses of the association between serum 25(OH)D levels and the non-HDL/HDL ratio in the diabetes group (individuals meeting American Diabetes Association criteria for diabetes). Regression coefficients (β), P-values for 25(OH)D, and R2 values are presented for three models with progressively adjusted covariates. Model 1: adjusted for age, sex; Model 2: adjusted for age, sex, BMI; and Model 3: adjusted for age, sex, BMI, WC, HbA1c, chronic diseases (hypertension, diabetes mellitus, dyslipidemia), smoking, drinking, aerobic physical activity.
25(OH)D, 25-hydroxyvitamin D; HDL, high-density lipoprotein; BMI, body mass index; WC, waist circumference; HbA1c, glycated hemoglobin.
