Serum 25-hydroxyvitamin D levels and risk of type 2 diabetes according to glycemic status: a prospective cohort study

Nutrition & Diabetes ARTICLE www.nature.com/nutd OPEN Serum 25-hydroxyvitamin D levels and risk of type 2 diabetes according to glycemic status: a prospective cohort study Sihan Song 1 , Mi Kyoung Son1, Bo Mi Song1, Joong-Yeon Lim1 and Hyun-Young Park 2✉ 1234567890();,: © The Author(s) 2026 BACKGROUND: Evidence on whether the association between vitamin D levels and diabetes risk differs across glycemic statuses is limited. We investigated the association between circulating vitamin D levels and type 2 diabetes risk according to glycemic status. METHODS: This prospective cohort study included 3604 participants without diabetes (mean age: 56 years) whose serum 25hydroxyvitamin D (25(OH)D) levels were measured in the community-based Korean Genome Epidemiology Study (2007–2009). Fasting and 2-h post-load glucose and glycated hemoglobin (HbA1c) levels were measured biannually over a 14-year follow-up period. Season-standardized serum 25(OH)D levels were categorized into <25 (deficient), 25–50 (insufficient), and ≥ 50 nmol/L (sufficient). We assessed the association between serum 25(OH)D levels and type 2 diabetes risk using Cox proportional hazards models with time-varying covariates, stratifying by glycemic status and individual glycemic markers. RESULTS: Over 43,176 person-years of follow-up, 796 participants developed type 2 diabetes. Compared with participants with vitamin D deficiency, the multivariable-adjusted hazard ratios (95% confidence intervals) for type 2 diabetes in those with vitamin D sufficiency were 0.80 (0.62–1.04) overall, 0.55 (0.32–0.97) in participants with normoglycemia, and 0.93 (0.69–1.24) in those with prediabetes (P for interaction by glycemic status = 0.04). The association between serum 25(OH)D levels and diabetes risk differed according to HbA1c level (P for interaction = 0.05); a trend toward a lower risk of type 2 diabetes was observed in participants with HbA1c < 5.7% but not in those with HbA1c 5.7–6.4%. CONCLUSIONS: The association between serum 25(OH)D levels and the risk of type 2 diabetes differed by glycemic status, with a more pronounced inverse association among participants with normoglycemia. Nutrition and Diabetes (2026)16:8 ; https://doi.org/10.1038/s41387-026-00416-y INTRODUCTION Diabetes is a leading public health issue worldwide, affecting nearly 529 million people in 2021, with projections indicating more than 1.31 billion cases by 2050 [1]. Diabetes was ranked as the eighth leading cause of death and disability combined in 2019 [2]. Type 2 diabetes, which accounts for over 90% of all diabetes cases, is primarily characterized by insulin resistance and relative insulin deficiency due to pancreatic β-cell dysfunction [1, 3]. Evidence from epidemiological studies suggests that type 2 diabetes is, to some extent, preventable, and that early management can lead to remission [4, 5]. Therefore, identifying modifiable risk factors is crucial to reduce the burden of diabetes. Vitamin D is primarily synthesized in the skin through sunlight exposure and is obtained to a lesser extent from dietary sources [6]. It plays a key role in bone and calcium metabolism and has also been implicated in the pathogenesis of diabetes [7]. Preclinical studies suggest that vitamin D deficiency may impair insulin secretion and glucose tolerance [8, 9]. Observational studies have reported an association between low vitamin D levels and higher risk of type 2 diabetes [10]. Although clinical evidence remains inconclusive, the 2024 Endocrine Society Clinical Practice Guideline states that vitamin D supplementation may be considered for adults with high-risk prediabetes to help reduce the risk of developing type 2 diabetes [11]. In this context, it is important to provide further evidence on the role of vitamin D in type 2 diabetes across different populations to develop effective risk management strategies. However, evidence on whether this relationship differs by glycemic status is limited. Furthermore, longitudinal data on the association between circulating vitamin D levels and type 2 diabetes in the Korean population, where the prevalence of type 2 diabetes among adults aged ≥30 years exceeded 15% in 2020 [12], remains scarce. In a previous Korean study of high-risk individuals for type 2 diabetes, vitamin D deficiency was associated with a higher risk of type 2 diabetes, compared with vitamin D sufficiency [13]. Therefore, in this prospective cohort study, we aimed to investigate the association between 25-hydroxyvitamin D (25[OH] D), the main form of circulating vitamin D, and the risk of type 2 diabetes among Korean adults. We examined whether this association varied according to the glycemic status at baseline. METHODS Study population The Ansan–Ansung cohort is an ongoing prospective study conducted as part of the Korean Genome Epidemiology Study (KoGES) by the National 1 Division of Population Health Research, Department of Precision Medicine, National Institute of Health, Cheongju 28159, Republic of Korea. 2National Institute of Health, Cheongju 28159, Republic of Korea. ✉email: Received: 15 January 2025 Revised: 6 February 2026 Accepted: 13 March 2026 S. Song et al. 2 Fig. 1 Flowchart of the study population. KoGES Korean Genome Epidemiology Study, 25(OH)D 25-hydroxyvitamin D, eGFR estimated glomerular filtration rate. Institute of Health at the Korea Disease Control and Prevention Agency (KDCA) [14]. This community-based study enrolled 10,030 adults aged 40–69 years from two cities in Gyeonggi Province, Ansan (an urban industrial area) and Ansung (a predominantly rural area), between June 2001 and January 2003. Participants were followed up biannually through interviewer-administered questionnaires, anthropometric measurements, and biochemical assessments, including blood and urine tests. In each survey, blood tests for diabetes were conducted, and fasting blood samples were collected and stored for future use. As of December 2022, the tenth follow-up survey was completed. The National Institute of Health conducted additional biochemical analyses between 2017 and 2018, using frozen serum samples obtained from the third follow-up survey (2007–2009). Circulating vitamin D levels were measured in 5578 of 6688 participants from that survey (Fig. 1). From these, we excluded 1383 participants based on the following criteria: a history of physician-diagnosed diabetes (n = 595), current or past treatment for diabetes (n = 542), undiagnosed diabetes according to the American Diabetes Association (ADA) criteria at any point from study entry to the third follow-up (n = 638), or missing data on relevant variables (n = 68). Additionally, we excluded 507 participants owing to cancer confirmed through registry data (n = 117); a history of physiciandiagnosed cardiovascular disease (n = 182) or renal disease (n = 157); chronic kidney disease, defined as an estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m² based on the 2021 Chronic Kidney D (...truncated)