Homocysteine (Hcy), a sulfur-containing amino acid, is a known risk factor for cardiovascular disease. Hyperhomocysteinemia (hHcy), characterized by elevated blood Hcy levels, is associated with various health problems due to its involvement in DNA hypomethylation, oxidative stress, and cellular damage. While the role of B vitamins in Hcy metabolism is well-established, the association between mineral intake and Hcy levels is less understood. Existing research suggests a relationship between individual minerals (calcium, zinc, selenium) and Hcy, but the effects of mixed mineral intake on blood Hcy levels and the relative contribution of each mineral remain unclear. This study aimed to investigate the association between the combined intake of ten common minerals (calcium, potassium, magnesium, sodium, iron, zinc, selenium, phosphorus, copper, and manganese) and blood Hcy levels, employing both traditional regression analysis and advanced machine learning techniques to account for the complexity of mineral interactions. The study used data from the Shanghai Suburban Adult Cohort and Biobank (SSACB), a large-scale population-based study offering a robust dataset to address this research gap. Understanding the combined effect of multiple minerals on Hcy levels holds significant potential for improving cardiovascular health strategies.

Several studies have explored the link between individual mineral intake and homocysteine levels. Studies on postmenopausal women have indicated a negative association between calcium intake and serum Hcy levels, although the generalizability to other populations was limited. Research also suggests a positive correlation between plasma zinc concentration and the risk of hHcy, with lower concentrations being associated with reduced risk. A non-linear relationship was observed between blood selenium levels and hHcy prevalence, with individuals in higher selenium quartiles exhibiting a lower risk. These findings highlight the potential role of minerals in Hcy metabolism, possibly through their involvement in enzymes such as Paraoxonase 1 and Betaine-Hcy methyltransferase. However, these studies primarily focused on individual minerals, neglecting the simultaneous intake of multiple minerals and potential synergistic or antagonistic interactions. Therefore, to assess the overall effect of combined mineral intake, the study will utilize advanced machine learning techniques capable of handling complex interactions among the minerals.

This cross-sectional study utilized baseline data from the 2016 Shanghai Suburban Adult Cohort and Biobank (SSACB) study. A total of 38,273 participants aged 20-74 years were included after applying inclusion and exclusion criteria (e.g., complete FFQ data, energy intake within specified ranges, no missing key information). Dietary intake of ten minerals was assessed using a validated food frequency questionnaire (FFQ), and blood Hcy concentrations were measured using fluorescence chromatography. Participants with blood Hcy concentrations ≥15 µmol/L were classified as having hHcy. Traditional multiple linear and multivariate logistic regression models were used to examine the associations between individual mineral intakes and blood Hcy levels/hHcy risk, adjusted for various covariates identified using a Directed Acyclic Graph (DAG). Three machine learning methods – Weighted Quantile Sum (WQS) regression, Quantile G-computation (Qg-comp), and Bayesian kernel machine regression (BKMR) – were employed to analyze the combined effect of the ten minerals on blood Hcy levels and hHcy prevalence. The BKMR analysis was performed on a 10% subsample to improve computational efficiency. All analyses were adjusted for age, sex, education, marital status, smoking, alcohol and tea consumption, energy intake, physical activity, BMI, hypertension, coronary heart disease, diabetes, and vitamin B6 and B12 intake. Statistical significance was set at p<0.05, and the Benjamini-Hochberg procedure was used to adjust p-values for multiple comparisons. The data were analyzed using SAS (version 9.4) and R software (version 4.3.2).

Traditional regression models showed that higher intake of calcium, phosphorus, potassium, magnesium, iron, zinc, copper, and manganese was associated with lower blood Hcy levels and reduced hHcy risk. The WQS model indicated a significant negative association between overall mineral intake and both blood Hcy levels (WQS: -0.201, P=0.001) and hHcy prevalence (WQS: -0.074, P<0.002). Calcium had the highest weight in the overall effect. The Qg-comp model also showed a negative association between mixed mineral intake and Hcy levels and hHcy prevalence, with iron showing the highest positive weight and manganese the highest negative weight. The BKMR analysis, conducted on a subsample, revealed a negative association between higher intake of nine minerals (excluding sodium) and both blood Hcy levels and hHcy prevalence. Phosphorus showed the highest posterior inclusion probability (PIP) for Hcy levels, while zinc had the highest PIP for hHcy prevalence. Exposure-response curves from BKMR indicated negative associations between intakes of several minerals and Hcy levels and hHcy risk, even when other minerals were held constant. Some interactions were found between certain minerals in their effects on hHcy.

This study provides substantial evidence for an association between higher mixed mineral intake and lower blood Hcy levels and reduced hHcy risk. The findings from both traditional regression and machine learning models consistently support this association. The utilization of advanced machine learning techniques, capable of handling the complexities of correlated mineral intakes and potential interactions, significantly enhances the understanding of this relationship. The varying weights of individual minerals in the joint effect highlight the need to consider the combined effect rather than focusing solely on individual minerals. The findings are supported by existing literature showing the involvement of minerals in key enzymes related to Hcy metabolism, such as Paraoxonase 1 and Betaine-Hcy methyltransferase. This study strengthens the understanding of the potential role of a balanced mineral intake in maintaining healthy Hcy levels and reducing the risk of associated health problems.

Higher intake of mixed minerals is associated with lower blood Hcy levels and reduced hHcy prevalence. Each mineral contributes differently to the joint effect, emphasizing the importance of a balanced dietary intake. Future research should investigate the underlying mechanisms, explore the impact on other health outcomes, and use more precise dietary assessments to strengthen the understanding of the relationship between mixed mineral intake and Hcy metabolism.

The cross-sectional design limits causal inference. The FFQ, while validated, might have limitations in accuracy and recall bias. The sex imbalance in the sample might influence the results, although sex was adjusted for in the models. The Chinese Food Composition Table might not encompass all minerals relevant to Hcy metabolism. Future studies should utilize longitudinal designs, incorporate objective measures of mineral intake (e.g., blood mineral levels), and consider a wider range of minerals to enhance the robustness of the findings.