Introduction

Diabetes mellitus is a metabolic disease characterized by chronic hyperglycemia.1 It is highly prevalent among middle-aged and older adults and can lead to damage in multiple organs, including the kidneys, nerves, and blood vessels.2–5 Coronary heart disease (CHD) is a common cardiovascular disorder resulting from atherosclerotic narrowing or occlusion of the epicardial coronary arteries, leading to myocardial ischemia, injury, or necrosis.6,7 It remains one of the leading causes of morbidity and mortality worldwide.8 Globally, the prevalence of CHD continues to rise, with an estimated age-standardized prevalence of approximately 1665 per 100,000 population in 2019, and an estimated global incidence of 197 million cases.8,9 The mortality burden remains substantial, with CHD accounting for nearly 9 million deaths annually.8 Among patients with CHD, the concurrent burden of diabetes is particularly alarming. The global prevalence of diabetes among CHD patients is estimated to be approximately 30–40%, and the presence of diabetes confers a two- to four-fold increased risk of cardiovascular events and mortality in this population.9,10 Diabetes is not only an important risk factor for the occurrence and progression of CHD, but may also further increase the risk of cardiovascular events and adverse outcomes in patients with CHD by accelerating atherosclerosis, impairing endothelial function, aggravating metabolic disorders, and promoting inflammatory responses.11–13 Therefore, early prevention of diabetes and improvement of glycemic status in patients with CHD are of great clinical significance for delaying disease progression and improving long-term prognosis.

At present, the prevention of diabetes in patients with CHD mainly relies on traditional strategies, including dietary control, weight management, increased physical activity, and lifestyle modification.14–17 These interventions have established benefits in diabetes prevention and cardiovascular risk control. However, in clinical practice, some patients with CHD still develop impaired glucose metabolism or even progress to diabetes despite receiving routine management. This suggests that, for this high-risk population with both vascular lesions and metabolic vulnerability, conventional interventions alone may have certain limitations, and new feasible adjunctive preventive strategies are urgently needed.

In recent years, the relationship between elevated homocysteine (Hcy) and glucose metabolism disorders as well as cardiovascular injury has attracted increasing attention.18–20 Evidence suggests that Hcy may participate in the development of hyperglycemia, insulin resistance, and diabetes-related vascular damage through mechanisms such as inducing oxidative stress, impairing endothelial function, promoting inflammation, and interfering with insulin signaling pathways.18,19,21,22 Meanwhile, in patients with CHD, elevated Hcy may also promote the progression of atherosclerosis, vascular calcification, and coronary stenosis, thereby further aggravating the severity of CHD.18,23,24 Therefore, lowering Hcy levels may not only help improve vascular status, but may also have potential value in regulating glucose metabolism.

Folic acid (FA) is an important nutritional intervention for reducing Hcy levels and has the advantages of low cost, wide availability, and favorable safety.25,26 Previous studies have shown that FA supplementation can effectively reduce Hcy levels and may exert cardiovascular protective effects by improving endothelial function and attenuating oxidative stress and inflammatory responses.27–29 In addition, FA has shown potential benefits in blood pressure (BP) control and in reducing the risk of cerebrovascular events such as stroke, as well as lowering the risk of CVD including aortic aneurysms.29–31 Therefore, for patients with CHD, FA use may have multiple potential benefits, including improving vascular function, reducing cardiovascular risk, and regulating metabolic status. However, direct evidence remains lacking regarding whether FA use can further reduce the future risk of diabetes and improve glycemic status in patients with CHD, and its specific effect remains unclear.

Based on the importance of diabetes prevention in patients with CHD and the potential role of FA in reducing Hcy levels and improving vascular function, this study aims to use multicenter clinical cohort data and propensity score matching (PSM) to control for potential confounding factors. The associations of FA use and different usage patterns with the risk of diabetes and improvement in glycemic status among patients with CHD were systematically evaluated, and the cumulative effects of long‑term FA use were further explored.

Materials and Methods Study Population

The study initially included a total of 8955 patients with CHD from four medical centers. Of these, 876 patients did not complete valid follow-up, leaving 8079 patients who completed regular follow-up. To control for potential confounding factors, further strict exclusion criteria were applied, sequentially removing: (1) patients with diabetes at baseline; (2) patients taking any weight‑loss medications or drugs affecting glucose metabolism; (3) patients with severe liver or kidney dysfunction (eg, eGFR < 30 mL/min/1.73 m2 or liver cirrhosis); (4) patients using any medications affecting FA metabolism, such as methotrexate or antiepileptic drugs; and (5) patients with malignancy. After these rigorous exclusion criteria were applied, a total of 6087 patients met the study criteria (Figure 1).

Figure 1 Screening and matching process of the study population.

The study was conducted in strict accordance with the ethical principles of the Declaration of Helsinki and received unanimous approval from the ethics committees of the four medical centers: Changzhi People’s Hospital (CZX20180213), Sichuan Provincial People’s Hospital (SCS20240512), the Sixth Affiliated Hospital of Kunming Medical University (XKY20200806), and Yunyang County People’s Hospital (Yy20200419). Before the study commenced, all patients were informed of the overall process and details of the study, and each provided written informed consent agreeing to participate.

Data Collection and Definitions

Covariate information was collected through multiple sources, including medical visit records, electronic medical records, and health insurance data from each center, as well as regular face-to-face interviews or telephone follow-ups. Anthropometric measurements [height, weight, body mass index (BMI)], systolic blood pressure (SBP), diastolic blood pressure (DBP), and fasting blood glucose (FBG) levels were obtained by trained nurses. Laboratory parameters, including total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), glycosylated hemoglobin (HbA1c), and Hcy, were measured by the central or local clinical laboratories. Information on comorbid conditions, such as hypertension and hyperlipidemia, was also collected, with detailed diagnostic criteria provided in Supplementary Methods 1. In addition, data on medication use were recorded, including lipid-lowering drugs, antiplatelet drugs, beta-blockers, and angiotensin-converting enzyme inhibitors (ACEIs)/ angiotensin receptor blockers (ARBs) (Table S1).

Definition of Medication Use and Cumulative Dose

According to the definitions adopted in previous medication-use studies, patients who had continuously used FA for at least six months before baseline enrollment were defined as folic acid users; otherwise, they were classified as non-users.32,33 Information on FA use was obtained through patient interviews, electronic medical records, or prescription refill records. The cumulative dose of FA was calculated as the daily dose multiplied by the actual number of days of use, expressed in milligram (mg).

Study Outcome

The primary outcome of this study was the occurrence of new-onset diabetes during follow-up in patients with CHD. The diagnosis of diabetes was made according to current clinical guidelines, based on any of the following criteria: (1) presence of characteristic symptoms (eg, polydipsia, polyuria, polyphagia, or unexplained weight loss) accompanied by a random venous plasma glucose level ≥ 11.1 mmol/L; (2) a 2-hour venous plasma glucose level ≥ 11.1 mmol/L during a 75 g oral glucose tolerance test; or (3) a HbA1c level ≥ 6.5%.

PSM to Control for Confounding Factors

To investigate the effect of FA use on new-onset diabetes in patients with CHD, this study employed a PSM method to control for potential confounding factors. First, propensity scores representing the probability of FA use were calculated for each patient using a logistic regression model. To minimize potential confounding, the matching model included the following variables: sex, age, BMI, smoking status, drinking status, SBP, DBP, LDL-C, HDL-C, antiplatelet drugs, and ACEIs/ARBs. We performed propensity‑score matching using a nearest‑neighbour matching algorithm on the logit scale, matching FA users to non‑users at a 1:4 ratio with a caliper of 0.2 standard deviations of the logit of the propensity score. Matching was performed without replacement. Ties were handled by random order selection. The standardized mean difference (SMD) was used to assess balance between the two groups before and after matching, with an SMD < 0.10 indicating good balance. We assessed the distribution of propensity scores between the two groups and confirmed adequate overlap, supporting the validity of the matching procedure. Prior to matching, the cohort included 492 FA users and 5595 non-users. After 1:4 nearest‑neighbour matching without replacement, 488 FA users were successfully matched to 1952 non-users. A total of 4 FA users were excluded because no suitable match could be found within the specified caliper distance.

Statistical analysis

Patients with CHD were divided into a folic acid user group and a non‑user group based on their FA use. A stepwise‑adjusted multivariable Cox regression model was used to analyze the association between FA use and the risk of diabetes. To compare the effects of different daily doses on diabetes risk, patients were further categorized into three groups: non‑user group, 0.4 mg/day user group, and 0.8 mg/day user group, and Cox regression analyses were also performed for these three groups. Kaplan‑Meier (KM) cumulative risk curves were plotted to visually compare the cumulative incidence of diabetes during follow‑up between FA users and non‑users, as well as among the different dose groups. Furthermore, based on four repeated measurements of HbA1c, we plotted the trajectory of HbA1c changes across groups and compared the differences in HbA1c change over time.

Considering the long‑term cumulative effect of medication use, we conducted a cumulative effect analysis using restricted cubic splines (RCS) to evaluate the dose‑response relationship between cumulative FA dose and diabetes risk, and to identify potential threshold effects. Based on the identified threshold, a two‑piecewise segmented regression analysis was further performed. Finally, subgroup analyses and sensitivity analyses were conducted to further confirm the robustness of the results.

All statistical analyses were performed using R software (version 4.2.2). A two‑sided P value < 0.05 was considered statistically significant.

Results Study Population Matching

Among the 6087 patients with CHD who met the study criteria, 492 were identified as receiving FA therapy, while the remaining 5595 were not receiving FA. After performing rigorous 1:4 propensity score matching PSM, a total of 2440 patients were included in the final analysis, consisting of 488 FA users and 1952 non-users. During a median follow-up of 3.35 years (interquartile range, 2.50–4.68), 510 incident diabetes cases were recorded. Figure 1 illustrates the stepwise selection process of the study population.

Baseline Characteristics of the Study Population

Table 1 presents the comparison of baseline characteristics between non-users and users before and after matching. Prior to matching, the user group had significantly lower BMI, SBP, and DBP compared to the non-user group; their glycemic parameters (FBG and HbA1c) were also significantly lower. In addition, Hcy levels were lower in the user group. However, the utilization rates of antiplatelet drugs and ACEIs/ARBs were relatively higher among users. To address these potential confounders, we performed matching on the major covariates. After matching, the balance of variables between the two groups was markedly improved, with all SMDs below 0.1. Furthermore, Figures S1 and S2 visually illustrate the balance of variables before and after matching.

Table 1 Baseline Characteristics Before and After PSM

Effect of FA Use (vs Non-Use) on Diabetes Risk in Patients with CHD

To compare the effect of FA use versus non-use on the risk of diabetes in patients with CHD, we first analyzed the incidence of diabetes during follow-up between the two groups. Before PSM, the cumulative incidence of diabetes in the non-user group was significantly higher than that in the user group (26.62% vs 15.45%) (Figure 2A). After matching, this difference persisted, with the non-user group having an incidence of 22.23%, still higher than the 15.57% in the user group (Figure 2B). Moreover, Table S2 presents the detailed data for each matched group and the Incidence per 100 person-years observed during follow-up. Multivariable-adjusted Cox regression models were then applied. Analysis of the pre-matching data showed that, in the unadjusted Model 1, the FA user group had a 38.6% lower risk of diabetes compared with the non-user group [hazard ratio (HR) = 0.614, 95% confidence interval (CI): 0.491–0.769, P<0.001]. After full adjustment for all covariates, the risk remained significantly reduced by 35.1% (HR = 0.649, 95% CI: 0.518–0.814, P<0.001) (Table 2). Analysis of the post-matching data similarly demonstrated that the FA user group had a 25.1% lower risk of diabetes compared with the non-user group (HR = 0.749, 95% CI: 0.586–0.957, P=0.021) (Table 3).

Table 2 FA Supplementation and Diabetes Risk (Pre-Matching)

Table 3 FA Supplementation and Diabetes Risk (Post-Matching)

Figure 2 Incidence of diabetes between the user group and the non-user group. (A) Pre-matching cohort; (B) Post-matching cohort.

Furthermore, the KM cumulative risk curves yielded consistent results: both before and after matching, the cumulative risk of diabetes during follow-up was significantly lower in the FA user group than in the non-user group (Figure 3).

Figure 3 Kaplan-Meier cumulative risk curves between the user group and the non-user group. (A) Pre-matching cohort; (B) Post-matching cohort.

Effect of Different Daily Doses of FA on Diabetes Risk in Patients with CHD

Considering that the daily dose of FA may vary among patients and whether this difference leads to different effects on diabetes risk remains unclear, we further subdivided FA users into a 0.4 mg/day group and a 0.8 mg/day group.

First, we compared the cumulative incidence of diabetes among the non‑user group, the 0.4 mg/day group, and the 0.8 mg/day group. Before matching, the cumulative incidence of diabetes in the 0.8 mg/day group was 14.98%, which was significantly lower than that in the non‑user group (26.62%) and the 0.4 mg/day group (18.05%) (Figure 4A). After matching, the results remained consistent, with the 0.8 mg/day group still having the lowest incidence at 13.78% (Figure 4B). Similarly, Table S3 presents the detailed data for each of the three matched groups and the Incidence per 100 person-years observed during follow-up.

Figure 4 Incidence of diabetes among the three groups: the non-user group and the groups receiving daily doses of 0.4 mg/day and 0.8 mg/day. (A) Pre-matching cohort; (B) Post-matching cohort.

Second, Cox regression analysis was performed on the three groups. Analysis of pre‑matching data showed that, compared with the non‑user group, the 0.4 mg/day group and the 0.8 mg/day group had reductions in diabetes risk of 32.0% (HR = 0.680, 95% CI: 0.491–0.944, P=0.021) and 37.4% (HR = 0.626, 95% CI: 0.462–0.849, P=0.003), respectively (Table 4). Results from post-matching data were consistent: the 0.8 mg/day group had a 36.3% lower risk of diabetes compared with the non-user group (HR = 0.637, 95% CI: 0.453–0.897, P=0.010), and this dose group exhibited the lowest HR relative to non-users (Table 5).

Table 4 Daily FA Dose and Diabetes Risk (Pre-Matching)

Table 5 Daily FA Dose and Diabetes Risk (Post-Matching)

Finally, the KM cumulative risk curves yielded consistent conclusions: both before and after matching, the cumulative risk of diabetes in the 0.8 mg/day group was the lowest and significantly lower than that in the FA non‑user group (Figure 5).

Figure 5 Kaplan-Meier cumulative risk curves for the three groups: the non-user group and the groups receiving daily doses of 0.4 mg/day and 0.8 mg/day. (A) Pre-matching cohort; (B) Post-matching cohort.

Effect of FA Use and Dosage on Long-Term Glycemic Status in Patients with CHD

To further evaluate the effect of FA use and its dosage on long-term glycemic status in patients with CHD, we analyzed repeated measurements of HbA1c collected during follow-up.

First, we compared the trajectory of HbA1c changes between the FA user group and the non-user group. The results showed that HbA1c levels in the FA user group exhibited a gradual decreasing trend, whereas those in the non-user group showed fluctuations with a mild increasing tendency (Figure 6A). Subsequently, we further compared the effect of different daily doses on long-term glycemic status. The results demonstrated that the non-user group still showed a mild increasing trend with fluctuations; the 0.4 mg/day group exhibited a mild decreasing trend with some fluctuations; while the 0.8 mg/day group showed a clear and stable decreasing trend in HbA1c levels (Figure 6B). These findings indirectly suggest that a daily dose of 0.8 mg of FA may be more beneficial for long-term glycemic control in patients with CHD.

Figure 6 Dynamic changes in HbA1c levels among different groups during follow-up. (A) Grouping by use or non-use; (B) Grouping by daily dose.

Effect of Cumulative FA Dose on Reducing Diabetes Risk in Patients with CHD

To further quantify the combined effect of daily dose and duration of FA use on diabetes risk in patients with CHD, we calculated the cumulative FA dose. Cox regression analysis showed that each 10 mg increase in cumulative FA dose was associated with a 4.8% reduction in diabetes risk (HR = 0.952, 95% CI: 0.931–0.974, P<0.001) (Table 6). RCS analysis was then used to evaluate the dose‑response relationship between cumulative FA dose and diabetes risk. The results revealed a significant progressive decreasing effect, with a particularly notable decline in diabetes risk when the cumulative dose exceeded 140 mg (Figure 7). Based on this finding, we performed a two‑piecewise comparative analysis using 140 mg as the threshold. The results showed that, compared with patients with a cumulative dose ≤ 140 mg, those with a cumulative dose > 140 mg had a 52.2% lower risk of diabetes (HR = 0.478, 95% CI: 0.351–0.650, P<0.001) (Table 6). These findings further indicate that the risk‑reducing effect of FA on diabetes becomes apparent once a certain cumulative dose is achieved in patients with CHD.

Table 6 Threshold Effect of Cumulative FA Dosage

Figure 7 Dose-response relationship and threshold effect between cumulative dose of folic acid and diabetes risk.

Subgroup and Sensitivity Analyses

First, to account for individual differences and variations in disease status, we performed subgroup analyses stratified by sex, age, BMI, smoking status, drinking status, hypertension, and hyperlipidemia. The cumulative dose of FA was significantly associated with a reduced risk of diabetes across all subgroups. However, formal interaction tests revealed no significant effect modification by any of the stratification factors (all p for interaction > 0.05, Table 7). Although the protective association appeared somewhat more pronounced in patients with hyperlipidemia and those without hypertension, the interaction p-values were 0.115 and 0.229, respectively, indicating that these observed differences did not reach statistical significance (Table 7).

Table 7 Subgroup Analysis of Cumulative FA Dose and Diabetes Risk

On this basis, we further conducted a series of sensitivity analyses to test the robustness of the findings. To address potential reverse causality, we re-analyzed the data after excluding patients with less than one year of follow-up, and the results remained largely consistent with the primary analysis (Tables S4–S7). Furthermore, considering that older patients may have worse comorbidities and glycemic control, we excluded those aged 75 years or older and re-analyzed the data; the results again remained stable (Tables S8–S11). Finally, to further mitigate potential confounding from other factors, we performed enhanced matching on top of the initial matching by additionally matching for FBG, HbA1c, lipid-lowering drugs, baseline Hcy, study center, renal function, and the clinical indication for FA treatment, and re-analyzed the data; the results did not change significantly (Tables S12 and S13).

In summary, the results of these subgroup and sensitivity analyses further confirm the robustness of our study conclusions.

Discussion

The findings of this multicenter study demonstrate that FA use is significantly associated with a lower risk of new‑onset diabetes and more favorable long‑term glycemic control in patients with CHD. Specifically, the incidence of diabetes was significantly lower among FA users than among non‑users, and Cox regression analyses further showed that FA use was associated with a 25–38% reduction in diabetes risk compared with non‑use. This association also exhibited a clear dose‑response pattern. Relative to non‑users, the 0.8 mg/day group showed numerically more favorable estimates for reduced diabetes risk, reduced HbA1c levels, and improved glycemic status. Consistently, cumulative dose analysis showed that each 10 mg increase in FA exposure was associated with a 4.8% reduction in diabetes risk, and the risk reduction reached 52.2% when cumulative exposure exceeded 140 mg. Overall, these findings suggest that FA may have potential value as a preventive strategy for reducing diabetes risk and improving glycemic control in patients with CHD.

Previous studies have increasingly highlighted a close relationship between hyperhomocysteinemia and disordered glucose metabolism.19,34–36 At the mechanistic level, Hcy may directly impair pancreatic β-cell function and exacerbate peripheral insulin resistance by inducing oxidative stress, endoplasmic reticulum stress, and inflammatory responses, thereby providing biological plausibility for its diabetogenic effects.19,34,37 Clinical evidence also supports this association.35,36,38 For example, studies in pregnant women have shown that elevated Hcy levels during pregnancy are associated with hyperglycemia and may serve as an independent risk factor for gestational diabetes.35 Mendelian randomization analyses have further suggested a potential causal relationship, emphasizing the importance of maintaining Hcy within the normal range during pregnancy.36 In addition, a cross-sectional case-control study reported a significantly higher prevalence of diabetes among individuals with hyperhomocysteinemia than among those with normal Hcy levels.38 This relationship is also particularly relevant in patients with CHD, because Hcy also promotes endothelial dysfunction and atherosclerosis, which may further worsen cardiovascular and metabolic outcomes.18,39,40 Therefore, existing evidence supports a critical role of Hcy in the intersection between glucose metabolism and CVD.

Given the pathogenic role of Hcy in glucose dysregulation and cardiovascular injury, FA may exert beneficial metabolic effects by serving as a key coenzyme in Hcy metabolism and effectively lowering plasma Hcy levels. Emerging evidence supports this potential role.41,42 For example, a randomized controlled trial in individuals with prediabetes showed that FA supplementation improved insulin sensitivity and delayed diabetes progression,41 while another study reported that combined supplementation with FA, vitamin B6, and vitamin B12 reduced Hcy levels and improved glycemic control in patients with type 2 diabetes.42 In addition to these metabolic effects, FA has long been recognized as a safe and inexpensive B vitamin with established cardiovascular benefits, including stroke prevention in patients with hypertension and vascular protection in patients with CHD.29,43,44 However, most existing studies have focused on pregnant women, individuals with prediabetes, or patients with established diabetes, and the potential preventive association between FA use and new-onset diabetes in patients with CHD remains insufficiently investigated.

To address this evidence gap, the present multicenter propensity score-matched study comprehensively examined the associations of FA use, including different daily doses and cumulative dose levels, with new-onset diabetes risk and long-term glycemic changes in patients with CHD. To our knowledge, this is the first study to systematically evaluate this issue in a CHD population. Our findings extend the potential clinical value of FA from cardiovascular protection to metabolic improvement and suggest dose‑related patterns that may inform future research. Specifically, the 0.8 mg daily dose and cumulative doses exceeding 140 mg showed more favorable associations with the outcomes compared with non‑use, providing preliminary quantifiable references for future clinical practice and research. However, these observations should be interpreted cautiously, as direct comparisons between dose groups were not formally tested, and the optimal dose remains to be established in future randomized controlled trials.

In the dose-stratified analysis, the 0.8 mg/day group showed a clear and stable downward trend in HbA1c, whereas the 0.4 mg/day group showed only a modest and fluctuating decrease. This dose-dependent difference may be partly explained by the stronger Hcy-lowering effect of the higher dose, which may more effectively reduce oxidative stress-related injury to pancreatic β-cells.34,37,45,46 The cumulative dose findings further complement this observation. The marked reduction in diabetes risk after cumulative dose exceeded 140 mg suggests that the metabolic benefits of FA may require sufficient duration and intensity of cumulative dosing before becoming clinically evident. This potential threshold effect is consistent with the characteristics of many interventions for chronic diseases and supports the clinical rationale for long-term and regular FA use in patients with CHD.

The beneficial effects of FA on glucose metabolism observed in this study may involve several biological mechanisms. First, FA reduces Hcy levels, thereby attenuating Hcy-induced toxicity to pancreatic β-cells and its interference with insulin signaling pathways.19,34,41 Second, FA may exert antioxidant effects by inhibiting hyperglycemia-induced reactive oxygen species production and protecting islet cell function.47–49 Third, FA may indirectly enhance glucose uptake and utilization by improving endothelial function and increasing peripheral tissue perfusion.50,51 Additionally, FA may upregulate DHFR, recouple eNOS, and enhance NO production, thereby improving peripheral tissue perfusion and insulin sensitivity, which in turn facilitates glucose uptake and reduces blood glucose.52,53 Finally, FA participates in one-carbon metabolism together with vitamin B12, thereby influencing DNA methylation and gene expression.54–56 Through this pathway, FA may regulate the expression of genes involved in glucose metabolism at the epigenetic level.37,56,57 Although these mechanisms are biologically plausible and supported by basic research, further experimental studies are needed to confirm their roles in CHD patients.

This study systematically evaluates the association between FA use and diabetes risk, as well as glycemic improvement, in patients with CHD, thereby translating mechanistic hypotheses into clinical evidence. The major strengths of this study include its multicenter large-sample design, the use of PSM to reduce confounding bias, detailed analyses of daily dose and cumulative dose, and dynamic assessment of repeated HbA1c measurements. However, several limitations should be acknowledged. First, our prevalent‑user design may introduce selection bias and confounding by prior treatment or disease severity. Although we used PSM and multivariable adjustment, residual confounding and unmeasured factors (eg, dietary folate intake, physical activity) may still affect the results. Second, the definition of FA exposure relied on comprehensive data from electronic prescription records, regular follow-up medication documentation, and patient self-reports; while this multi-source approach reduced misclassification, recall or recording errors may still persist. Third, all baseline covariates were measured after folic acid initiation, which aligns with the prevalent-user design but may not fully capture pre-exposure status. In addition, although diabetes was diagnosed according to clinical criteria, some patients with early-stage asymptomatic diabetes may have remained undetected. Finally, because the study population consisted exclusively of Chinese patients with CHD, the generalizability of these findings to other ethnic groups or populations requires further validation.

Conclusion

This study provides the first systematic evaluation of the association between FA use and the risk of new-onset diabetes as well as long-term glycemic changes in patients with CHD. Our findings suggest that FA use is significantly associated with a reduced risk of diabetes in CHD patients, with observed dose-related patterns suggesting a more favorable association for the 0.8 mg daily dose and cumulative doses exceeding 140 mg. These findings expand the clinical scope of FA, implying that, in addition to its cardiovascular protective effects, it may represent a potentially beneficial adjunctive strategy for primary diabetes prevention in patients with CHD. However, given the observational nature of this study, causality cannot be definitively established, and further large‑scale, prospective randomized controlled trials are warranted to confirm these findings and to determine the optimal dose.

Institutional Review Board Statement

This study was conducted in strict accordance with the ethical principles of the Declaration of Helsinki and received unanimous approval from the ethics committees of the four medical centers: Changzhi People’s Hospital (CZX20180213), Sichuan Provincial People’s Hospital (SCS20240512), the Sixth Affiliated Hospital of Kunming Medical University (XKY20200806), and Yunyang County People’s Hospital (Yy20200419). Before the study commenced, all patients were informed of the overall process and details of the study, and each provided written informed consent agreeing to participate.

Data Sharing Statement

The data analyzed in this study are available from the corresponding author, Shengqin Li, upon reasonable request.

Author Contributions

Shuaiwei Song: Conceptualization, Methodology, Investigation, Data Curation, Formal Analysis, Writing – Original Draft, Writing – review and editing. Xintian Cai: Methodology, Data Curation, Investigation, Formal Analysis, Validation Writing – Original Draft. Xiyang Li: Methodology, Investigation, Formal Analysis, Validation Writing – Original Draft. Penghui Cui: Methodology, Investigation, Data Curation, Formal Analysis, Writing – review and editing. Ziliang Zou: Data Curation, Investigation, Writing – review and editing. Yazhe Liu: Methodology, Data Curation, Writing – review and editing. Kaiyu Wang: Conceptualization, Methodology, Investigation, Supervision, Data Curation, Writing – review and editing. Shengqin Li: Conceptualization, Methodology, Investigation, Supervision, Data Curation, Formal Analysis, Writing – Original Draft, Writing – review and editing. All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors report no conflicts of interest.

References

1. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care. 2013;36(Suppl 1):S67–16. doi:10.2337/dc13-S067

2. NCD Risk Factor Collaboration (NCD-RisC). Worldwide trends in diabetes since 1980: a pooled analysis of 751 population-based studies with 4.4 million participants. Lancet. 2016;387(10027):1513–1530. doi:10.1016/S0140-6736(16)00618-8

3. Li Y, Liu Y, Liu S, et al. Diabetic vascular diseases: molecular mechanisms and therapeutic strategies. Signal Transduct Target Ther. 2023;8(1):152. doi:10.1038/s41392-023-01400-z

4. Forbes JM, Cooper ME. Mechanisms of diabetic complications. Physiol Rev. 2013;93(1):137–188. doi:10.1152/physrev.00045.2011

5. Demir S, Nawroth PP, Herzig S, et al. Emerging targets in type 2 diabetes and diabetic complications. Adv Sci. 2021;8(18):e2100275. doi:10.1002/advs.202100275

6. Qiu ZX, Wang RY, Zhang Y, et al. Mechanisms, precision therapies, and technological frontiers in coronary atherosclerosis: a comprehensive review. Acta Pharmacol Sin. 2026;47(5):1103–1116. doi:10.1038/s41401-025-01729-x

7. Zaman S, Wasfy JH, Kapil V, et al. The Lancet Commission on rethinking coronary artery disease: moving from ischaemia to atheroma. Lancet. 2025;405(10486):1264–1312. doi:10.1016/S0140-6736(25)00055-8

8. Roth GA, Mensah GA, Johnson CO, et al. Global Burden of Cardiovascular Diseases and Risk Factors, 1990-2019: update From the GBD 2019 Study. J Am Coll Cardiol. 2020;76(25):2982–3021. Erratum in: J Am Coll Cardiol. 2021;77(15):1958-1959. doi: 10.1016/j.jacc.2021.02.039. doi:10.1016/j.jacc.2020.11.010

9. GBD 2019 Diseases and Injuries Collaborators. Global burden of 369 diseases and injuries in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2020;396(10258):1204–1222. doi:10.1016/S0140-6736(20)30925-9

10. Dörr M. Big data and cardiovascular risk-insights into obesity, diabetes, and coronary heart disease. Herz. 2025;50(4):260–267. English. doi:10.1007/s00059-025-05323-z

11. Khan AW, Jandeleit-Dahm KAM. Atherosclerosis in diabetes mellitus: novel mechanisms and mechanism-based therapeutic approaches. Nat Rev Cardiol. 2025;22(7):482–496. doi:10.1038/s41569-024-01115-w

12. Zhong M, Wu X, Liu Y, et al. Molecular mechanisms underlying the comorbidity of type 2 diabetes mellitus and coronary artery disease: from insulin resistance and inflammation to endothelial dysfunction and therapeutic implications. Cardiovasc Drugs Ther. 2026. doi:10.1007/s10557-026-07882-7

13. Gyldenkerne C, Maeng M, Kjøller-Hansen L, et al. Coronary artery lesion lipid content and plaque burden in diabetic and nondiabetic patients: PROSPECT II. Circulation. 2023;147(6):469–481. doi:10.1161/CIRCULATIONAHA.122.061983

14. Demissie GD, Birungi J, Shrestha A, et al. The effectiveness of lifestyle interventions in reducing cardiovascular risk and risk factors in people with prediabetes: a systematic review and meta-analysis. Nutr Metab Cardiovasc Dis. 2025;35(10):104130. doi:10.1016/j.numecd.2025.104130

15. Dansinger ML, Gleason JA, Maddalena J, et al. Lifestyle modification in prediabetes and diabetes: a large population analysis. Nutrients. 2025;17(8):1333. doi:10.3390/nu17081333

16. Sood S, Sylivris AL, Sualeheen A, et al. The effects of Mediterranean diet with and without exercise on body composition in adults with chronic disease: a systematic review and meta-analysis of clinical trials. Clin Nutr. 2025;51:222–239. doi:10.1016/j.clnu.2025.06.019

17. Song S, Cai X, Xu T, et al. Inflammation mediates the relationship between serum uric acid and new-onset diabetes risk in patients with coronary heart disease: results from a multicenter cohort study. Diabetes Metab Syndr Obes. 2026;19:602643. doi:10.2147/DMSO.S602643

18. Jakubowski H, Witucki Ł. Homocysteine Metabolites, Endothelial Dysfunction, and Cardiovascular Disease. Int J Mol Sci. 2025;26(2):746. doi:10.3390/ijms26020746

19. Ding T, Wen B, Chen J, et al. Excess homocysteine inhibits pancreatic β-cell secretory function by repressing Zbtb20 expression. Mol Cell Endocrinol. 2024;586:112195. doi:10.1016/j.mce.2024.112195

20. Fu D, Gong W, Bao X, et al. Homocysteine levels are associated with diabetes mellitus in Chinese with H-type hypertension. Nutr Res Pract. 2024;18(4):511–522. doi:10.4162/nrp.2024.18.4.511

21. Agostini D, Bartolacci A, Rotondo R, et al. Homocysteine, nutrition, and gut microbiota: a comprehensive review of current evidence and insights. Nutrients. 2025;17(8):1325. doi:10.3390/nu17081325

22. Sheikh AB, Jathar SM, Tammara V, et al. Characterization of S-homocysteinylation of human insulin and its implications in diabetes. Protein J. 2025;44(5):598–610. doi:10.1007/s10930-025-10282-9

23. Karger AB, Steffen BT, Nomura SO, et al. Association between homocysteine and vascular calcification incidence, prevalence, and progression in the Mesa cohort. J Am Heart Assoc. 2020;9(3):e013934. doi:10.1161/JAHA.119.013934

24. Zheng LC, Liu F, Zheng PM, et al. Serum sST2, IL-33, and Hcy expression in older adults patients with myocardial infarction and their predictive value for MACE. Clin Interv Aging. 2025;20:495–504. doi:10.2147/CIA.S519437

25. Turck D, Bohn T, Castenmiller J, et al; EFSA Panel on Nutrition, Novel Foods and Food Allergens (NDA Panel). Scientific opinion on the tolerable upper intake level for folate EFSA J. 2023;21(11):e08353 doi:10.2903/j.efsa.2023.8353

26. Rodrigues VB, Silva END, Santos MLP. Cost-effectiveness of mandatory folic acid fortification of flours in prevention of neural tube defects: a systematic review. PLoS One. 2021;16(10):e0258488. doi:10.1371/journal.pone.0258488

27. Title LM, Cummings PM, Giddens K, et al. Effect of folic acid and antioxidant vitamins on endothelial dysfunction in patients with coronary artery disease. J Am Coll Cardiol. 2000;36(3):758–765. doi:10.1016/s0735-1097(00)00809-3

28. Mokgalaboni K, Mashaba GR, Phoswa WN, et al. Folic acid supplementation on inflammation and homocysteine in type 2 diabetes mellitus: systematic review and meta-analysis of randomized controlled trials. Nutr Diabetes. 2024;14(1):22. doi:10.1038/s41387-024-00282-6

29. Liu Y, Geng T, Wan Z, et al. Associations of serum folate and Vitamin B12 levels with cardiovascular disease mortality among patients with type 2 diabetes. JAMA Network Open. 2022;5(1):e2146124. Erratum in: JAMA Netw Open. 2022;5(3)e224936doi: 10.1001/jamanetworkopen.2022.4936. doi:10.1001/jamanetworkopen.2021.46124

30. Wang X, Qin X, Demirtas H, et al. Efficacy of folic acid supplementation in stroke prevention: a meta-analysis. Lancet. 2007;369(9576):1876–1882. doi:10.1016/S0140-6736(07)60854-X

31. Asbaghi O, Salehpour S, Rezaei Kelishadi M, et al. Folic acid supplementation and blood pressure: a GRADE-assessed systematic review and dose-response meta-analysis of 41,633 participants. Crit Rev Food Sci Nutr. 2023;63(13):1846–1861. doi:10.1080/10408398.2021.1968787

32. Song S, Cai X, Hu J, et al. Effectiveness of spironolactone in reducing osteoporosis and future fracture risk in middle-aged and elderly hypertensive patients. Drug Des Devel Ther. 2024;18:2215–2225. doi:10.2147/DDDT.S466904

33. Carbone LD, Cross JD, Raza SH, et al. Fracture risk in men with congestive heart failure risk reduction with spironolactone. J Am Coll Cardiol. 2008;52(2):135–138. doi:10.1016/j.jacc.2008.03.039

34. Patterson S, Flatt PR, McClenaghan NH. Major metabolic homocysteine-derivative, homocysteine thiolactone, exerts changes in pancreatic beta-cell glucose-sensing, cellular signal transduction and integrity. Arch Biochem Biophys. 2007;461(2):287–293. doi:10.1016/j.abb.2007.02.011

35. Liu W, Huang Z, Tang S, et al. Changes of serum sex hormone-binding globulin, homocysteine, and hypersensitive CRP levels during pregnancy and their relationship with gestational diabetes mellitus. Gynecol Obstet Invest. 2021;86(1–2):193–199. doi:10.1159/000515085

36. Zhan X, Chen H, Lin Y, et al. Investigating the role of homocysteine and B vitamins in gestational diabetes mellitus: mendelian randomization study. Nutr J. 2025;24(1):167. doi:10.1186/s12937-025-01234-1

37. Zhu J, Saikia G, Zhang X, et al. One-carbon metabolism nutrients, genetic variation, and diabetes mellitus. Diabetes Metab J. 2024;48(2):170–183. doi:10.4093/dmj.2023.0272

38. Yang X, Lan Q, Wang Q, et al. Association of homocysteine and uric acid with type 2 diabetes mellitus: a case-control study. Sci Rep. 2025;15(1):1914. doi:10.1038/s41598-025-85812-7

39. Esse R, Barroso M, Tavares de Almeida I, et al. The contribution of homocysteine metabolism disruption to endothelial dysfunction: state-of-the-art. Int J Mol Sci. 2019;20(4):867. doi:10.3390/ijms20040867

40. Tian W, Ju J, Guan B, et al. Role of hyperhomocysteinemia in atherosclerosis: from bench to bedside. Ann Med. 2025;57(1):2457527. doi:10.1080/07853890.2025.2457527

41. Lind MV, Lauritzen L, Kristensen M, et al. Effect of folate supplementation on insulin sensitivity and type 2 diabetes: a meta-analysis of randomized controlled trials. Am J Clin Nutr. 2019;109(1):29–42. doi:10.1093/ajcn/nqy234

42. Sudchada P, Saokaew S, Sridetch S, et al. Effect of folic acid supplementation on plasma total homocysteine levels and glycemic control in patients with type 2 diabetes: a systematic review and meta-analysis. Diabet Res Clin Pract. 2012;98(1):151–158. doi:10.1016/j.diabres.2012.05.027

43. An P, Wan S, Luo Y, et al. Micronutrient supplementation to reduce cardiovascular risk. J Am Coll Cardiol. 2022;80(24):2269–2285. doi:10.1016/j.jacc.2022.09.048)

44. Zhang N, Zhou Z, Chi X, et al. Folic acid supplementation for stroke prevention: a systematic review and meta-analysis of 21 randomized clinical trials worldwide. Clin Nutr. 2024;43(7):1706–1716. doi:10.1016/j.clnu.2024.05.034

45. Huang X, Bao H, Ding C, et al. Optimal folic acid dosage in lowering homocysteine: precision folic acid trial to lower homocysteine (PFAT-Hcy). Eur J Nutr. 2024;63(5):1513–1528. doi:10.1007/s00394-024-03344-8

46. Homocysteine Lowering Trialists’ Collaboration. Dose-dependent effects of folic acid on blood concentrations of homocysteine: a meta-analysis of the randomized trials. Am J Clin Nutr. 2005;82(4):806–812. doi:10.1093/ajcn/82.4.806

47. Hsu HC, Chiou JF, Wang YH, et al. Folate deficiency triggers an oxidative-nitrosative stress-mediated apoptotic cell death and impedes insulin biosynthesis in RINm5F pancreatic islet β-cells: relevant to the pathogenesis of diabetes. PLoS One. 2013;8(11):e77931. doi:10.1371/journal.pone.0077931

48. Hunter-Lavin C, Hudson PR, Mukherjee S, et al. Folate supplementation reduces serum hsp70 levels in patients with type 2 diabetes. Cell Stress Chaperones. 2004;9(4):344–349. doi:10.1379/csc-28r.1

49. Bhattacharjee A, Prasad SK, Pal S, et al. Synergistic protective effect of folic acid and vitamin B12 against nicotine-induced oxidative stress and apoptosis in pancreatic islets of the rat. Pharm Biol. 2016;54(3):433–444. doi:10.3109/13880209.2015.1043561

50. Wang Z, Xing W, Song Y, et al. Folic acid has a protective effect on retinal vascular endothelial cells against high glucose. Molecules. 2018;23(9):2326. doi:10.3390/molecules23092326

51. Hong L, Lou X, Zha Y, et al. Folic acid alleviates endothelial dysfunction caused by T2DM-induced ferroptosis by activating the SIRT6/NRF2/GPX4 signaling pathway. Food Funct. 2026;17(8):3587–3599. doi:10.1039/d5fo04381f

52. Youn JY, Gao L, Cai H. The p47phox- and NADPH oxidase organiser 1 (NOXO1)-dependent activation of NADPH oxidase 1 (NOX1) mediates endothelial nitric oxide synthase (eNOS) uncoupling and endothelial dysfunction in a streptozotocin-induced murine model of diabetes. Diabetologia. 2012;55(7):2069–2079. doi:10.1007/s00125-012-2557-6

53. Zhang Y, Youn JY, Huang K, et al. Alleviation of accelerated diabetic atherogenesis in STZ-treated apoE/NOX1 DKO mice, apoE-/-/tg-EC-DHFR mice, and by folic acid. Redox Biol. 2025;82:103570. doi:10.1016/j.redox.2025.103570

54. Lyon P, Strippoli V, Fang B, et al. B Vitamins and One-Carbon Metabolism: implications in Human Health and Disease. Nutrients. 2020;12(9):2867. doi:10.3390/nu12092867

55. Kadam I, Dalloul M, Hausser J, et al. Associations between nutrients in one-carbon metabolism and fetal DNA methylation in pregnancies with or without gestational diabetes mellitus. Clin Clin Epigenet. 2023;15(1):137. doi:10.1186/s13148-023-01554-1

56. Kranenburg E, Kubant R, Cho CE, et al. Folic acid reduces insulin resistance in mice with diet-induced obesity by altering one-carbon metabolism and DNA methylation patterns of hypothalamic and hepatic insulin receptor gene. Mol Nutr Food Res. 2025;69(20):e70181. doi:10.