This 12-week prospective cohort study was conducted in Poland during the autumn–winter season (October 1, 2024 to March 31, 2025). Participants were recruited consecutively from a single outpatient metabolic clinic; all eligible patients attending routine follow-up visits during the recruitment period were screened and invited to participate, ensuring a non-selective recruitment process.

No participants were lost to follow-up. All 200 enrolled patients completed baseline and 12-week assessments and were included in the final analyses.

Seasonal variability in endogenous vitamin D synthesis was minimized by conducting the entire study within a single autumn–winter period, when cutaneous vitamin D production is minimal at the study latitude.

Adherence to vitamin D supplementation was assessed using pill counts and self-reports. In addition, adherence was indirectly supported by the marked and consistent increase in serum 25(OH)D concentrations observed in the supplementation group.

No missing data were observed for primary or secondary endpoints; therefore, no imputation procedures were required and all analyses were performed on complete cases.

The study protocol was predefined and approved by the institutional bioethics committee (see: 2.7 Ethical Approval). However, the study was not prospectively registered in a public registry, reflecting its observational design. This is acknowledged as a limitation.

The decision to include a patient in the study was made by the attending physician. All recruited participants were fully informed about the study’s aims and conditions and provided written informed consent to participate. The study protocol was approved by the Bioethics Committee of the Poznan University of Medical Sciences (approval no. KB-664/23, September 13, 2023).

Participants were eligible if they met the following criteria:

Patients were excluded for any of the following:

Participants in the supplementation group received oral vitamin D₃ - 4,000 IU daily for 12 weeks. The control group received no supplementation. Both cohorts continued stable metformin monotherapy throughout the study to ensure treatment uniformity. Adherence was verified by pill counts and self-reports.

Primary endpoint:

• Change in serum 25(OH)D levels after 12 weeks.

Secondary endpoints:

Fasting blood samples were collected at baseline and after 12 weeks.

Laboratory Methods: All laboratory analyses were performed at the hospital’s certified clinical laboratory using standardized protocols.

Biochemical Measurements (29–31):

Quality Control: Internal quality control samples were analyzed with each batch. External quality assurance was maintained through participation in national proficiency testing programs.

BP was measured in the seated position with an automated sphygmomanometer (average of three readings). BMI was calculated from measured weight and height.

Continuous variables are expressed as mean ± standard deviation (SD). Baseline group comparisons were performed with independent Student’s t-tests or chi-square tests for categorical variables. Within-group changes were analyzed using paired t-tests, and between-group differences in mean change (Δ) were compared with independent t-tests.

Multiple logistic regression identified predictors of failure to achieve HbA1c ≤6.5% at 12 weeks. Covariates included age (per 10 years), gender, FSG (per 2 mmol/L), BMI (per 5 kg/m²), and duration of T2DM (per 6 months). Results are presented as odds ratios (OR) with 95% confidence intervals (CI).

A two-sided p-value <0.05 was considered statistically significant. Analyses were performed using SPSS version 27.0 (IBM Corp., Armonk, NY, USA).

The study cohort included 200 patients with T2DM (51% men, 49% women) with a mean age of 58.6 ± 7.35 years and a mean BMI of 30.1 ± 3.5 kg/m². In the entire study cohort, the mean duration of T2DM was 1.4 ± 0.6 years, the mean HbA1c level was 6.8 ± 0.9%, and the mean FSG level was 5.8 ± 0.2 mmol/l. At baseline, mean serum 25(OH)D concentrations in both groups (~21–22 ng/mL) corresponded to vitamin D deficiency/insufficiency according to established clinical thresholds. The baseline clinical characteristics of the study participants and controls are presented in Table 1. No significant differences were observed between groups at baseline with respect to age, sex distribution, diabetes duration, serum 25(OH)D, FSG, HbA1c, BP, serum calcium, or BMI (all p > 0.5). (Table 1).

After 12 weeks, the supplementation group achieved a mean serum 25(OH)D concentration of approximately 45 ng/mL, corresponding to vitamin D sufficiency, whereas the control group remained within the deficient/insufficient range. Significant improvements were also observed in FSG and HbA1c, with mean reductions of –0.4 mmol/L and –0.6%, respectively, in the supplementation group. Changes in SBP and DBP, serum calcium, and BMI were not statistically significant (Table 2; Figures 1, 2).

Multiple logistic regression identified predictors of failure to achieve target HbA1c ≤ 6.5% (≤ 48 mmol/mol). Higher baseline FSG and longer T2DM duration were strongly associated with suboptimal glycemic control, with odds ratios (OR) of 1.34 (95% CI: 1.12–1.61; p = 0.001) and 1.28 (95% CI: 1.07–1.54; p = 0.006), respectively. Elevated BMI was also significantly associated with poor glycemic control (OR 1.21; 95% CI: 1.01–1.47; p = 0.04). In contrast, age and sex were not significant predictors (Table 3).

The substantial increase in serum 25(OH)D among supplemented participants confirms effective absorption and bioavailability of the administered high-dose regimen. The observed reductions in FSG and HbA1c indicate that VDS was associated with modest improvements in glycemic parameters in patients with T2DM managed on stable metformin therapy.

These findings are consistent with several meta-analyses reporting modest but significant improvements in glycemic indices following VDS in diabetic or prediabetic populations. For example, Musazadeh et al. (2023) observed mean reductions in FSG (–3.08 mg/dL; 95% CI –3.97 to –2.19) and HbA1c (–0.05%; 95% CI –0.10 to –0.01) across 14 studies (17,136 participants) (32). Similarly, Farahmand et al. (2023) demonstrated significant decreases in FSG, HbA1c, and fasting insulin in RCTs, although the magnitude of effect was modest and heterogeneity was noted (2).

Conversely, other trials have failed to demonstrate significant HbA1c reductions, particularly in cohorts with sufficient baseline vitamin D levels or when using lower doses or shorter durations of therapy (33). Results of meta-analysis conducted by Qi et al. (2022) showed that VDS had no impact on HbA1c level and serum lipids, but reduced IR and BP values (34).

Although modest in magnitude, the observed HbA1c reduction (–0.6%) exceeds expected short-term biological variability and occurred in a population with relatively well-controlled T2DM at baseline, suggesting potential clinical relevance when interpreted as an adjunctive effect.

In the present study, both SBP and DBP declined modestly in the intervention group but the changes were not statistically significant. Evidence regarding vitamin D’s influence on BP remains inconsistent. Lee and Lee (2016) reported reductions in DBP among patients with T2DM receiving vitamin D, whereas Jafari et al. (2018) found a significant reduction in SBP but not DBP (35, 36). As it was mentioned above, meta-analysis by Qi et al. (2022) showed the improvement in hypertension control in patients with VDS (34). In contrast, analyses in the general population have shown minimal or no effect of VDS on BP, suggesting limited causal influence (37). An umbrella meta-analysis by Meng et al. (2023) reported that the overall effects of vitamin D supplementation on blood pressure were modest (38).

Mechanistically, vitamin D may modulate BP through regulation of the renin–angiotensin–aldosterone system (RAAS), improvement of endothelial function, and enhanced vascular smooth muscle tone. These mechanisms may underlie variable BP responses among susceptible individuals (38).

Numerous systematic reviews, meta-analyses, and an umbrella review have examined the relationship between vitamin D and obesity, combining data from RCTs and cohort studies (4, 26, 39). The findings vary depending on factors such as population characteristics, initial vitamin D status, dosage, and accompanying interventions. Overall, meta-analytic and umbrella evidence suggests modest reductions in BMI and weight circumference (WC), though the effects on body weight and fat mass remain inconsistent. Multimodal strategies involving exercise, protein supplementation, or caloric restriction achieve more substantial and consistent improvements in body composition and function (40).

In the present study there was no significant change in BMI. One of the key factor is a short study period. In the meta-analysis conducted by Musazadeh et al. (2022), the effects of vitamin D on body weight and fat mass were not considerable, but VDS significantly improved BMI, and WC (39). Similarly, Hu et al. in an umbrella meta-analysis reported small pooled reductions in BMI and WC, but no meaningful weight or fat−mass change overall (26).

In turn, Aquino et al. (2023) did not show a decrease of WC in patients with MS (41). Roth et al. (2022) reported that the effect of VDS on the preservation of free fat mass (FFM) during non-surgical and surgical weight loss remains controversial (40).

Logistic regression analysis identified higher baseline FSG, longer disease duration, and higher BMI as significant predictors of failure to achieve target HbA1c levels. These findings are physiologically plausible and consistent with clinical experience, as patients with more pronounced baseline hyperglycemia, longer disease duration, or obesity typically exhibit reduced responsiveness to metabolic interventions. The lack of association with age and sex suggests that metabolic factors, rather than demographic characteristics, primarily determine glycemic response.

This study offers several methodological strengths. The use of a standardized high-dose vitamin D regimen achieved robust improvement in serum 25(OH)D concentrations, enabling clear evaluation of metabolic effects. Homogeneous treatment with metformin monotherapy minimized pharmacologic confounding. The prospective, ambulatory design enhances external validity and clinical relevance, while multivariable modeling accounted for key covariates, strengthening internal validity. Seasonal variation in 25(OH)D concentrations could not have influenced the results, as the study was conducted during the autumn–winter 2024/2025 season.

Nonetheless, limitations must be acknowledged. The non-randomized design introduces potential residual confounding, including behavioral and unmeasured factors. The 12-week duration limits assessment of long-term outcomes. The sample size, although moderate, may have been underpowered to detect small changes in calcium or BMI. Finally, findings may not generalize to patients using complex or combination antidiabetic regimens.

The observational design precludes causal inference. Potential residual confounding, lack of placebo control, reliance on pragmatic adherence measures, and absence of prospective registration are acknowledged limitations. Observational designs often underestimate the true effect of vitamin D due to seasonal and temporal variability in serum levels. As emphasized by Pilz et al. (2022), large RCTs have generally failed to demonstrate consistent reductions in adverse outcomes, possibly owing to study design and dosing heterogeneity (42). Observational studies allowing variable supplementation and repeated 25(OH)D measurements may provide broader insight into dose–response relationships and health outcomes (43, 44).

Proposed biological mechanisms discussed herein are derived from prior experimental and clinical studies and were not directly investigated in the present study. They are included to provide biological plausibility and should be considered hypothesis-generating.

The present findings support the hypothesis that high-dose VDS may offer adjunctive benefits in glycemic control among obese T2DM patients with baseline deficiency. However, due to the observational nature of this study, causality cannot be inferred. Future RCTs with longer follow-up, uniform treatment regimens, and stratification by baseline vitamin D status and body composition are warranted. Mechanistic studies exploring effects on insulin sensitivity, β-cell function, inflammation, RAAS activity, and endothelial function would further elucidate the metabolic and cardiovascular implications of vitamin D in T2DM and obesity.

Recent research suggests that vitamin D supplementation may offer beneficial effects on oxidative stress as well as on reproductive, endocrine, and metabolic functions (45). In Hashimoto’s thyroiditis VDS shows promise in influencing immune activity, reducing symptom severity, and enhancing overall quality of life (46). Further investigation is needed to clarify the full range of potential therapeutic applications of VDS in diabetic neuropathy and statin-associated muscle symptoms (SAMS) (47, 48). So far, evidence suggest that vitamin D may support fertility and overall metabolic well-being by modulating hormone concentrations and key metabolic indicators (45). Despite these encouraging findings, additional studies are required to clarify the most effective dosage, treatment length, and timing for supplementation. It is essential for scientists, healthcare professionals, and policy leaders to work together to develop clear, evidence-based recommendations for the use of vitamin D in above mentioned indications (45–48).

High-dose VDS was associated with improved vitamin D status and modest improvements in glycemic control in obese patients with T2DM. These findings should be interpreted as associative rather than causal and support further investigation in randomized, placebo-controlled trials with extended follow-up and mechanistic endpoints.