Vitamin D supplementation in individuals with type 1 diabetes mellitus (T1DM): a systematic review and meta-analysis focusing on pediatrics

Abstract

Backgrounds

Type 1 diabetes mellitus (T1DM) is a significant global health concern, particularly among pediatric populations. Emerging evidence suggests that vitamin D supplementation may support glycemic control in individuals with T1DM; however, existing findings are inconsistent, underscoring the need for further investigation.

Methods

A comprehensive search of PubMed, Scopus, Cochrane Central, and Web of Science was conducted up to January 2024 to identify relevant English-language randomized controlled trials (RCTs). Risk of bias (ROB) was assessed using the ROB-2 tool. A random-effects model was employed for the meta-analysis using Stata version 17.

Results

From 2,744 records screened, twelve RCTs including 485 participants with T1DM were included. Vitamin D supplementation did not significantly impact HbA1c levels (-1.60 [-3.78, 0.57]; I² = 98.07%) but was associated with significant reductions in C-peptide levels (-2.54 [-4.97, -0.11]; I² = 97.03%), fasting blood sugar (FBS) (-1.44 [-2.67, -0.22]; I² = 91.10%), and daily insulin requirements (-0.44 [-0.82, -0.06]; I² = 58.64%). Moreover, 25(OH)D concentrations significantly increased following supplementation (4.19 [3.26, 5.13]; I² = 82.85%). No serious adverse events were reported, supporting the safety of supplementation.

Conclusion

Vitamin D supplementation showed potential benefits in reducing insulin requirements and fasting blood glucose in pediatric T1DM populations. However, its effect on HbA1c remains inconclusive, and the observed reduction in C-peptide levels requires cautious interpretation due to high heterogeneity and possible confounding factors. The certainty of evidence for all outcomes was rated as “low” to “very low” per GRADE assessment. Further large-scale, long-term RCTs are warranted.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13098-025-02008-9.

Introduction

Type 1 diabetes mellitus (T1DM) is a chronic autoimmune disease characterized by the destruction of pancreatic β-cells, resulting in absolute insulin deficiency and lifelong insulin dependency [1, 2]. The global incidence of T1DM is estimated at 15 per 100,000 person-years [3], with over 1.5 million affected children and adolescents worldwide [4]. The incidence is rising, particularly in low socioeconomic settings, influenced by genetic predisposition, environmental exposures, infections, gut microbiota, diet, toxins, and psychosocial stressors [5].

Among the modifiable environmental factors, vitamin D has garnered considerable attention for its potential immunomodulatory and anti-inflammatory effects, which may influence the onset and progression of autoimmune diseases, including T1DM [6, 8]. Vitamin D deficiency has been implicated in β-cell dysfunction and apoptosis through the promotion of systemic inflammation, suggesting a potential role in T1DM pathogenesis [7, 9]. Vitamin D receptors (VDR) are expressed in pancreatic β-cells, as well as in various immune cells, including B and T lymphocytes, indicating a regulatory role in immune homeostasis and autoimmunity [10, 11].

Observational studies have reported associations between low serum 25(OH)D levels and poor glycemic control in T1DM patients, while vitamin D supplementation has been linked to improved β-cell preservation and metabolic outcomes [12–14]. Notably, early-life supplementation with vitamin D has been associated with a reduced risk of developing T1DM, with some studies reporting a 29% decrease in risk among supplemented infants [11–13]. However, despite these promising findings, clinical trials investigating the efficacy of vitamin D supplementation in individuals with T1DM have yielded inconsistent results [15–19].

Current clinical guidelines generally do not recommend routine supplementation for glycemic control in T1DM, except to address established vitamin D deficiency [16]. Nevertheless, the widespread use of over-the-counter vitamin D supplements necessitates further clarification of their safety and therapeutic utility in this population [17–19]. Therefore, this systematic review and meta-analysis aimed to evaluate the efficacy and safety of vitamin D supplementation on glycemic indices, insulin requirements, and serum 25(OH)D levels in individuals with T1DM, with a specific focus on pediatric populations.

Methods

Protocol registration and reporting guidelines

This systematic review and meta-analysis adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines [20]. The protocol was prospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO) under the ID CRD42024588550. The completed PRISMA checklist is provided in Supplementary Files 1.

Search strategy

A comprehensive literature search was conducted to identify relevant English-language studies published up to January 2024. The electronic databases searched included PubMed, Scopus, Cochrane Central Register of Controlled Trials (CENTRAL), and Web of Science. The search strategy incorporated combinations of Medical Subject Headings (MeSH) and free-text terms such as: “vitamin D,” “type 1 diabetes mellitus,” “cholecalciferol,” “ergocalciferol,” “calcitriol,” “alfacalcidol,” and “randomized controlled trial.” Additionally, grey literature was explored through Google Scholar, ClinicalTrials.gov, and WHO ICTRP. No additional eligible trials were identified. The full search strategies are provided in Supplementary Table 1. Only English-language publications were included, which may introduce language bias.

Eligibility criteria

The eligibility criteria were defined using the PICO framework as follows:

• Population: Pediatric individuals (≤ 18 years) diagnosed with T1DM. One study (Li et al., 2009) included young adults (mean age 18.4 years); sensitivity analysis excluding this study was performed.

• Intervention: Vitamin D supplementation (including cholecalciferol, ergocalciferol, alfacalcidol, or calcitriol) in various dosages, as an adjunct to routine insulin therapy.

• Comparison: Placebo or no vitamin D supplementation, both combined with routine insulin therapy.

• Outcomes: Changes in glycemic control (HbA1c, fasting blood sugar [FBS], C-peptide), total daily insulin requirement, and serum 25(OH)D concentration.

Only randomized controlled trials (RCTs) reporting outcome measures at both baseline and the end of treatment were included in the meta-analysis. Studies were included in qualitative synthesis if they reported only directional trends (e.g., ‘improved HbA1c’) without means, standard deviations, or confidence intervals.

Study selection and data extraction

Two reviewers (TZ and SF) independently screened the titles and abstracts for eligibility, followed by full-text assessments of potentially relevant studies. Discrepancies were resolved through discussion or consultation with a third reviewer (LY). A standardized data extraction form was used to record the following information from each study: author, country, study design, sample size, mean age, sex distribution, type and dose of vitamin D intervention, duration of treatment, reported adverse effects, and primary outcome measures.

Risk of bias assessment

The Cochrane Risk of Bias 2 (RoB-2) tool was employed to assess methodological quality across five domains: randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result [21, 22]. Each study was categorized as having low risk, some concerns, or high risk of bias.

Statistical analysis

Statistical analyses were performed using Stata software version 17. For continuous outcomes, the mean difference (MD) and 95% confidence interval (CI) were calculated using a random-effects model. Heterogeneity among studies was assessed using the I² statistic and Chi-square test. Subgroup analyses were performed based on nationality, disease duration, ROB score, and study duration. Publication bias was evaluated using Begg’s test and funnel plot inspection for outcomes with < 10 studies. For outcomes with ≥ 10 studies, Egger’s test was also applied. Trim-and-fill analysis was performed where funnel plot asymmetry was observed. Sensitivity analyses were conducted by excluding individual studies to examine the robustness of pooled estimates. A p-value < 0.05 was considered statistically significant.

Results

Study selection

A total of 2,744 records were initially identified through database searches. After removing 467 duplicates, 2,277 records remained for screening. Based on title and abstract screening, 55 articles were deemed potentially relevant and assessed in full text. Of these, 12 randomized controlled trials (RCTs) met the inclusion criteria for the systematic review, and 9 studies were included in the meta-analysis [23–30]. The study selection process is illustrated in Fig. 1(PRISMA diagram).

Fig. 1.

PRISMA flowchart. The main reasons for exclusion at the full-text stage: Not RCT (n = 18), No quantitative data (n = 12), Wrong population (e.g., type 2 diabetes, n = 9), Duplicate publication (n = 5), Inadequate intervention (n = 3)

Study characteristics

The 12 included RCTs encompassed 485 participants diagnosed with T1DM. Two studies employed a crossover design, while the remaining followed parallel group designs. The trials were conducted between 2009 and 2022 across various countries, including the USA, Italy, India, Iran, Iraq (two each), Germany, Brazil, China, and Austria. Seven trials were conducted in newly diagnosed T1DM patients, while others included participants with longer disease durations. Participant ages ranged from 8.5 (± 2.5) to 44 (± 12.6) years.

Vitamin D formulations varied: cholecalciferol (eight studies), alfacalcidol (two), calcitriol (two), and ergocalciferol (one). Intervention durations ranged from 90 days to two years. Further study characteristics are summarized in Table 1.

Table 1.

Characteristics of included randomized controlled trials (n = 12)

Study (year)	Country	Sample size (n)	Age (mean ± SD)	Pediatric subset?	Vitamin D form	Dose (IU/day) or regimen	Duration	Adverse effects
Li (2009)	China	24	18.4 ± 5.2	Young adult	Alfacalcidol	20 IU/day	6 mo	None
Bizzarri (2010)	Italy	37	12.6 ± 3.4	Yes	Calcitriol	10 IU/day	24 mo	None
Gabbay (2012)	Brazil	38	12.0 ± 3.3	Yes	Cholecalciferol	2,000 IU/day	18 mo	None
Ataie-jafari (2013)	Iran	54	10.1 ± 2.1	Yes	Alfacalcidol	20 IU/day	6 mo	Not reported
Shih (2014)	USA	30	13.2 ± 3.1	Yes	Cholecalciferol	2,857 IU/day	6 mo	1 case (ideation)
Treiber (2015)	Austria	34	10.9 ± 2.2	Yes	Cholecalciferol	3,434 IU/day	12 mo	None
Bogdanou (2017)	Germany	36	12.1 ± 2.3	Yes	Cholecalciferol	4,000 IU/day	3 mo	None
Nafei (2017)	Iraq	30	8.5 ± 2.5	Yes	Cholecalciferol	2,000 IU/day	3 mo	Not reported
Sharma (2017)	India	74	9.5 ± 3.0	Yes	Cholecalciferol	2,000–4,000 IU/day	6 mo	None
Kadhim (2018)	Iraq	40	10.5 ± 3.6	Yes	Calcitriol	2,000 IU/day	3 mo	Not reported
Panjiyar (2018)	India	80	10.6 ± 3.1	Yes	Cholecalciferol	3,000 IU/day	12 mo	None
Nwosu (2022)	USA	38	13.3 ± 2.8	Yes	Ergocalciferol	7143 IU/day	12 mo	None

- Pediatric subset defined as participants ≤ 18 years.

- Dose standardized to IU/day for comparability.

- Li et al. (2009) was retained in primary analysis but excluded in sensitivity analyses for pediatric-specific outcomes.

Adverse effects

Two trials [23, 25] did not report data on adverse events. Among the remaining studies, only Shih et al. [32] reported a single case of suicidal ideation during the non-supplementation period. No other adverse effects were observed in intervention or control groups, supporting the safety of vitamin D supplementation in pediatric T1DM populations.

Risk of bias assessment

Risk of bias was assessed using the Cochrane ROB-2 tool. Four studies [23, 25, 26, 29] were classified as low risk, and eight studies [22, 29–25, 27–29] were categorized as having “some concerns.” The overall quality assessment is presented in Fig. 2.

Fig. 2.

Risk of bias summary (ROB-2 tool)

Meta-analysis

HbA1c

Six studies (n = 275 participants) reported HbA1c outcomes. As shown in Table 2. pooled analysis indicated no statistically significant effect of vitamin D supplementation compared to controls (MD: −1.60; 95% CI: −3.78 to 0.57; I² = 98.07%) (Fig. 3). Sensitivity analysis (Table 3) excluding Treiber et al. [33] showed a significant reduction (MD: −2.37; 95% CI: −4.26 to −0.47). No publication bias was detected (Egger p = 0.20; Begg p = 0.26).

Table 2.

Pooled effect estimates of vitamin D supplementation in T1DM

Outcome	No. of studies	Participants (n)	Mean difference (95% CI)	I² (%)	Significant?
HbA1c (%)	6	275	−1.60 [−3.78, 0.57]	98.07%	No
Fasting C-peptide	4	173	−2.54 [−4.97, −0.11]	97.03%	Yes
Daily insulin (U/kg)	6	278	−0.44 [−0.82, −0.06]	58.64%	Yes
Fasting blood sugar	3	158	−1.44 [−2.67, −0.22]	91.10%	Yes
25(OH)D (ng/mL)	6	331	4.19 [3.26, 5.13]	82.85%	Yes

Fig. 3.

Forest plot (down) and funnel plot (up) for the effect of vitamin D supplementation on HbA1C level in comparison to placebo in T1DM

Table 3.

Sensitivity analysis of pooled outcomes

Outcome	Study removed	Adjusted MD (95% CI)	I² (%)	Effect
HbA1c (%)	Treiber et al. [33]	−2.37 [−4.26, −0.47]	↓	Significant
C-peptide	Nafei et al. [31]	−1.49 [−3.33, 0.35]	↓	Non-sig
Daily insulin	Treiber et al. [33]	−0.33 [−0.69, 0.03]	↓	Non-sig
FBS	Nafei et al. [31]	−0.78 [−1.17, −0.39]	↓	Significant
25(OH)D	Bogdanou et al. [28]	3.85 [3.03, 4.68]	↓	Significant

C-peptide

Four studies (n = 173) were included. Vitamin D supplementation was associated with a reduction in fasting C-peptide levels (MD: −2.54; 95% CI: −4.97 to −0.11; I² = 97.03%) (Fig. 4; Table 2), which may reflect altered β-cell demand or assay variability — however, this finding requires cautious interpretation, as lower C-peptide typically indicates reduced β-cell function. Sensitivity analysis excluding Nafei et al. [31] rendered the effect non-significant (MD: −1.49; 95% CI: −3.33 to 0.35) (Table 3).

Fig. 4.

Forest plot (down) and funnel plot (up) for the effect of vitamin D supplementation on C- peptide level in comparison to placebo in T1DM

Daily insulin requirement

Six studies (n = 278) showed a significant decrease in daily insulin needs (MD: −0.44; 95% CI: −0.82 to −0.06; I² = 58.64%) (Fig. 5; Table 2). Sensitivity analysis excluding Treiber et al. [30] led to a non-significant estimate (MD: −0.33; 95% CI: −0.69 to 0.03) (Table 3).

Fig. 5.

Forest plot (down) and funnel plot (up) for the effect of vitamin D supplementation on daily insulin requirement in comparison to placebo in T1DM

Fasting blood sugar (FBS)

Three studies (n = 158) demonstrated a significant reduction in FBS with vitamin D supplementation (MD: −1.44; 95% CI: −2.67 to −0.22; I² = 91.10%) (Fig. 6; Table 2). The result remained significant after omitting Nafei et al. [31] (Table 3).

Fig. 6.

Forest plot (down) and funnel plot (up) for the effect of vitamin D supplementation on daily insulin requirement in comparison to placebo in T1DM

Serum 25(OH)D levels

Six studies (n = 331) reported significant increases in 25(OH)D concentrations following supplementation (MD: 4.19; 95% CI: 3.26 to 5.13; I² = 82.85%) (Fig. 7; Table 2). This remained robust in sensitivity analysis (Table 3).

Fig. 7.

Forest plot (down) and funnel plot (up) for the effect of vitamin D supplementation on 25(OH)D in comparison to placebo in

GRADE assessment

The certainty of evidence for all primary outcomes was rated as “low” to “very low” using the GRADE framework (Supplementary Table 2), primarily due to high heterogeneity, small sample sizes, and risk of bias.

Discussion

This systematic review and meta-analysis synthesized the available evidence on the efficacy and safety of vitamin D supplementation in pediatric patients with T1DM. Despite accumulating interest in vitamin D as a potential adjunctive therapy for autoimmune diseases, our findings revealed mixed results. While vitamin D supplementation significantly improved certain metabolic markers—including reduced insulin requirements, fasting blood glucose (FBS), and elevated 25(OH)D levels—it did not show a statistically significant effect on hemoglobin A1c (HbA1c). Moreover, an unexpected finding was a decrease in fasting C-peptide levels, which may indicate variability in β-cell function or study design limitations.

Glycemic control and HbA1c

HbA1c is a gold-standard marker for long-term glycemic control, reflecting average glucose concentrations over 2–3 months. In this meta-analysis, the pooled estimate of six randomized controlled trials (RCTs) indicated no significant reduction in HbA1c following vitamin D supplementation. This finding aligns with prior systematic reviews that also reported inconsistent effects of vitamin D on HbA1c in T1DM patients [31, 32]. The high heterogeneity (I² = 98.07%) in our analysis suggests underlying variability in study populations, disease stage (newly diagnosed vs. established T1DM), duration of follow-up, and intervention regimens. Some individual studies, particularly those involving newly diagnosed patients and higher baseline vitamin D deficiency, did report favorable reductions in HbA1c [30, 33, 34]. These results suggest that the glycemic response to vitamin D may be modulated by baseline deficiency status, timing of intervention, and immune activity.

Fasting blood glucose and daily insulin requirement

A more consistent benefit of vitamin D supplementation was observed in short-term glycemic indices. Three RCTs demonstrated a significant reduction in fasting blood glucose (MD: −1.44 mg/dL), and six studies reported a meaningful decrease in daily insulin requirements. These findings may reflect improved insulin sensitivity and reduced systemic inflammation, two mechanisms supported by preclinical and observational research [17, 35–38]. Vitamin D has been shown to enhance insulin receptor expression, regulate calcium flux in β-cells, modulate the T-regulatory cell function, and suppress proinflammatory cytokines (e.g., IL-6, TNF-α), all of which contribute to improved insulin action and glycemic homeostasis. While the reduction in daily insulin requirement (−0.44 U/kg) was statistically significant, its clinical meaningfulness remains uncertain. A reduction of >0.5 U/kg is generally considered clinically relevant in pediatric T1DM; thus, our finding may represent a modest effect.

C-peptide and β-cell function

Unexpectedly, our pooled analysis revealed a reduction in fasting C-peptide levels, which is physiologically counterintuitive and potentially misleading. While vitamin D is hypothesized to preserve β-cell function [26, 30], this finding may be influenced by: inclusion of participants with long-standing T1DM and minimal residual β-cell mass, use of fasting (rather than stimulated) C-peptide measurements, which are less sensitive, and high heterogeneity driven by one outlier study (Nafei et al.) [31]. Sensitivity analysis excluding this study nullified the effect, supporting cautious interpretation. No included trial measured inflammatory or oxidative stress biomarkers; therefore, proposed immunomodulatory mechanisms remain speculative.

Serum 25(OH)D levels and supplementation efficacy

As expected, serum 25(OH)D concentrations significantly increased in the vitamin D groups, confirming bioavailability and compliance. The included studies used various forms of vitamin D (e.g., cholecalciferol, alfacalcidol, calcitriol) and administered doses ranging from physiological (400–1,000 IU/day) to pharmacological (>2,000 IU/day). This diversity underscores a lack of consensus on the optimal type and dosage for immunomodulatory effects. Evidence from observational and interventional studies suggests that maintaining serum 25(OH)D levels above 30 ng/mL may confer immune and metabolic benefits in autoimmune populations, including T1DM [14, 36, 39].

Immunological and genetic mechanisms

Several mechanistic pathways support the immunoregulatory role of vitamin D in T1DM. Vitamin D receptors (VDRs) are expressed in pancreatic islets, T cells, and antigen-presenting cells. VDR activation may shift immune responses from Th1 to Th2 phenotypes, enhance regulatory T cell (Treg) activity, and suppress autoimmune β-cell destruction [10, 11, 40]. Genetic polymorphisms in VDR, vitamin D-binding protein (VDBP), and vitamin D metabolizing enzymes (e.g., CYP2R1, CYP27B1) have been associated with increased T1DM susceptibility and variable responses to vitamin D supplementation [35, 41–51]. Ethnic variation in these polymorphisms may explain differences in outcomes observed across studies from different geographical regions. It should be emphasized that none of the included trials measured biomarkers of inflammation (e.g., IL-6, TNF-α) or oxidative stress (e.g., MDA, GSH); therefore, the proposed immunomodulatory mechanisms remain speculative and require validation in future studies.

Clinical and research implications

Our findings have important clinical implications. While vitamin D supplementation appears safe and beneficial in improving short-term glycemic markers and reducing insulin requirements, its effect on long-term outcomes such as HbA1c and β-cell preservation remains uncertain. These findings support routine vitamin D screening and supplementation in pediatric T1DM patients, particularly in those with suboptimal vitamin D status. However, personalized strategies based on genetic background, disease duration, and baseline vitamin D levels may be needed to optimize therapeutic outcomes.

Strengths and limitations

This review has several strengths: it was conducted in accordance with PRISMA guidelines, included only RCTs, employed robust statistical methods, and explored heterogeneity through sensitivity and subgroup analyses. However, limitations include: small sample sizes, short intervention durations (≤ 12 months in most studies), Unmeasured confounders (e.g., physical activity, BMI), lack of data on dietary vitamin D intake or sun exposure, high heterogeneity not fully explained by subgroup analyses, inclusion of one study with young adults (Li et al.), and low certainty of evidence per GRADE. Language bias may also exist due to English-only inclusion.

Conclusion

This systematic review and meta-analysis suggest that vitamin D supplementation showed potential benefits in reducing insulin requirements and fasting blood glucose in pediatric T1DM populations. However, effects on HbA1c remain inconclusive, and the reduction in C-peptide levels requires cautious interpretation due to high heterogeneity and possible confounding factors. We recommend routine vitamin D screening in pediatric T1DM, but emphasize that supplementation for glycemic control should be individualized and not generalized. Future trials should standardize dosing, include biomarkers of inflammation/β-cell function, and extend follow-up beyond 12 months.

Author contributions

SF and TTZ worked together to create the manuscript title. The first draft was written by SF, TTZ, and LYY, each bringing their distinct insights and expertise to build a strong base for the paper. They then collaborated on an extensive editing phase to improve the clarity and coherence of the final version. All authors affirm their agreement for publication.

Funding

Not applicable.

Data availability

All details are included; for more information, contact the corresponding author via email.

Declarations

Ethics approval

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.