Participant characteristicsParticipants were grouped into non-CHD (n = 32,211) and CHD (n = 1429). We recorded the following variables for included samples: Vitamin E, Vitamin B6, Vitamin B12, age, gender, household annual income, education level, race, BMI, smoking status, diabetes status, hypertension status, and drinking status. There was a significant difference in Vitamin B12 levels between the two groups, with the CHD group having a level of 4.93 (0.21) and the non-CHD group having a level of 5.24 (0.05). Except for smoking status, there were significant differences in baseline characteristics between the two groups. Participants in the CHD group were older (67.25 (0.36) vs 46.68 (0.23), P < 0.0001), had lower levels of Vitamin B6 (P = 0.001) and Vitamin E (P = 0.003) than the non-CHD group, and had a higher BMI than the non-CHD group (P < 0.0001). Participants in the CHD group were more likely to have hypertension and a history of alcohol consumption (both P < 0.0001). Additionally, noticeable differences between the two groups in terms of racial distribution and education level (both P < 0.0001) (Table 1), and their basic characteristics are shown in Table 1.
Table 1 Baseline characteristics of the study population based on the presence or absence of coronary heart disease.Association of Vitamin E, Vitamin B6, Vitamin B12 intake with coronary heart diseaseMultivariate regression analysis revealed that Vitamin E and Vitamin B6 were negatively associated with CHD and exerted protective effects, while Vitamin B12 had little correlation with CHD.In view of the quartiles of Vitamin E percentage, the third quartile (Q3) in all four models manifested a negative relevance between Vitamin E percentage and the suffering from CHD, indicating a protective effect. In comparison with the first quartile (Q1) of the overall, the fully adjusted model manifested a stronger protective effect against CHD in the Q3 (OR 0.66; 95% CI 0.52 to 0.83). The P value for the trend was < 0.03, with a P value of < 0.001 (Table 2). Among all the models, the strongest protective effect was observed in the third quartile (Q3), where the intake of Vitamin E was 6.45 to 9.89 mg, which maximally reduced the prevalence of CHD.
Table 2 Results from a multiple logistic regression analysis of the association between Vitamin E intake, Vitamin B6 intake, Vitamin B12 intake and coronary heart disease, weighted.In view of the quartiles of Vitamin B6 percentage, all four models manifested a negative relevance between Vitamin B6 percentage and the suffering from CHD, indicating a protective effect. In comparison with the first quartile (Q1) of the overall, the fully adjusted model manifested a stronger protective effect against CHD in the third quartile (Q3) (OR 0.77; 95% CI 0.62 to 0.94). The P value for the trend was < 0.01 (Table 2). Among all the models, the strongest protective effect was observed in the third quartile (Q3), where the intake of Vitamin B6 was 1.703 to 2.466 mg, which maximally reduced the prevalence of CHD.In view of the quartiles of Vitamin B12 percentage, most models manifested no statistically significant relevance between Vitamin B12 percentage and the suffering from CHD. Vitamin B12 intake was found to maybe increase the risk of coronary heart disease. In the original model (OR 1.23; 95% CI 1.02 to 1.49, P = 0.03) Vitamin B12 quartile 2 intake was statistically significant.In addition, dose–response relationships between the prevalence of CHD and intake of Vitamin E, Vitamin B6, and Vitamin B12 were examined using RCS curves. We identified a nonlinear relationship between Vitamin E, Vitamin B6 and CHD (nonlinear P-value = 0.0004, nonlinear P-value = 0.0403). As Vitamin E intake increased, the OR curve for CHD decreased first, followed by a slow increase, finally to stabilize. The turning point of the OR curve was found to be 7.54 mg/day of Vitamin E intake (Fig. 2A). According to Fig. 2B, it was found that the OR curve for CHD dropped sharply at first with the increase in Vitamin B6 intake, then rose slowly and finally leveled off. The turning point of the OR curve was found to be 2.65 mg/day of Vitamin B6 intake. For Vitamin B12 there was no non-linear relationship found with CHD (non-linear P-value = 0.1322).Figure 2(A) RCS curves describing the dose–response relationship between Vitamin E intake and coronary heart disease. (B) RCS curves describing the dose–response relationship between Vitamin B6 intake and coronary heart disease. (C) RCS curves describing the dose–response relationship between Vitamin B12 intake and coronary heart disease. The following covariates were adjusted for: age, sex, education level, race, family income, BMI, drinking status, smoking status, hypertension and diabetes mellitus.We also conducted RCS curve analyses for the drinking as well as smoking status subgroups which revealed that the OR curve for the non-drinking population started with an increase and then decreased sharply with a turning point at 3.18 mg/day of Vitamin E intake, whereas the OR curve for the drinking population was the opposite of that of the non-drinking population, with the curve decreasing sharply and then increasing, which turned at 7.48 mg/day of Vitamin E intake (Fig. 3A). For the smoking status subgroup, we can find that the OR curve for the nonsmoking population first rises, then falls sharply, and then rises again, with a turning point at 7.54 mg/day of Vitamin E intake, while the OR curve for the drinking population is the opposite of that of the non-drinking population, with the curve first falling sharply, then rising, with a turning point at 7.48 mg/day of Vitamin E intake (Fig. 3B). For the RCS curve analysis of Vitamin B6, we found that the OR curve of the drinking population decreased and then increased, with a turning point at 2.53 mg/day of Vitamin B6 intake, while the curve of the non-drinking population showed a decreasing trend (Fig. 3C); the OR curve of the non-smoking population decreased with the increase of Vitamin B6 intake, whereas the smoking population showed the opposite trend, which increased with the increase of intake (Fig. 3D). All of the above analyses were performed under the fully adjusted model.Figure 3(A) RCS curves describing the dose–response relationship between Vitamin E intake and coronary heart disease by grouping based on drinking status. (B) RCS curves describing the dose–response relationship between Vitamin E intake and coronary heart disease by grouping based on smoking status. (C) RCS curves describing the dose–response relationship between Vitamin B6 intake and coronary heart disease by grouping based on drinking status. (D) RCS curves describing the dose–response relationship between Vitamin B6 intake and coronary heart disease by grouping based on smoking status. The following covariates were adjusted for: age, sex, education level, race, family income, BMI, drinking status, smoking status, hypertension and diabetes mellitus.Analyses of subgroups and interactionsFigure 4 presents the results of stratified analysis based on age, gender, education level, body BMI, household annual income, race, diabetes status, hypertension status, smoking status, and drinking status. These factors significantly influence the risk of CHD, therefore we conducted stratified analysis for a more insightful understanding of the links between Vitamin E, Vitamin B6, and Vitamin B12 intake and the occurrence of CHD. The results of stratified analysis showed that in racial subgroup analysis, the intake of Vitamin B6 in Mexican Americans possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.77, P = 0.002). The intake of Vitamin B6 in non Hispanic black individuals possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.81, P = 0.03), while the intake of Vitamin B6 in non Hispanic white individuals possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.94, P = 0.03), The intake of Vitamin B12 in Mexican Americans possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.92, P = 0.01). In racial subgroup analysis, the intake of Vitamin B6 in women possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.85, P < 0.0001), while the intake of Vitamin E in women possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.97, P = 0.001). In BMI subgroup analysis, when the participant’s BMI was in the 25–30 range, the intake of Vitamin B6 possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.03, P < 0.0001), the P-value of the interaction was 0.01, and the intake of Vitamin E possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.43, P < 0.0001), and the P-value of the interaction was 0.005. In the subgroup analysis of smoking status, when participants smoked, the intake of Vitamin B6 possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.90, P < 0.001), and the intake of Vitamin E possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.97, P < 0.001). In the subgroup analysis of household annual income, when the participant’s household annual income was less than $20,000, the intake of Vitamin E possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.93, P < 0.0001), and the interaction was P < 0.001. The intake of Vitamin B12 possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.94, P = 0.002); When the annual household income of participants is ≥ $20,000, the intake of Vitamin B6 possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.95, P = 0.02). In the subgroup analysis of education level, when the participants had a high school education level, the intake of Vitamin B6 possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.88, P = 0.02), and the intake of Vitamin E possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.95, P = 0.02), with a p-value of 0.01 for the interaction; When the participant’s education level is higher than high school level, the intake of Vitamin B6 possessed a negative correlation with the probability of suffering from CHD (odds ratio [OR] 0.94, P = 0.02).Figure 4The Forest plots illustrate stratified subgroup analyses by age, gender, BMI, education level, race, family income, drinking status, smoking status, hypertension, diabetes mellitus. Subgroup analyses of Vitamin E, Vitamin B6, and Vitamin B12 intake in association with coronary heart disease.
Association between Vitamin E, Vitamin B6, and Vitamin B12 with coronary heart disease
