Vitamin D status and sarcopenia in older adults: population-based evidence for synergy with diabetes and experimental validation in an aged diabetic rat model Authors: Cheng Zhang, Mingkai Li, Chenyang Li, Lushan Xue, Chenhui Lv, Qinfei Guo, Yafei Zhao, Haifeng Zhao Published: 2026 ABSTRACT Graphical abstractKeywords: Vitamin D deficiency, Sarcopenia, Diabetes, Older adults Introduction Sarcopenia is a geriatric syndrome characterized by accelerated loss of skeletal muscle mass and function and is closely associated with multiple adverse health outcomes [1]. As global population aging accelerates, sarcopenia has emerged as a major public health challenge that undermines physical function and quality of life in later life. Globally, sarcopenia is estimated to affect approximately 16% of older adults, with a disproportionate burden among the oldest-old and among individuals with multimorbidity and suboptimal nutritional status [2,3]. Sarcopenia is associated with increased risks of falls, fractures, mobility limitation, and disability, and it is also linked to higher rates of hospitalization and mortality, thereby intensifying healthcare burden [4]. Because skeletal muscle is essential for locomotion and metabolic homeostasis [5], progressive declines in muscle mass and strength may exacerbate a self-reinforcing cycle of muscle loss and metabolic dysfunction by reducing basal metabolic rate, impairing glucose uptake and lipid oxidation, and promoting insulin resistance (IR) and chronic inflammation [6,7]. Sarcopenia is more prevalent among individuals with diabetes than in the general population [2]. Type 2 diabetes mellitus (T2DM) is frequently accompanied by IR, chronic low-grade inflammation, oxidative stress, and mitochondrial dysfunction. These pathophysiological disturbances may accelerate declines in muscle mass and strength by suppressing muscle protein synthesis and promoting proteolysis [8]. Accordingly, identifying modifiable risk factors for sarcopenia in the context of T2DM is critical to delaying functional deterioration and reducing disability risk in older adults. Vitamin D is essential for regulating calcium and phosphate homeostasis. As a fat-soluble micronutrient, it exerts its biological effects primarily by binding to the vitamin D receptor (VDR), thereby modulating downstream signaling and transcriptional programs [9]. Emerging evidence also implicates vitamin D in skeletal muscle biology, including myoblast proliferation and differentiation, as well as intramuscular fat accumulation [10,11]. Because skeletal muscle cells express VDR, vitamin D may help maintain muscle integrity by regulating calcium homeostasis, supporting myofiber differentiation and regeneration, and influencing intramuscular lipid content [12]. Epidemiological studies further indicate that vitamin D deficiency is highly prevalent among older adults [13]. Although many studies have reported associations between serum 25-hydroxyvitamin D [25(OH)D] concentrations and muscle mass, muscle strength, and sarcopenia risk, findings remain inconsistent across populations [14,15]. In addition, the contribution of diabetes to these associations and the underlying biological mechanisms are not fully understood. To address these limitations, we applied an integrated population evidence–experimental validation framework. First, we evaluated the association between 25(OH)D and sarcopenia risk and tested for interaction with diabetes in two cohorts of older adults. We then validated the contribution of vitamin D status to diabetes-associated acceleration of age-related muscle loss using a D-galactose–induced accelerated-aging rat model. Diabetes was induced via a high-sugar/high-fat diet combined with streptozotocin (STZ) injection, and we focused on key functional and structural phenotypes, including hindlimb grip strength and gastrocnemius muscle (GM) mass. We further investigated mechanisms by which vitamin D may ameliorate sarcopenia by modulating skeletal muscle fat infiltration. Collectively, these findings provide population-based evidence and mechanistic support for precision prevention and management of sarcopenia in older adults with T2DM, and identify potential intervention targets. Methods Epidemiological study Study population The Health and Retirement Study (HRS) was an ongoing, nationally representative longitudinal study of U.S. adults aged ≥50 years, funded by the National Institute on Aging (NIA; U01AG009740) and conducted by the University of Michigan. The English Longitudinal Study of Ageing (ELSA) was a prospective, nationally representative cohort of community-dwelling adults aged ≥50 years in England, supported by the NIA (R01AG017644). Further details on the study design, sampling framework, and data collection procedures for HRS and ELSA are available in prior publications. The HRS and ELSA protocols were approved by the University of Michigan Ethics Review Committee and the London Multicentre Research Ethics Committee, respectively, and all participants provided written informed consent. We used HRS wave 13 and ELSA wave 6 as the baseline waves. HRS wave 13 was selected because it was the first wave to include large-scale venous blood collection and measurement of serum 25-hydroxyvitamin D [25(OH)D] concentrations. In ELSA, nurse visits occur every four years to collect biological samples; wave 6 was selected because it contained complete venous blood collection records and serum 25(OH)D measurements. Supplementary Figure S1 presented the participant selection process. We initially included participants with blood test data (HRS: 9,934; ELSA: 8,054). After excluding individuals with missing serum 25(OH)D data or missing grip strength or Body mass index (BMI) measurements, 7,520 participants remained for analysis (HRS: 3,246; ELSA: 4,274). Assessment of sarcopenia Sarcopenia was defined according to the European Working Group on Sarcopenia in Older People 2 (EWGSOP2) criteria as the coexistence of low muscle strength and low skeletal muscle mass [16]. Muscle strength was assessed using handgrip strength, with low strength defined as <27 kg for men and <16 kg for women. Skeletal muscle mass (SMM) was estimated using the anthropometric equation validated by Lee et al. [17]: SMM = 0.244 × weight + 7.8 × height + 6.6 × sex – 0.098 × age + race – 3.3 (where female = 0 and male = 1; race = 0 [White and Hispanic], race = 1.4 [Black] and race = −1.2 [Asian]). Where sex was coded as male = 1 and female = 0, and race was coded as White/Hispanic = 0, Black = 1.4, and Asian = −1.2. SMM was standardized by body mass index (BMI) to derive the skeletal muscle mass index (SMI = SMM/BMI) [18]. Low skeletal muscle mass was defined as the sex-specific lowest quartile of SMI within the study population, consistent with prior studies [19]. Assessment of vitamin D Vitamin D status was assessed using serum 25-hydroxyvitamin D [25(OH)D] concentrations. In HRS, serum 25(OH)D was quantified by liquid chromatography–tandem mass spectrometry (LC–MS/MS), with 25(OH)D2 and 25(OH)D3 measured concurrently; total 25(OH)D was calculated as the sum of 25(OH)D2 and 25(OH)D3 and expressed in nmol/L. In ELSA, serum 25(OH)D was measured using the DiaSorin LIAISON immunoassay, which detects 25(OH)D2 and 25(OH)D3, and results were reported in nmol/L. Details on assay methods and quality control were provided in the official technical documentation. Low serum 25(OH)D was defined as ≤50 nmol/L, and higher serum 25(OH)D as >50 nmol/L [20]. Assessment of covariates Demographic covariates included age, sex, race/ethnicity, education, and marital status. Prior studies indicate that physical activity (PA) was important determinants of sarcopenia [21]. To ensure comparability between HRS and ELSA, we harmonized the PA measures using questionnaire-reported frequency of vigorous, moderate, and light activities. For each intensity, the original response options were recoded into a four-level ordinal scale: >1 time/week (or almost daily) = 3, 1 time/week = 2, 1–3 times/month = 1, and almost never/never = 0; responses such as “don’t know,” refusal, and proxy responses were coded as missing. Moderate-to-vigorous PA frequency was defined as the maximum of the vigorous and moderate frequency scores and was further dichotomized as Active versus Inactive. Participants were classified as Active if they reported engaging in moderate or vigorous activity at least once per week; otherwise, they were classified as Inactive (<1 time/week, including 1–3 times/month or almost never/never). Additional lifestyle and health-related covariates included smoking status, alcohol consumption, BMI, diabetes and hypertension, and depressive symptoms (Center for Epidemiologic Studies Depression Scale; CES-D). Statistical analysis For descriptive analyses, continuous variables are summarized as the mean (standard deviation [SD]), and categorical variables are presented as n (%). The proportion and pattern of missingness for each covariate are reported in Supplementary Table S1, Table S2. Given that missing data were common in older-adult cohorts and complete-case analysis can reduce statistical power and introduce selection bias, we used multiple imputation for missing covariates (R package “mice”). Estimates were obtained within each imputed dataset and pooled across imputations for subsequent regression analyses to reduce bias and improve the robustness and generalizability of the findings. To evaluate the association between serum 25(OH)D and sarcopenia, we fitted multivariable logistic regression models with sarcopenia as the outcome, adjusting for potential confounders. Serum 25(OH)D was categorized as “low” versus “higher” to compare sarcopenia risk across vitamin D status. Given evidence that diabetes may influence sarcopenia and related mechanisms [22,23], diabetes was incorporated in two ways: (1) analyses restricted to participants with diabetes to assess whether the association persisted and whether its magnitude differed; and (2) formal tests of interaction between vitamin D status and diabetes, evaluated on both multiplicative and additive scales. All statistical analyses were conducted in R (version 4.3.2), with statistical significance defined as a two-sided P < 0.05. Experimental study Animals and animal care This research protocol was reviewed and approved by the Laboratory Animal Ethics Committee of Shanxi Cancer Hospital (Approval No. 2022002). Two-month-old male Sprague–Dawley (SD) rats were obtained from the China National Institutes for Food and Drug Control (Beijing) (Animal Production License No. SCXK (Jing) 2022-0002). Rats were 8 weeks old at the beginning of the experiment. After a 1-week acclimation period and 14 weeks of intervention, rats were approximately 22 weeks old, or 5.5 months old, at the end of the experiment. D-galactose administration was used to induce accelerated aging-like phenotypes; therefore, this model represents a D-galactose-induced aging-like diabetic rat model rather than a naturally aged rat model [24]. Animals were housed under controlled conditions (20–26 °C; 40%–70% relative humidity) with a standardized 12-h light/12-h dark cycle and provided ad libitum access to food and water. At the end of the study, all rats were euthanized under anesthesia induced by 20% urethane. Animal experimental design and group After a 1-week acclimation, rats were randomly assigned to four groups according body weight: the control, the aged type 2 diabetes mellitus (T2DM) group; the aged T2DM with vitamin D deficiency (VDD); and VDD rats (with T2DM) supplemented with vitamin D₃ at 2000 IU/kg.bw (2000 IU). To induce an accelerated-aging T2DM phenotype, rats received daily intraperitoneal injections of D-galactose (50 mg/kg) [25,26], whereas control rats received an equal volume of saline. In parallel, rats in the T2DM groups were fed a high-sugar/high-fat diet for 14 weeks (vitamin D: 1,000 IU/kg diet). At week 5, T2DM-model rats were administered streptozotocin (STZ; 30 mg/kg, intraperitoneal) [[27], [28], [29]], while control rats received an equal volume of citrate buffer. Fasting blood glucose (FBG) was measured on days 3th, 7th and 14th after STZ injection; diabetes was confirmed when FBG was ≥16.7 mmol/L [30]. Rats assigned to the VDD group received the same D-galactose injection and a vitamin D–deficient high-sugar/high-fat diet for 14 weeks (vitamin D: 0 IU/kg diet), with STZ administered at week 5 using the same protocol. Serum 25(OH)D3 concentrations were measured by ELISA at baseline (week 0) and at weeks 4 th, 6 th, 10 th, and 14 th. Vitamin D deficiency was defined as serum 25(OH)D3 < 20 ng/mL [31]. In the 2000 IU group, after T2DM induction and establishment of vitamin D deficiency (D-galactose, vitamin D–deficient high-sugar/high-fat diet, and STZ at week 5th), vitamin D3 was given once daily by oral gavage at 2,000 IU/kg body weight from week 7 th until study completion. This dosing strategy was supported by prior work showing that oral vitamin D3 at 2,000 IU/kg rapidly increases serum 25(OH)D in Sprague–Dawley rats without materially altering serum calcium or phosphorus [32,33]. The experimental timeline and intervention protocol were shown in Figure S1 C. Throughout the study, body weight, food intake, water intake, and urine output were recorded weekly, and FBG was assessed every 2 weeks. Notably, D-galactose was used to induce accelerated aging-like phenotypes rather than natural chronological aging. Chronic D-galactose administration has been widely used in rodent studies to model aging-related biological changes, including oxidative stress, mitochondrial dysfunction, inflammation, and cellular senescence [34]. Hind limb grip strength test Hind limb muscle strength was assessed using a digital grip strength meter (GripStrengthMeterB; Beijing Youbiai, Beijing, China). Briefly, rats were gently lifted by the base of the tail and allowed to grasp the pull bar with the hind paws only, while the forepaws were kept clear of the sensor. After a stable grip was achieved, the investigator pulled the rat backward by the tail at a constant speed, parallel to the surface, until the grip was released. The peak tensile force (N) generated by the hind limbs was recorded automatically. Each animal completed three consecutive trials, with a 30-s rest interval between trials. Hematoxylin and Eosin Staining (H&E) To assess histomorphological changes in the gastrocnemius muscle (GM), GM tissues were fixed in 4% paraformaldehyde, paraffin-embedded, and sectioned into 4-μm serial slices. After deparaffinization and rehydration, sections were stained with hematoxylin for 5 min, differentiated in 1% hydrochloric acid alcohol for 10 s, and counterstained with eosin. Sections were then dehydrated through graded ethanol, cleared, and mounted. Images were captured at 200× magnification using a light microscope, with three randomly selected fields per animal. Transmission electron microscopy (TEM) observation GM tissue was cut into 1 mm3 pieces, fixed in 3% glutaraldehyde, and post-fixed in 1% osmium tetroxide. After washing in phosphate-buffered saline (PBS), samples were dehydrated through a graded ethanol series (30%, 50%, 70%, 80%, 90%, and 100%; 15–20 min per step), transferred to acetone, and incubated for 20 min. It was then transferred to acetone and incubated for 20 min, processed into ultrathin sections (70–80 nm) on copper grids, double-stained with uranyl acetate and lead citrate, and examined by transmission electron microscopy (Hitachi, Tokyo, JEOL, Tokyo, Japan). Homeostasis model assessment of insulin resistance (HOMA-IR) Fasting serum insulin (FINS) was quantified using an insulin enzyme-linked immunosorbent assay (ELISA) kit (Mlbio, Shanghai, China). IR was estimated using the HOMA-IR index: HOMA-IR = FINS (mIU/L) × FBG (mmol/L) / 22.5. Immunofluorescence staining for muscle fiber typing Frozen GM sections were fixed in 4% paraformaldehyde, washed with PBS, and blocked with bovine serum albumin (BSA). Sections were incubated overnight at 4 °C with primary antibodies against type I (slow) and type II (fast) myosin heavy chains, washed with PBS, and then incubated for 1 h at room temperature in the dark with the corresponding fluorescent secondary antibodies. Type I fibers were detected with Alexa Fluor 488 (green) and type II fibers with Alexa Fluor 594 (red); nuclei were counterstained with DAPI. Sections were mounted with an anti-fade mounting medium and imaged using a fluorescence microscope. Oil Red O staining To assess neutral lipid deposition in GM, frozen rat gastrocnemius sections were fixed for 15 min, stained with Oil Red O for 15 min, differentiated with 60% isopropanol, and counterstained with Mayer’s hematoxylin for 2 min. Sections were rinsed with running water and then mounted. Western blot analysis Total protein was extracted from GM using RIPA lysis buffer (G2002, Servicebio, China) and quantified with a BCA assay kit (AR1189, Boster, China). Equal amounts of protein were separated by 8%, 10%, or 12% SDS–PAGE (Seven, Beijing, China) and transferred to PVDF membranes (IPVH00010, Millipore, USA). Membranes were blocked with 5% skim milk in Tris-buffered saline containing 0.1% Tween-20 (TBST) at 37 °C for 2 h and incubated overnight at 4 °C with primary antibodies (Supplementary Table S3). After TBST washes, membranes were incubated for 1 h at room temperature with HRP-conjugated anti-rabbit IgG (1:5,000–1:10,000; BA1056, Boster, Wuhan, China). Bands were detected using an enhanced chemiluminescence (ECL) kit (BMU101, Abbkine, Wuhan, China). Statistical analysis Data were presented as mean ± SD and analyzed using one-way analysis of variance (ANOVA), followed by the least significant difference (LSD) or Dunnett’s T3 post hoc test. Graphs were generated using GraphPad Prism 10.0. P < 0.05 was considered statistically significant. Results Associations between Vitamin D levels and sarcopenia in humans This study included 3,246 participants from the HRS cohort, with a mean age of 72.10 ± 8.53 years, and 4,274 participants from the ELSA cohort, with a mean age of 69.91 ± 7.27 years. In ELSA, 318 participants had sarcopenia. Compared with non-sarcopenic individuals, those with sarcopenia were older (78.94 vs 69.19 years), were more often women (60.4% vs 39.6%), had lower educational attainment, were less physically active, and had lower grip strength (15.33 vs 29.02) and serum 25(OH)D concentrations (44.43 vs 50.07). In HRS, 187 participants had sarcopenia. Participants with sarcopenia were older than those without sarcopenia (81.48 vs 71.53 years), were more often women (63.6% vs 36.4%), and had lower grip strength (16.24 vs 29.70). Serum 25(OH)D level did not differ significantly between groups. Baseline characteristics were provided in Supplementary Tables S5 and S6. Table S4 showed the association between serum 25(OH)D status and sarcopenia risk. In the HRS cohort, after full adjustment for potential confounders, participants with low 25(OH)D levels (<50 nmol/L) showed a higher, although not statistically significant, odds of sarcopenia compared with those with higher levels (≥50 nmol/L) (OR = 1.292, 95% CI: 0.823–2.028). Similarly, in the ELSA cohort, low 25(OH)D levels were associated with a slightly higher but non-significant odds of sarcopenia (OR = 1.028, 95% CI: 0.782–1.350). These findings suggest that low vitamin D status may be a potential risk factor for sarcopenia, although the associations did not reach statistical significance in either cohort. Associations between Vitamin D levels and sarcopenia stratified by diabetes Status Table 1 summarized the association between serum 25(OH)D status and sarcopenia among participants with diabetes. In the HRS cohort, low 25(OH)D levels were associated with directionally higher, but statistically non-significant, odds of sarcopenia in the fully adjusted model (OR = 1.778, 95% CI: 0.843–3.750). In the ELSA cohort, a similar direction of association was observed, and the association reached statistical significance only in the fully adjusted model (OR = 2.242, 95% CI: 1.055–4.764). These findings suggest that low vitamin D status may be associated with greater sarcopenia susceptibility among older adults with diabetes, although the evidence was statistically significant only in ELSA and should therefore be interpreted cautiously. Table 1Association between serum 25(OH)D level and sarcopenia in diabetes.Table 1Cohort25(OH)D levelModel 1Model 2Model 3OR (95% CI)OR (95% CI)OR (95% CI)HRS≥50 nmol/L(higher)refrefref<50 nmol/L(low)1.053 (0.547–2.028)1.739 (0.839–3.604)1.778 (0.843–3.750)ELS≥50 nmol/L(higher)refrefref<50 nmol/L(low)2.257 (1.195–4.263)2.658 (1.293–5.467)2.242 (1.055–4.764)Note: HRS: The Health and Retirement Study; ELSA: The English Longitudinal Study of Ageing; OR: odds ratio; CI: confidence interval. Model 1 represents the unadjusted model; Model 2 was adjusted for confounders, including age, gender, race, education and marital status; Model 3 further adjusted for BMI, smoking, alcohol consumption, CES-D, hypertension, and physical activity in addition to the covariates included in Model 2. Association between serum 25(OH)D level and sarcopenia in diabetes. Note: HRS: The Health and Retirement Study; ELSA: The English Longitudinal Study of Ageing; OR: odds ratio; CI: confidence interval. Model 1 represents the unadjusted model; Model 2 was adjusted for confounders, including age, gender, race, education and marital status; Model 3 further adjusted for BMI, smoking, alcohol consumption, CES-D, hypertension, and physical activity in addition to the covariates included in Model 2. Table 2 presented the joint associations and interaction between serum 25(OH)D status and diabetes in relation to sarcopenia. In the HRS cohort, with participants who had higher 25(OH)D and no diabetes as the reference group, diabetes alone was associated with higher odds of sarcopenia (OR = 1.15, 95% CI: 0.78–1.71), and low 25(OH)D alone showed a similar direction of association (OR = 1.19, 95% CI: 0.67–2.12). Participants with both low 25(OH)D and diabetes had the highest odds (OR = 1.79, 95% CI: 0.89–3.60). However, no interaction was detected on either the multiplicative or additive scale. Table 2Interaction effects of serum 25(OH)D levels and diabetes on sarcopenia.Table 2Cohort25(OH)D * DiabetesOR (95% CI)Multiplicative interactionAdditive interaction (RERI)HRSHigher + No diabetesrefHigher + Diabetes1.15 [0.78, 1.71]Low + No diabetes1.19 [0.67, 2.12]Low + Diabetes1.79 [0.89, 3.6]1.3 [0.52, 3.24]0.44 [-0.87, 2.23]ELSAHigher + No diabetesrefHigher + Diabetes0.7 [0.36, 1.36]Low + No diabetes0.88 [0.66, 1.19]Low + Diabetes1.66 [1.06, 2.61]2.68 [1.23, 5.81]1.08 [0.25, 1.97]Note: HRS: The Health and Retirement Study; ELSA: The English Longitudinal Study of Ageing; OR: odds ratio; CI: confidence interval; RERI: relative excess risk due to interaction. Higher means serum 25(OH)D >50 nmol/l. Low means serum 25(OH)D ≤50 nmol/l. Model was adjusted for confounders, including age, gender, race, education, marital status, BMI, smoking, alcohol consumption, CES-D, hypertension, and physical activity.The null value is 1 for the multiplicative interaction and 0 for Additive interaction (RERI). Thus, multiplicative interaction is supported when the 95% CI for the interaction OR excludes 1, whereas additive interaction is supported when the 95% CI for RERI excludes 0 [56,57]. Interaction effects of serum 25(OH)D levels and diabetes on sarcopenia. Note: HRS: The Health and Retirement Study; ELSA: The English Longitudinal Study of Ageing; OR: odds ratio; CI: confidence interval; RERI: relative excess risk due to interaction. Higher means serum 25(OH)D >50 nmol/l. Low means serum 25(OH)D ≤50 nmol/l. Model was adjusted for confounders, including age, gender, race, education, marital status, BMI, smoking, alcohol consumption, CES-D, hypertension, and physical activity. The null value is 1 for the multiplicative interaction and 0 for Additive interaction (RERI). Thus, multiplicative interaction is supported when the 95% CI for the interaction OR excludes 1, whereas additive interaction is supported when the 95% CI for RERI excludes 0 [56,57]. In the ELSA cohort, neither diabetes alone nor low 25(OH)D alone was associated with increased odds of sarcopenia. Consistent with findings from the HRS cohort, the combined presence of low 25(OH)D and diabetes was associated with higher odds of sarcopenia (OR = 1.66, 95% CI: 1.06–2.61). Interaction analyses further indicated statistically significant multiplicative and additive interactions in ELSA. (interaction OR = 2.68, 95% CI: 1.23–5.81; RERI = 1.08, 95% CI: 0.25–1.97). 3.4 Effects of vitamin D deficiency on diabetic phenotypes, gastrocnemius senescence, hyperglycemia, and IR in aged T2DM rats with skeletal muscle atrophy At week 1th, serum 25(OH)D₃ level did not differ among groups. By week 6th, 25(OH)D₃ level in the vitamin D–deficient (VDD) group declined to <20 ng/mL and was significantly lower than those in the Control and T2DM groups (P < 0.05; Fig. 1A), confirming successful induction of vitamin D deficiency. By week 14th, body weight and cumulative weight gain were significantly less than in both the T2DM and VDD groups compared to the control group (P < 0.05; Fig. 1B). Fasting insulin (FINS) levels differed significantly among groups. Relative to the Control group, FINS levels were significantly higher in both the T2DM and VDD groups (P < 0.05). Furthermore, FINS level was higher in the VDD group than the T2DM group (P < 0.05). Compared with the VDD group, FINS levels were significantly decreased in the 2000 IU treatment group (Fig. 1C; P < 0.05). In parallel, fasting blood glucose (FBG) in the T2DM group exceeded 16.7 mmol/L two weeks after STZ administration (week 7th), and by week 14 th, FBG was significantly lower in the 2000 IU treatment group than in the VDD group (P < 0.05; Fig. 1D). HOMA-IR was significantly worsened in the VDD group compared with the Control and T2DM groups, whereas vitamin D treatment improved insulin resistance (P < 0.05; Fig. 1F). Typical diabetic manifestations—24 -h food intake, water intake, and urine output—were also significantly increased in the T2DM and VDD groups compared with Controls (P < 0.05; Fig. 1E, G, and H), with more pronounced changes in the VDD group, suggesting that vitamin D deficiency aggravated diabetic symptoms. Finally, senescence-associated markers in the gastrocnemius muscle were elevated in the VDD group, with increased expression of p53, p21, and β-galactosidase compared with the Control and T2DM groups; these markers were significantly reduced in the 2000 IU treatment group (P < 0.05; Fig. 1I–K), indicating a protective effect of vitamin D against gastrocnemius senescence. Collectively, vitamin D deficiency exacerbated core diabetic features and insulin resistance and intensified gastrocnemius senescence, whereas vitamin D supplementation mitigated muscle senescence and improved hyperglycemia and insulin sensitivity in aged T2DM rats with skeletal muscle atrophy. Effects of vitamin D deficiency on diabetic phenotypes, gastrocnemius muscle aging markers, hyperglycemia, and IR in aged rats with T2DM. (A) Serum 25(OH)D3 concentration (n = 6). (B) Body weight (n = 6). (C) Fasting serum insulin concentration (n = 6). (D) Fasting blood glucose (FBG; n = 6). (E) 24-h food intake (n = 6). (F) Homeostasis model assessment of insulin resistance (HOMA-IR; n = 6). (G) 24-h water intake (n = 6). (H) 24-h urine output (n = 6). (I) Densitometric quantification of p53 (n = 3). (J) Densitometric quantification of p21 (n = 3). (K) Densitometric quantification of β-galactosidase (n = 3). (L) Representative Western blot images of p53, p21, and β-galactosidase in gastrocnemius muscle. Compared with the control group, aP < 0.05; compared with the T2DM group, bP < 0.05; compared with the VDD group, cP < 0.05. IR: insulin resistance; T2DM: type 2 diabetes mellitus. 3.5 Effects of vitamin D supplementation on hind limb grip strength and muscle mass in the aged T2DM rats with skeletal muscle atrophy Muscle mass and strength are the key for evaluating skeletal muscle atrophy. Histological assessment by H&E staining showed obvious gastrocnemius atrophy in the VDD group, as evidenced by shrunken myofibers, marked fiber-size heterogeneity, disorganized fiber architecture, and increased interstitial spacing (Fig. 2A). In contrast, vitamin D supplementation markedly improved muscle morphology. Quantification indicated that the muscle fiber cross-sectional area in the 2000 IU group increased by approximately 32.5% relative to the VDD group (P < 0.05; Fig. 2D–E). The functional study results further confirmed the findings on the structure of the GM. Compared with the VDD group, the 2000 IU treatment group exhibited significantly greater hind limb grip strength, with average and maximal values increasing by approximately 88.1% and 31.0%, respectively (P < 0.05; Fig. 2H–I). Consistent with these findings, the wet weight and wet weight ratio of the GM were significantly increased in the 2000 IU treatment group (P < 0.05, Fig. 2B, F, G). Effects of Vitamin D Deficiency on GM mass in aged T2DM rats (A) Representative H&E-stained images of the gastrocnemius muscle (100×; scale bar = 200 μm). (B) Representative gross images of the gastrocnemius muscle. (C) Immunofluorescence staining for muscle fiber types. (D) Grayscale intensity analysis of H&E staining, n = 3. (E) Cross-sectional area analysis based on H&E staining, n = 3. (F) Gastrocnemius muscle weight, n = 6. (G) Wet weight ratio of the gastrocnemius muscle, n = 6. (H) Mean hindlimb grip strength (N), n = 6. (I) Maximum hindlimb grip strength, n = 6. (J) Atrogin-1 expression levels in the gastrocnemius muscle, n = 3. (K) MuRF1 expression levels in the gastrocnemius muscle, n = 3. (L) Area ratio of type I to type II muscle fibers. (M) Grayscale intensity ratio of type I to type II muscle fibers. (N) MyHC I expression levels in the gastrocnemius muscle, n = 3. (O) ACTN3 expression levels in the gastrocnemius muscle, n = 3. (P) Representative Western blot images of MuRF1, Atrogin-1, MyHC I, and ACTN3 in the gastrocnemius muscle. Compared with the control group, aP < 0.05; compared with the T2DM group, bP < 0.05; compared with the VDD group, cP < 0.05. IR: insulin resistance; T2DM: type 2 diabetes mellitus; GM: gastrocnemius muscle. Immunofluorescence staining showed that the proportion of slow-twitch (type I) fibers was significantly increased in the VDD group, whereas the 2000 IU treatment group exhibited an increased proportion of fast-twitch (type II) fibers compared with the VDD group. Relative to the VDD group, the 2000 IU treatment group showed a significant reduction in type I fiber indices, including a 75.1% decrease in the ratio of mean cross-sectional area and an 81.8% decrease in the ratio of mean gray value (P < 0.05; Fig. 2C,L,M). These results indicate that vitamin D supplementation mitigated the VDD-associated slow-fiber predominance and shifted fiber composition toward a faster phenotype. To validate fiber-type remodeling at the protein level, we quantified fiber-specific markers by Western blotting. In agreement with the immunofluorescence results, Myosin Heavy Chain I (MyHC I; type I marker) was significantly downregulated in the 2000 IU treatment group compared with the VDD group (P < 0.05; Fig. 2N), whereas α-Actinin-3 (ACTN3; type II marker) was significantly upregulated (P < 0.05; Fig. 2O). Together, these data indicate that vitamin D promotes a more favorable fiber-type profile by repressing type I–associated proteins while enhancing type II–associated proteins. Finally, to explore molecular pathways related to muscle mass preservation, we assessed the expression of the muscle atrophy–associated E3 ubiquitin ligases MuRF1 and Atrogin-1. Both proteins were significantly reduced in the gastrocnemius of the 2000 IU treatment group relative to the VDD group (P < 0.05; Fig. 2J–K), suggesting that vitamin D attenuates muscle wasting, at least in part, by suppressing key catabolic mediators. 3.6 Vitamin D deficiency exacerbates intramuscular fat infiltration by disrupting the balance between myogenic and adipogenic differentiation Intramuscular adipose tissue (IMAT) infiltration impairs skeletal muscle structure and function; however, the mechanisms by which vitamin D deficiency promotes IMAT accumulation remain incompletely defined. To address this question, we first examined gastrocnemius ultrastructure using transmission electron microscopy (TEM). As shown in Fig. 3A, myofibers in the Control group were regularly aligned and exhibited intact ultrastructural organization. In contrast, the VDD group displayed marked myofiber disarray, accompanied by prominent “bead-like” lipid droplet deposition within intermyofibrillar regions. We next quantified intramuscular lipid accumulation by Oil Red O staining. 2000 IU treatment group markedly reduced lipid deposition, with the mean intramuscular fat area decreased by approximately 72.0% and the mean gray value reduced by 77.4% (P < 0.05; Fig. 3B–D). Effects of Vitamin D Deficiency on fat infiltration of the GM in aged T2DM rats (A) Representative transmission electron microscopy images of gastrocnemius muscle fibers (30,000×; scale bar = 1 μm). (B) Oil Red O staining of the gastrocnemius muscle (100×; scale bar = 200 μm). (C) Quantification of Oil Red O–positive area, n = 3. (D) Grayscale intensity analysis of Oil Red O staining, n = 3. (E) Wnt10b expression levels in the gastrocnemius muscle, n = 3. (F) β-catenin expression levels in the gastrocnemius muscle, n = 3. (G) MyoG expression levels in the gastrocnemius muscle, n = 3. (H) MyoD expression levels in the gastrocnemius muscle, n = 3. (I) Myf5 expression levels in the gastrocnemius muscle, n = 3. (J) PPARγ expression levels in the gastrocnemius muscle, n = 3. (K) C/EBPβ expression levels in the gastrocnemius muscle, n = 3. (L) C/EBPδ expression levels in the gastrocnemius muscle, n = 3. (M) FABP4 expression levels in the gastrocnemius muscle, n = 3. (N) Representative protein bands for Wnt10b, β-catenin, MyoG, MyoD, Myf5, PPARγ, C/EBPβ, C/EBPδ, FABP4, and GAPDH. T2DM: type 2 diabetes mellitus; GM: gastrocnemius muscle (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article). To elucidate the molecular basis for these histological changes, we assessed key regulators of myogenic and adipogenic differentiation. Because the Wnt/β-catenin pathway promotes myogenesis and restrains adipogenesis, we hypothesized that VD deficiency would suppress this axis and shift lineage commitment toward adipogenic differentiation. Western blot analyses showed that the VDD group exhibited significant reductions in Wnt10b and β-catenin, accompanied by the decrease of canonical myogenic regulatory factors (MyoG, MyoD, and Myf5) relative to Controls. Importantly, 2000 IU treatment group significantly restored the expression of these myogenic proteins compared with the VDD group (P < 0.05; Fig. 3E–I). Because adipocyte differentiation is initiated by early C/EBP family members (C/EBPβ and C/EBPδ), which induce PPARγ and activate downstream effectors, we further evaluated the adipogenic signaling pathways markers as indicators of adipogenic commitment. Compared with the VDD group, the 2000 IU treatment group demonstrated significant suppression of C/EBP β, C/EBP δ, PPAR γ, and FABP4 (P < 0.05; Fig. 3J–M). Representative immunoblot bands were shown in Fig. 3N. Taken together, these findings indicate that vitamin D deficiency aggravates IMAT infiltration in GM, whereas vitamin D supplementation attenuates lipid accumulation. Mechanistically, these effects are consistent with VD-dependent regulation of the Wnt/β-catenin pathway, which influences the balance between myogenic maintenance and adipogenic differentiation. Discussion Sarcopenia is a systemic condition characterized by progressive skeletal muscle degeneration that substantially compromises quality of life in older adults, yet effective pharmacological therapies remain limited [35]. Accordingly, identifying early risk factors and developing feasible intervention strategies are critical.In two independent population-based cohorts, lower vitamin D levels were associated with directionally higher odds of sarcopenia in older adults, although these associations did not reach statistical significance after multivariable adjustment. These findings do not establish low vitamin D status as an independent determinant of sarcopenia, but the consistent direction of the estimates suggests that it may represent a potential risk factor. Stratified and joint analyses further suggested that this association may be more evident among individuals with diabetes, with participants exposed to both low vitamin D and type 2 diabetes showing the highest sarcopenia risk. A significant interaction was observed in the ELSA cohort, supporting a possible joint contribution of vitamin D deficiency and diabetes to sarcopenia susceptibility, although further validation is needed. To explore the biological plausibility of this relationship, we established an aging T2DM rat model and found that vitamin D deficiency impaired muscle composition and quality, accompanied by dysregulated expression of MyHC I and ACTN3 and increased intramuscular fat infiltration. Vitamin D supplementation partially reversed these phenotypes. Previous studies have found that patients with T2DM are more prone to developing sarcopenia [36]. In our stratified analyses among participants with diabetes, low serum 25(OH)D levels showed directionally higher odds of sarcopenia in both cohorts, but statistical significance was observed only in the fully adjusted model in ELSA, suggesting that this potential diabetes-related vulnerability should be interpreted cautiously. T2DM is characterized by IR and systemic, chronic low-grade inflammation [37]. Insulin resistance compromises skeletal muscle uptake and utilization of energy substrates and suppresses myofibrillar protein synthesis. As a principal regulator of whole-body glucose homeostasis, skeletal muscle under physiological conditions depends largely on the IRS-1/PI3K/Akt/mTORC1 insulin-signaling axis to stimulate protein translation and anabolism [38], while promoting GLUT4 translocation to the sarcolemma to increase glucose uptake [39]. In the diabetic milieu, elevated free fatty acids, chronic inflammation, and oxidative stress activate stress kinases and induce aberrant serine phosphorylation of IRS-1, thereby attenuating Akt activation and downstream mTORC1 signaling [40,41]. The resulting impairment in anabolic signaling and substrate utilization contributes to a progressive decline in muscle mass and strength. Skeletal muscle dysfunction can reciprocally worsen insulin resistance and disrupt glucose homeostasis, establishing a self-perpetuating pathogenic cycle [42]. Notably, our findings suggest that vitamin D deficiency may exacerbate fatty infiltration of the gastrocnemius muscle. Intramuscular fat infiltration, a key pathological feature of sarcopenia, is regulated in part by Wnt/β-catenin signaling. Moderate activation of the Wnt pathway stabilizes β-catenin and suppresses adipogenic transcription factors, including PPARγ and C/EBPβ, thereby inhibiting adipocyte differentiation and intramuscular lipid deposition [43]. Inhibition of Wnt/β-catenin signaling promotes adipocyte differentiation, leading to increased intramuscular fat infiltration and impairing the maintenance of muscle mass and function. Consistent with prior evidence, vitamin D deficiency is associated with a higher risk of sarcopenia [44,45]. In a cohort of 1,705 adults aged ≥70 years followed for 5 years, participants with lower baseline serum 25(OH)D concentrations had a significantly greater risk of sarcopenia (OR = 2.40; 95% CI, 1.02–5.64) [46]. Aging is commonly accompanied by declining serum 25(OH)D levels and alterations in bone-related metabolism. Vitamin D is activated and inactivated by metabolic enzymes, including CYP27B1 and CYP24A1, and exerts its biological effects through binding to the VDR [47]. Because VDR is expressed in skeletal muscle, vitamin D may directly contribute to the regulation of muscle function and mass; accordingly, some studies report that VDR expression is associated with greater muscle mass [48]. Notably, the ≥50 nmol/L category in our study should be interpreted as the reference group rather than as a marker of vitamin D adequacy. Although 50 nmol/L was used as a pragmatic cutoff to define lower vitamin D status, several geriatric and bone-health recommendations suggest that higher serum 25(OH)D concentrations, such as ≥75 nmol/L, may be preferable in older adults at risk of falls, fractures, or osteoporosis [[49], [50], [51], [52]]. The administration of a low-dose streptozotocin (30 mg/kg) in this study was strategically chosen to partialy impair pancreatic β-cell function rather than causing complete destruction. When combined with a high-sugar/high-fat diet, this protocol more accurately replicates the chronic pathophysiological progression of type 2 diabetes mellitus and persistent insulin resistance. In this T2DM backdrop, our study demonstrated that vitamin D intervention reduced intramuscular fat infiltration and was accompanied by improvements in muscle architecture and overall tissue quality. Consistent with these findings, prior experimental studies suggest that vitamin D deficiency promotes differentiation of mesenchymal progenitors toward adipogenic lineages [53]. Intramuscular fat infiltration is a key pathological feature of skeletal muscle dysfunction in aging and metabolic disorders [54]. Excess lipid accumulation can widen intermuscular spaces, reduce the density of functional contractile units, and further disrupt myocellular metabolic homeostasis through lipotoxicity. Mechanistic reviews propose that vitamin D may attenuate ectopic lipid deposition by regulating pathways involved in lipid uptake and oxidation, and that the vitamin D–VDR axis modulates lipid-handling capacity [55]. Moreover, 1,25(OH)2D supplementation enhances oxidative phosphorylation in skeletal muscle under vitamin D–deficient conditions and mitigates palmitate-induced mitochondrial dysfunction and triglyceride accumulation in myocytes [47]. In summary, these findings support vitamin D as a potential nutritional target to improve sarcopenia in older adults, particularly those with diabetes. This study has several limitations that warrant cautious interpretation. Firstly, cohort studies are inherently observational studies, and although various confounding factors have been adjusted for and multiple imputation has been used to handle missing values, it is still difficult to completely rule out the possibility of residual confounding and reverse causality. Moreover, the association between low serum 25(OH)D levels and sarcopenia was not statistically significant in the fully adjusted models of both cohorts, with statistical significance observed only in limited analyses such as the unadjusted model in ELSA; therefore, the epidemiological findings should be interpreted as suggestive rather than confirmatory. Second, HRS and ELSA differ in 25(OH)D assay platforms, sampling procedures, and variable collection protocols, which may introduce measurement error and between-cohort heterogeneity and contribute to differences in effect-size estimates. Third, the STZ–induced diabetes model primarily reflects pancreatic β-cell injury and hyperglycemia-related metabolic stress and does not fully recapitulate the long-term, progressive insulin resistance and metabolic syndrome phenotype characteristic of human T2DM. Third, the STZ-induced diabetes model primarily reflects pancreatic β-cell injury and hyperglycemia-related metabolic stress, and does not fully recapitulate the long-term, progressive insulin resistance and metabolic syndrome phenotype characteristic of human type 2 diabetes. In addition, repeated oral administration in the vitamin D-supplemented group may have introduced procedure-related stress. Future studies should therefore consider vehicle-treated controls across relevant groups or less stressful oral administration methods to minimize this potential influence. Furthermore, although D-galactose is widely used to induce aging-like phenotypes in rodents, this model does not fully reproduce the biological complexity of natural aging. Finally, sarcopenia-related phenotypes are mainly based on grip strength, gastrocnemius muscle mass, and histological parameters. While covering the two core dimensions of muscle strength and mass, these studies do not adequately evaluate overall functional performance and endurance (such as treadmill running and gait analysis) or more standardized body composition measurements (such as DEXA/BIA or muscle mass corrected for tibial length). Future studies should incorporate more comprehensive measures of physical function and body composition, and further validate the effects of vitamin D status on diabetes-related sarcopenia using naturally aged animal models. Conclusion In conclusion, this study demonstrates a significant association between vitamin D status and sarcopenia risk in older adults and suggests that maintaining higher serum 25(OH)D concentrations may be beneficial, particularly in the setting of diabetes or other metabolic disorders. Further research is warranted to validate and extend these findings using more comprehensive phenotyping and stronger causal-inference approaches, especially randomized controlled trials and mechanistic studies in high-risk older adults with type 2 diabetes mellitus (T2DM). In parallel, our animal experiments showed that vitamin D intervention was associated with improved grip strength, preservation of GM mass, and reduced intramuscular fat infiltration, thereby providing experimental support for the population-based results. These findings provide stronger evidence for a potential role of vitamin D in the “aging–diabetes–muscle wasting” cascade. CRediT authorship contribution statement Cheng Zhang: Writing – original draft, Formal analysis, Data curation, Conceptualization. Mingkai Li: Writing – original draft, Formal analysis, Data curation, Conceptualization. Chenyang Li: Supervision, Methodology, Conceptualization. Lushan Xue: Methodology, Conceptualization. Chenhui Lv: Methodology, Conceptualization. Qinfei Guo: Conceptualization. Yafei Zhao: Conceptualization. Haifeng Zhao: Supervision, Project administration, Funding acquisition, Writing – review and editing, Conceptualization. Ethics statement This study was conducted in accordance with the Declaration of Helsinki. The Health and Retirement Study (HRS) was approved by the University of Michigan Institutional Review Board, and the English Longitudinal Study of Ageing (ELSA) was approved by the National Research Ethics Service Committee South Central–Berkshire. This research protocol was reviewed and approved by the Laboratory Animal Ethics Committee of Shanxi Cancer Hospital (Approval No. 2022002). All participants provided written informed consent. Declaration of Generative AI and AI-assisted technologies in the writing process The authors did not use generative artificial intelligence (AI) and AI- assisted technologies in the writing process. Funding This study was supported by the Central Guiding Fund for the Development of Local Science and Technology (YDZJSX20231A056), the National Natural Science Foundation of China (No. 82574092), the Research Project Supported by Shanxi Scholarship Council of China (2023−103), and Shanxi Province Higher Education “Billion Project” Science and Technology Guidance Project. Data availability The original data of cohort can be obtained upon application on their respective official websites. HRS (https://hrs.isr.umich.edu/data-products), ELSA (https://beta.ukdataservice.ac.uk/datacatalogue/series/series?id=200011). Declaration of competing interest The authors declare that the research was conducted without any commercial or financial relationships that could be construed as a po tential conflict of interest. Acknowledgements The authors are grateful to the participants and staff of the Health and Retirement Study and the English Longitudinal Study of Ageing. We thank Mingkai Li for contributions to manuscript preparation. We also acknowledge Cheng Zhang and Mingkai Li as joint last authors who contributed equally to this work. Footnotes Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jnha.2026.100886. Supplementary data The following is Supplementary data to this article: References Associated Data Supplementary Materials Data Availability Statement The original data of cohort can be obtained upon application on their respective official websites. HRS (https://hrs.isr.umich.edu/data-products), ELSA (https://beta.ukdataservice.ac.uk/datacatalogue/series/series?id=200011).