Abstract
Background
Cardiovascular disease (CVD) is a major global health concern with increasing incident cases and deaths. Homocysteine (Hcy) has been investigated for its potential association with CVD, researchers have debated the extent to which Hcy should be considered a risk factor for cardiovascular diseases, as only 50% of CVD can be explained by classical risk factors.
Methods
We conducted a prospective cohort study using NHANES 1999–2006 data, analyzing 1,739 US patients aged at least 30 with CVD. Cox proportional hazards regression and restricted cubic splines were used to examine the relationship between Hcy levels and mortality, adjusting for covariates.
Result
A total of 1,739 participants with cardiovascular disease (CVD) were enrolled, with a median follow-up period of 126 months. Among them, 1,194 participants died, including 501 deaths due to cardiovascular causes. After adjusting for covariates, the hazard ratios (HR) and 95% confidence intervals (CI) for CVD mortality at different levels of homocysteine (Hcy) (T1 (< 9.3), T2 (9.3–12.5), T3 (> 12.5)) were 1.26 (0.92, 1.73) (T2), and 1.69 (1.14, 2.51) (T3) (P for trend = 0.0086). The HR and 95% CI for all-cause mortality at different levels of Hcy were 1.22 (1.05, 1.42) (T2) and 1.64 (1.29, 2.09) (T3) (P for trend < 0.0001). Elevated Hcy levels were associated with increased risks of all-cause mortality and CVD deaths, even at levels below the conventional threshold. The nonlinear relationship was observed, with inflection points at 14.5 µmol/L for all-cause mortality and 14.6 µmol/L for CVD mortality. Subgroup analyses revealed interactions with age, serum vitamin B12, and smoking.
Conclusion
Our study supports the notion that elevated Hcy levels are associated with higher all-cause and CVD mortality risks in CVD participants. The impact of Hcy on health outcomes can be observed at lower concentrations than previously thought.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12872-024-04317-9.
Keywords: Homocysteine, Cardiovascular disease, All-cause death, NHANES
Background
Cardiovascular disease (CVD) is a group of disorders of the heart and blood vessels, including coronary heart disease, cerebrovascular disease, and heart failure, among others [1]. Global epidemiologic trends in CVDs have shown an increase in both incident cases and deaths over the years. From 1990 to 2019, CVD incident cases increased by 77.12%, reaching 55.45 million, while deaths rose by 53.81%, reaching 25.07 million. However, age-standardized incidence and mortality have shown a consistent decrease in most countries, indicating progress in addressing CVDs. Nonetheless, specific regions, especially low to middle income ones, still face worrying increases in CVD cases and deaths [2].
Homocysteine (Hcy) is an amino acid that has been studied for its potential association with cardiovascular health. Elevated levels of Hcy, known as hyper homocysteinemia, have been proposed as a risk factor for cardiovascular diseases such as coronary artery disease, heart attacks, and strokes [3, 4]. However, the role of Hcy in the development of cardiovascular disease is still a topic of debate, and the results from studies examining this association have been inconsistent [5].
A meta-analysis of prospective studies found that elevated Hcy levels are an independent predictor for subsequent cardiovascular mortality or all-cause mortality. The risks were more pronounced among the general population, with a 66% increase in coronary heart disease (CHD) mortality, a 68% increase in cardiovascular mortality, and a 93% increase in all-cause mortality when comparing the highest to lowest Hcy level categories. For each 5 µmol/L Hcy increment, the pooled risk ratio was 1.52 for CHD mortality, 1.32 for cardiovascular mortality, and 1.27 for all-cause mortality [6]. However, it also illustrated an Inconclusive association in the elderly population. A community-based prospective cohort study in Beijing evaluated the association between Hcy and CVD and all-cause death in the elderly population. The study found that every 5 µmol/L increment in Hcy concentration was associated with a 4% higher risk of CVD events and a 5% higher risk of all-cause death in all participants. However, there was no significant association between moderate homocysteinemia and the risk of CVD events and all-cause death in CVD patients [7]. Additionally, researchers have debated the extent to which Hcy should be considered a risk factor for cardiovascular diseases, as only 50% of CVD can be explained by classical risk factors [5].
We aim to clarify the impact of Hcy on the risk of CVD and all-cause mortality in individuals with pre-existing CVD. This involves investigating whether moderately elevated Hcy levels, or even those within the normal range as mentioned earlier, contribute to the risk. Such insights are crucial for the proactive management of Hcy levels.
Method
Study population
The analyzed data is derived from National Health and Nutrition Examination Survey (NHANES) 1999–2006. This dataset employs a complex, stratified, multi-stage probability sampling procedure to collect health and nutrition data from a representative, non-institutionalized civilian population in the United States. It combines household interviews and physical examinations, including blood sample collection at Mobile Examination Centers (MEC). Trained interviewers use the Computer-Assisted Personal Interview (CAPI) system to inquire about demographic information and characteristics during household interviews and MEC visits. Variables such as age, gender, race, marital status, income, smoking status are obtained from household visit data. Data on Hcy, serum vitamin B12, blood uric acid, blood creatinine, blood total cholesterol, blood triglycerides, and Body Mass Index (BMI) are sourced from MEC examination data. Given that the median follow-up time is only about 10 years, we focused on the group with a higher probability of death within 10 years. Therefore, we selected participants over 30 years of age. In this study, a total of 41,474 participants were surveyed and 16,445 participants aged at least 30 years. A total of 1739 participants were included in the final analysis, after excluding those with missing serum Hcy concentrations (n = 4043), 10,663 individuals with non-CVD history (n = 10663). (see Fig. 1).
Fig. 1.
Flow chart of study participants
Exposure factor
In NHANES 1999–2001, serum Hcy was measured through a fully automated fluorescence polarization immunoassay (FPIA) using the Abbott Hcy IMX (HCY) assay from Abbott Laboratories, Chicago, IL, USA [8]. In NHANES 2002–2005, Hcy measurements were conducted using FPIA on the Abbott AxSym analyzer (Abbott Diagnostics, Abbott Laboratories, Chicago, IL, USA) [9]. According to NHANES documentation, these two methods are considered equivalent [10, 11].
Results definition
The primary outcomes were CVD mortality and all-cause mortality. CVD death was defined as death associated cardiovascular events from baseline examination, the specific ICD codes related to CVD mortality can be found in footnote 4 of Table S1. And all-cause death was defined as death from any cause. Data were sourced from NCHS Surveys Linked to NDI Mortality Data.
Covariates
Multivariable analysis controlled for the following covariates: age, sex, race, diabetes, hypertension, BMI, eGFR, smoking status, income level, serum vitamin B12 and marital status. Marital status was considered based on whether one currently lived alone or cohabited with a partner. Poverty Income Ratio (PIR) was used to measure income, calculated by dividing household income by the poverty guidelines for the survey year. Poverty guidelines vary by family size and geographic location. In this study, PIR was used to create two income categories: low (PIR < 1.3) and high (PIR ≥ 1.3), indicating socioeconomic status based on eligibility for Supplemental Nutrition Assistance Program (SNAP) benefits [12]. Smoking status was categorized as current or never smoked. Considering that over one-third of Americans are obese [13], BMI was classified as normal (< 25 kg/m2) and obese (≥ 25 kg/m2).
Medical conditions was self-reported by the participants. Regarding CVD, individuals who answered “yes” to the question “Has a doctor or health professional ever told you that you have coronary heart disease/angina, also known as angina/heart attack (also known as myocardial infarction)/stroke/congestive heart failure (CHF)?” were considered to have CVD. In the Nt-proBNP test, individuals under 50 with levels exceeding 450 pg/ml, those aged 50 to 70 with levels exceeding 900 pg/ml, and those aged 75 and above with levels exceeding 1800 pg/ml were considered positive for CVD. The diagnostic criteria for diabetes included recalling a doctor’s diagnosis of diabetes in the medical history, HbA1c (%) > 6.5, random blood glucose (mmol/L) ≥ 11.1, 2-hour OGTT blood glucose (mmol/L) ≥ 11.1, and the use of diabetes medications or insulin. The diagnostic criteria for hypertension included recalling a doctor’s diagnosis of hypertension in the medical history and the use of antihypertensive medications. The new CKD-EPI formula [14]was used to calculate creatinine clearance in this study. eGFR below 60 ml/min was defined as the CKD population.
Statistical analysis
We conducted the analysis using NHANES sampling weights, taking into account the complex multi-stage cluster survey design. For continuous variables, displayed as mean and standard deviation (SD), and categorical variables displayed as counts and percentages, the Kruskal-Wallis rank sum test was used for continuous variables conforming to a normal distribution, and the chi-squared test for categorical variables.
Event time analysis was used to determine the association between Hcy and all-cause mortality. Cox proportional hazards regression analysis was employed for both univariate and multivariable analyses. The Schoenfeld residual was used for the proportional hazard assumption test. To address multicollinearity, the median of each group was used as a continuous variable in the model for trend testing. The variance inflation factor (VIF) was calculated to assess multicollinearity among variables in the multivariable Cox model, with a VIF < 10 indicating that multicollinearity may not affect the estimates [15].
Given the potential for non-linear associations in linear analysis, restricted cubic splines were used in the multivariablemodel. The number of nodes for the spline equation was chosen based on Harrell’s recommendation. In our two models, the AIC was smallest when the number of nodes was 4. Additionally, Harrell suggests that when there are four nodes, their positions should be at the 5th, 35th, 65th, and 96th percentiles [16]. relying on the Akaike Information Criterion (AIC) to evaluate model fit. The model with the minimum AIC was selected as the final model. The relationship between Hcy and CVD mortality, as well as Hcy and all-cause mortality, was examined. Using Hcy = 15umol/L as the reference point, the changes in the hazard ratio (HR) and 95% confidence interval (95% CI) were observed. The Wald test was employed to examine the nonlinearity of variables, and a recursive algorithm was used to calculate the inflection points for Hcy and all-cause mortality and Hcy and CVD mortality. A P-value less than 0.05 was considered statistically significant. All analyses considered the complex survey design and were appropriately weighted.
All analyses were conducted using R software (4.2.3) following CDC guidelines.
Result
In this study, the analyzed dataset included 1739 patients with CVD, with a total follow-up of 126.222 (68.789) months. There were some differences in age and mortality between participants with missing data and those included in the analysis (see Table S1). The concentrations of Hcy in their serum were divided into three groups based on tertiles: T1 (< 9.3), T2 (9.3–12.5), T3 (> 12.5). The baseline characteristics of the study population are shown in Table 1. It can be observed that the levels of Hcy in females are generally lower than in males, and the probability of higher Hcy levels increases with age. Smokers have a higher probability of having elevated Hcy levels than non-smokers. Additionally, higher levels of serum VB12 and eGFR are associated with a greater likelihood of lower Hcy levels. And the Kaplan-Meier (KM) curve can be seen in Supplement Fig. 1. The higher Hcy levels, the higher the risk of cardiovascular death and all-cause death.
Table 1.
Baseline characteristics of the study population with CVD from the NHANES dataset
| Homocysteine levels | ||||||
|---|---|---|---|---|---|---|
| Characteristic | Overall, N = 1,7391 | 1, N = 5831 | 2, N = 5761 | 3, N = 5801 | p-value 2 | q-value 3 |
| Sex | < 0.001 | < 0.001 | ||||
| female | 746 (43%) | 303 (52%) | 224 (39%) | 219 (38%) | ||
| male | 993 (57%) | 280 (48%) | 352 (61%) | 361 (62%) | ||
| Age(years) | 68 (13) | 62 (13) | 70 (12) | 72 (11) | < 0.001 | < 0.001 |
| Race | 0.068 | > 0.9 | ||||
| Non-Hispanic Black | 312 (18%) | 109 (19%) | 98 (17%) | 105 (18%) | ||
| Non-Hispanic White | 1,088 (63%) | 340 (58%) | 372 (65%) | 376 (65%) | ||
| Other Race | 339 (19%) | 134 (23%) | 106 (18%) | 99 (17%) | ||
| Hypertension | < 0.001 | 0.017 | ||||
| No | 568 (33%) | 217 (37%) | 194 (34%) | 157 (27%) | ||
| Yes | 1,171 (67%) | 366 (63%) | 382 (66%) | 423 (73%) | ||
| Diabetes | 0.3 | > 0.9 | ||||
| No | 1,184 (68%) | 409 (70%) | 392 (68%) | 383 (66%) | ||
| Yes | 555 (32%) | 174 (30%) | 184 (32%) | 197 (34%) | ||
| Body mass index (kg/m2) | 29.2 (6.3) | 29.9 (6.4) | 29.0 (5.7) | 28.9 (6.8) | < 0.001 | 0.016 |
| Smoking status | 0.14 | > 0.9 | ||||
| No | 655 (38%) | 237 (41%) | 215 (37%) | 203 (35%) | ||
| Yes | 1,084 (62%) | 346 (59%) | 361 (63%) | 377 (65%) | ||
| Marital Status | < 0.001 | < 0.001 | ||||
| couple | 1,019 (59%) | 377 (65%) | 350 (61%) | 292 (50%) | ||
| single | 720 (41%) | 206 (35%) | 226 (39%) | 288 (50%) | ||
| Poverty income ratio | 0.2 | > 0.9 | ||||
| High income | 1,144 (66%) | 396 (68%) | 384 (67%) | 364 (63%) | ||
| Low income | 595 (34%) | 187 (32%) | 192 (33%) | 216 (37%) | ||
| Education | 0.002 | 0.033 | ||||
| Less than high school | 705 (41%) | 209 (36%) | 231 (40%) | 265 (46%) | ||
| High school or equivalent | 422 (24%) | 157 (27%) | 124 (22%) | 141 (24%) | ||
| College or above | 612 (35%) | 217 (37%) | 221 (38%) | 174 (30%) | ||
| Homocysteine(umol/L) | 11.8 (6.4) | 7.5 (1.2) | 10.6 (0.9) | 17.3 (8.4) | < 0.001 | < 0.001 |
| Serum Vitamin B12 (pg/mL) | 553 (402) | 638 (480) | 547 (339) | 474 (356) | < 0.001 | < 0.001 |
| eGFR(ml/min) | 73 (25) | 89 (17) | 74 (19) | 56 (25) | < 0.001 | < 0.001 |
| Total cholesterol (mg/dl) | 196 (48) | 197 (45) | 195 (51) | 196 (48) | 0.3 | > 0.9 |
| Uric acid (mg/dL) | 6.03 (1.66) | 5.34 (1.32) | 5.98 (1.44) | 6.76 (1.87) | < 0.001 | < 0.001 |
| Triglycerides (mg/dL) | 164 (133) | 162 (99) | 158 (99) | 172 (183) | 0.7 | > 0.9 |
| CVD death | 501 (29%) | 120 (21%) | 174 (30%) | 207 (36%) | < 0.001 | < 0.001 |
| All-cause Death | 1,194 (69%) | 285 (49%) | 411 (71%) | 498 (86%) | < 0.001 | < 0.001 |
| following-up time(months) | 126 (69) | 158 (62) | 126 (67) | 95 (62) | < 0.001 | < 0.001 |
1n (%); Mean (SD)
2Pearson’s Chi-squared test; Kruskal-Wallis rank sum test
3Bonferroni correction for multiple testing
During the 18291.67 person-years of follow-up, a total of 1194 deaths occurred, including 501 CVD-related deaths. Through three COX regression models studying the independent effect of Hcy on mortality(see Table 2), it was found that, after adjusting for age, sex, race/ethnicity, education, poverty income ratio, smoking status, history of hypertension or diabetes, eGFR, serum total cholesterol, serum uric acid, marital status, serum triglycerides, BMI, and serum vitamin B12, the HR and 95% CI for CVD mortality at different levels of Hcy (T1 (< 9.3), T2 (9.3–12.5), T3 (> 12.5)) were 1.00 (ref), 1.26 (0.92, 1.73), 1.69 (1.14, 2.51) (P for trend = 0.0086). The HR and 95% CI for all-cause mortality at different levels of Hcy (T1 (< 9.3), T2 (9.3–12.5), T3 (> 12.5)) were 1.00 (ref),
Table 2.
HRs (95% CIs) for mortality according to hcy concentrations among participants
| Serum Homocysteine levels(umol/L) | P for trend | |||
|---|---|---|---|---|
| T1(< 9.3) | T2(9.3–12.5) | T3(> 12.5) | ||
| CVD Mortality | ||||
|
Model 1 HR( 95%CI) |
ref | 1.98(1.48, 2.66) | 3.35(2.52, 4.47) | < 0.0001 |
|
Model 2 HR( 95%CI) |
ref | 1.42 (1.06, 1.90) | 2.04(1.49, 2.77) | < 0.0001 |
|
Model 3 HR( 95%CI) |
ref | 1.26(0.92, 1.73) | 1.69(1.14, 2.51) | 0.0086 |
| All-cause Mortality | ||||
|
Model 1 HR( 95%CI) |
ref | 1.95(1.64, 2.31) | 3.41(2.85, 4.09) | < 0.0001 |
|
Model 2 HR( 95%CI) |
ref | 1.40(1.21, 1.62) | 2.11(1.75, 2.54) | < 0.0001 |
|
Model 3 HR( 95%CI) |
ref | 1.22(1.05, 1.42) | 1.64(1.29, 2.09) | < 0.0001 |
Model 1: Non-adjusted
Model 2: Adjusted for age, sex, race/ethnicity
Model 3: Adjusted for age, sex, race/ethnicity, education, poverty income ratio, smoking status, history of hypertension or diabetes, eGFR, serum total cholesterol, serum uric acid, marital status, serum triglycerides, BMI, and serum vitamin B12
1.22 (1.05, 1.42), 1.64 (1.29, 2.09) (P for trend < 0.0001).
Through RCS curve analysis, we observed a nonlinear correlation between Hcy and CVD mortality, as well as between Hcy and all-cause mortality(see Fig. 2). The inflection points for all-cause mortality and CVD mortality were calculated to be 14.5 umol/L and 14.6 umol/L, respectively, using a recursive algorithm. Refer to Table 3 for specific data. Subsequently, we employed a two-piecewise Cox proportional hazards model and compared it with a linear Cox model. The two-piecewise model demonstrated significant superiority in fitting compared to the linear Cox model (P for log-likelihood ratio < 0.05).
Fig. 2.
Dose–response relationships between Hcy and CVD mortality (A) and all-cause mortality (B), Adjusted for age, sex, race/ethnicity, education, poverty income ratio, smoking status, history of hypertension or diabetes, eGFR, serum total cholesterol, serum uric acid, marital status, serum triglycerides, BMI, and serum vitamin B12
Table 3.
Threshold effect analysis of hcy concentrations on all-cause and CVD mortality in CVD patients
| All-cause mortality | Adjusted HR (95% CI) P-value |
|---|---|
| Fitting by the standard linear model | 1.02(1.01,1.03) 0.0005 |
| Fitting by the two-piecewise linear model | |
| Inflection point | 14.5 |
| Hcy < 14.5 umol/L | 1.06(1.01,1.11)0.012 |
| Hcy ≥ 14.5 umol/L | 1.01(1.00,1.02)0.016 |
| P for Log-likelihood ratio | 0.032 |
| CVD mortality | |
| Fitting by the standard linear model | 1.02(1.00,1.03)0.011 |
| Fitting by the two-piecewise linear model | |
| Inflection point | 14.6 umol/L |
| Hcy < 14.6 umol/L | 1.08(1.01,1.16)0.02 |
| Hcy ≥ 14.6 umol/L | 1.01(0.99,1.03)0.20 |
| P for Log-likelihood ratio | 0.044 |
Regarding CVD mortality (see Fig. 3), compared to lower serum Hcy concentration (< 14.5 umol/L) as reference, there are differences in the impact of elevated serum Hcy concentration (≥ 14.5 umol/L) on CVD mortality in the subgroups of age, race, hypertension, serum Vitamin B12 and smoking status with an interaction between age, serum Vitamin B12, smoking status and this variable (P < 0.05). Regarding all-cause mortality (see Fig. 4), compared to lower serum Hcy concentration (< 14.5 umol/L) as reference, there are differences in the impact of elevated serum Hcy concentration (≥ 14.5 umol/L) on all-cause mortality in the subgroups of age and hypertension with an interaction between age and this variable (P < 0.05) and the causes of death are available in Supplement Fig. 2.
Fig. 3.
Forest plots of stratified analyses of Hcy and CVD mortality. Adjusted for age, sex, race/ethnicity, education, poverty income ratio, smoking status, history of hypertension or diabetes, eGFR, serum total cholesterol, serum uric acid, marital status, serum triglycerides, BMI, and serum vitamin B12
Fig. 4.
Forest plots of stratified analyses of Hcy and all-cause mortality. Adjusted for age, sex, race/ethnicity, education, poverty income ratio, smoking status, history of hypertension or diabetes, eGFR, serum total cholesterol, serum uric acid, marital status, serum triglycerides, BMI, and serum vitamin B12
Discussion
Our study is a prospective study with a relatively large sample size, aimed at analyzing the relationship between serum Hcy levels and all-cause as well as CVD mortality in patients with CVD. In our cohort study, we observed that elevated serum Hcy levels were associated with an increased risk of both all-cause mortality and CVD deaths. While the typical definition of high homocysteine levels is set at greater than 15 µmol/L, our analysis revealed a significant association between Hcy levels and the risk of all-cause mortality and cardiovascular death even at levels below 14.5 µmol/L and 14.6 µmol/L. This suggests that the impact of Hcy on these health outcomes can be observed at lower concentrations than the commonly accepted threshold.
Our research has reached conclusions consistent with previous studies, suggesting that elevated levels of Hcy are associated with an increased risk of all-cause mortality [17, 18]. A cohort study indicated that in the population with coronary heart disease, even when Hcy levels fell within the normal range, the mortality rate for this group increased by 1.9 times compared to the reference group. This finding underscores the potential benefits of maintaining lower Hcy levels for patients with cardiovascular disease [19]. Another study highlighted the vulnerability of elderly individuals, as high Hcy levels in the elderly were significantly associated with an increased risk of cardiovascular events, cardiovascular mortality and all-cause mortality [1.68 (95% CI 1.06–2.67), 1.97 (95% CI 0.95–4.29), and 2.02 (95% CI 1.26–3.24)]. It’s worth noting that this effect was not statistically significant in the female subgroup, emphasizing the need for further research to clarify the potential gender-specific impact [7]. However, in VISP research, a randomized double-blind trial, high dose vitamin therapy had no effect on the outcome measures of stroke, CHD events, or death [20]. One possible reason that the treatment was not effective may have been that patients enrolled in this study had levels of total Hcy that were too low to show a large effect. Another consideration is that a longer duration of treatment may be necessary. Therefore, we also focus on exploring the long-term prognosis of higher levels of Hcy.
As previously mentioned, numerous studies have demonstrated a dose-response relationship between Hcy levels and the risk of mortality. Each 5 µmol/L increment in Hcy is significantly associated with an increased risk of all-cause and CVD mortality [6, 17]. The underlying pathophysiological mechanisms are intricate and multifaceted. One such mechanism involves the direct toxicity of Hcy on tissues, which can result in various detrimental effects, including oxidative stress, smooth muscle cell proliferation, the formation of reactive oxygen species, and the induction of unfolded protein responses [21–23]. These effects can contribute to the development of atherosclerosis, a major contributor to CVD [24]. On the other hand, homocysteinemia is also associated with the metabolic deregulation of the methionine cycle, leading to an imbalance between the biosynthesis and catabolism of Hcy [25]. This imbalance is linked to an increased risk of CVD, stroke, and myocardial infarction through mechanisms such as atherosclerosis, thrombosis, endothelial cell dysfunction, and oxidative stress [26, 27]. Another critical factor is the role of B vitamins in Hcy metabolism. Vitamin B and folate serve as important cofactors in Hcy metabolism, and an insufficient dietary supply of these nutrients can result in homocysteinemia [17, 28, 29]. Additionally, homocysteinemia can stimulate procoagulant factors and suppress anticoagulant factors, thereby enhancing thrombotic events [24, 30, 31].
We conducted further analysis and literature review regarding confounding factors in our study, which included age, gender, smoking, hypertension, diabetes, chronic kidney disease and serum vitamin B12 levels. A study involving 7872 subjects from the general population of China found that the Hcy level was significantly higher in males than in females in each age range [32]. Factors such as exposure to environmental cigarette smoking and alcohol consumption, genetic variation of methylene tetrahydrofolate reductase (MTHFR), and rates of re-methylation and transmethylation of Hcy may contribute to sexual differences in Hcy levels [33]. Our study revealed that Hcy posed a risk for both CVD death and all-cause mortality, and this risk was consistent across genders. Importantly, the degree of risk did not exhibit a statistically significant difference between the two gender groups. Elevated Hcy levels can lead to endothelial injury and smoking can exacerbate endothelial dysfunction by promoting oxidative stress and inflammation [34, 35]. Meanwhile, smoking is associated with lower levels of B-vitamins which are crucial for the metabolism of homocysteine [36]. The interaction between smoking and Hcy may enhance vascular inflammation and thrombus formation. The interaction between smoking and Hcy may enhance vascular inflammation and thrombus formation, leading to more severe cardiovascular conditions and ultimately higher mortality rates [34].
A cross-sectional study of the general population of China found that Hcy levels significantly increased after 50 years of age [37]. High levels of Hcy have been linked to the development and progression of age-associated disorders, including cardiovascular diseases, neurodegenerative disorders, and chronic kidney disease [38, 39]. It is important to acknowledge that the outcomes were negative in the age group of less than 60 years. This is likely attributable to the limited follow-up time, which resulted in a lower rate of endpoint events in this specific age group.
Hyperhomocysteinemia emerged as a risk factor for both all-cause mortality and cardiovascular death in both the smoking and nonsmoking groups. Notably, the elevated risk of cardiovascular death attributed to hyperhomocysteinemia was significantly greater in the smoking group when compared to the nonsmoking group, indicating that risk is further amplified in individuals who smoke [18]. Another study found that smokers with a plasma homocysteine level above 12µmol/L had a 12-fold increased risk of cardiovascular disease compared to non-smokers with normal plasma homocysteine levels [40].
In our study, we observed a higher proportion of hypertension among patients with elevated Hcy levels. One proposed mechanism for this association involves arteriolar constriction, renal dysfunction, increased sodium reabsorption, and heightened arterial stiffness [41]. Some studies have suggested that lowering Hcy levels with B vitamins can be an effective means of reducing blood pressure, especially in the management of drug-resistant hypertension [42]. Compared to the non-hypertensive group, Hcy was associated with a higher risk of all-cause mortalityd and CVD death in the hypertensive group, although no statistically significant differences were demonstrated. The relationship between Hcy and diabetes, particularly type 2 diabetes mellitus (T2DM), is complex. Some studies have found a positive association between elevated Hcy levels and T2DM, while others have found a negative association [43, 44]. In terms of diabetes complications, elevated Hcy levels have been associated with an increased risk of diabetic retinopathy [45].
Our study identified homocysteine Hcy as a risk factor for all-cause mortality irrespective of the presence of renal insufficiency. Hcy levels emerge as a consistent risk factor for both all-cause mortality and CVD mortality, irrespective of renal function.
Vitamin B12 is involved in the metabolic pathway that converts Hcy to methionine, an essential amino acid. This conversion process is catalyzed by the enzyme methionine synthase, which is dependent on Vitamin B12 [46]. Therefore, a deficiency in Vitamin B12 can lead to increased Hcy levels, leading to an increased risk of cardiovascular diseases [47]. In our study, we did not observe a statistically significant difference in the increased risk of cardiovascular death among individuals with higher vitamin B12 levels, implying that vitamin B12 supplementation might be considered an effective approach for reducing the risk of cardiovascular death.
Nevertheless, it’s essential to recognize certain limitations in our study. Our research is based on a cross-sectional observational design, which allows us to establish a correlation between Hcy levels and CVD risk but falls short of establishing a definitive causal relationship. While we can observe an association, causation cannot be implied. Furthermore, the limitations of the NHANES database, which provides data at specific time points, make it challenging to determine the precise temporal relationship between Hcy levels and the risk of mortality. A more comprehensive assessment of temporal relationships would necessitate longitudinal data with multiple time points. Additionally, in Table S1, we compared the groups with missing data and those included in the analysis. The results showed significant differences in age and mortality, but no significant differences in other aspects. This may introduce some selection bias into our study.
Our findings imply that Hcy levels might serve as potential predictors of all-cause mortality and CVD mortality. It is plausible that Hcy testing could be incorporated into clinical practice to assess cardiovascular risk and predict mortality. Additionally, tailored Hcy management strategies based on individual patient characteristics could be developed. Further research can delve into the pathophysiological mechanisms underlying the relationship between Hcy and CVD. Prospective clinical studies could be conducted to elucidate the causal relationship between Hcy and the risk of CVD. This could provide valuable insights into preventive and therapeutic strategies for cardiovascular diseases.
Conclusion
Our research supports the idea that serum Hcy levels should be considered a potential predictor of all-cause and CVD mortality. Reevaluating the clinical thresholds and conducting further research to explore the causative factors will contribute to a deeper understanding of the role of Hcy in cardiovascular health and inform strategies to reduce the burden of CVD-related mortality.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
The authors thank the participants of the NHANES databases.
Abbreviations
- BMI
Body Mass Index
- CDC
Centers for Disease Control and Prevention
- CKD
Chronic kidney disease
- CVD
Cardiovascular disease
- eGFR
Estimated glomerular filtration rate
- PIR
Poverty Income Ratio
- Hcy
Homocysteine
- MEC
Mobile Examination Centers
- NHANES
National Health and Nutrition Examination Survey
- NDI
National death index
Author contributions
LDH and FCS conceived the study design and are responsible for the overall content, analyzed and interpreted the data. LDH and WJ prepared the manuscript. TYQ drafted of the Figure, WJ prepared the tables, ZT reviewed the manuscript. All authors approved the submitted and final versions.
Funding
This research received the financial support by the National High Level Hospital Clinical Research Funding & Fundamental Research Funds for the Central Universities (Project Number: BJ-2022-192) and the Sichuan Provincial Health Commission Medical Science and Technology Project (Project Number:21PJ200).
Data availability
The NHANES data is publically available and can be downloaded from the following sites (https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?BeginYear=1999; https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?BeginYear=2001; https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?BeginYear=2003; https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?BeginYear=2005;)
Declarations
Ethics approval and consent to participate
The protocols of NHANES were approved by the institutional review board of the National Center for Health Statistics, CDC (https://www.cdc.gov/nchs/nhanes/). NHANES has obtained written informed consent from all participants.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Clinical trial number
Not applicable.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Donghao Liu, Chuangsen Fang and Jia Wang contributed equally to this work.
References
- 1.North BJ, Sinclair DA. The intersection between aging and cardiovascular disease. Circ Res. 2012;110(8):1097–108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Li Y, Cao GY, Jing WZ, Liu J, Liu M. Global trends and regional differences in incidence and mortality of cardiovascular disease, 1990–2019: findings from 2019 global burden of disease study. Eur J Prev Cardiol. 2023;30(3):276–86. [DOI] [PubMed] [Google Scholar]
- 3.Baszczuk A, Kopczyński Z. [Hyperhomocysteinemia in patients with cardiovascular disease]. Postepy Hig Med Dosw (Online). 2014;68:579–89. [DOI] [PubMed] [Google Scholar]
- 4.Marcus J, Sarnak MJ, Menon V. Homocysteine lowering and cardiovascular disease risk: lost in translation. Can J Cardiol. 2007;23(9):707–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ganguly P, Alam SF. Role of homocysteine in the development of cardiovascular disease. Nutr J. 2015;14:6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Peng HY, Man CF, Xu J, Fan Y. Elevated homocysteine levels and risk of cardiovascular and all-cause mortality: a meta-analysis of prospective studies. J Zhejiang Univ Sci B. 2015;16(1):78–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Zhang Z, Gu X, Fang X, Tang Z, Guan S, Liu H, Wu X, Wang C, Zhao Y. Homocysteine and the risk of Cardiovascular events and all-cause death in Elderly Population: A Community-based prospective cohort study. Ther Clin Risk Manag. 2020;16:471–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Wang Z, Klipfell E, Bennett BJ, Koeth R, Levison BS, Dugar B, Feldstein AE, Britt EB, Fu X, Chung YM, et al. Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease. Nature. 2011;472(7341):57–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kursa MB, Rudnicki WR. Feature selection with the Boruta Package. J Stat Softw. 2010;36(11):1–13. [Google Scholar]
- 10.NHANES 2001–2002. Data Documentation, Codebook, and Frequencies. Available online: https://wwwn.cdc.gov/Nchs/Nhanes/2001-2002/L06_2_B.htm
- 11.NHANES. 2003–2004 Data Documentation, Codebook, and Frequencies. Available online: https://wwwn.cdc.gov/Nchs/Nhanes/2003-2004/L06MH_C.htm
- 12.Fadeyev K, Nagao-Sato S, Reicks M. Nutrient and Food Group Intakes among U.S. Children (2–5 Years) Differ by Family Income to Poverty Ratio, NHANES 2011–2018. Int J Environ Res Public Health. 2021;18(22). [DOI] [PMC free article] [PubMed]
- 13.Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood and adult obesity in the United States, 2011–2012. JAMA. 2014;311(8):806–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Inker LA, Eneanya ND, Coresh J, Tighiouart H, Wang D, Sang Y, Crews DC, Doria A, Estrella MM, Froissart M, et al. New Creatinine- and cystatin C-Based equations to Estimate GFR without Race. N Engl J Med. 2021;385(19):1737–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kutner MH, Nachtsheim CJ, Neter J, Wasserman W. Applied linear regression models. Volume 4. McGraw-Hill/Irwin New York; 2004.
- 16.Harrell FE. General Aspects of Fitting Regression Models. In: Regression Modeling Strategies: With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis. Edited by Harrell JFE. Cham: Springer International Publishing; 2015 13–44.
- 17.Fan R, Zhang A, Zhong F. Association between homocysteine levels and all-cause mortality: a dose-response Meta-analysis of prospective studies. Sci Rep. 2017;7(1):4769. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Peng H-y, Man C-f, Xu J, Fan Y. Elevated homocysteine levels and risk of cardiovascular and all-cause mortality: a meta-analysis of prospective studies. J Zhejiang Univ Sci B. 2015;16(1):78–86. [DOI] [PMC free article] [PubMed]
- 19.Nygård O, Nordrehaug JE, Refsum H, Ueland PM, Farstad M, Vollset SE. Plasma homocysteine levels and mortality in patients with coronary artery disease. N Engl J Med. 1997;337(4):230–6. [DOI] [PubMed] [Google Scholar]
- 20.Toole JF, Malinow MR, Chambless LE, Spence JD, Pettigrew LC, Howard VJ, Sides EG, Wang C-H, Stampfer M. Lowering homocysteine in patients with ischemic stroke to prevent recurrent stroke, myocardial infarction, and DeathThe Vitamin Intervention for Stroke Prevention (VISP) Randomized Controlled Trial. JAMA. 2004;291(5):565–75. [DOI] [PubMed] [Google Scholar]
- 21.Krumdieck CL, Prince CW. Mechanisms of homocysteine toxicity on connective tissues: implications for the morbidity of aging. J Nutr. 2000;130(2S Suppl):s365–8. [DOI] [PubMed] [Google Scholar]
- 22.Perna AF, Ingrosso D, Lombardi C, Acanfora F, Satta E, Cesare CM, Violetti E, Romano MM, De Santo NG. Possible mechanisms of homocysteine toxicity. Kidney Int Suppl 2003(84):S137–140. [DOI] [PubMed]
- 23.Pushpakumar S, Kundu S, Sen U. Endothelial dysfunction: the link between homocysteine and hydrogen sulfide. Curr Med Chem. 2014;21(32):3662–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Huang T, Yuan G, Zhang Z, Zou Z, Li D. Cardiovascular pathogenesis in hyperhomocysteinemia. Asia Pac J Clin Nutr. 2008;17(1):8–16. [PubMed] [Google Scholar]
- 25.Yang Q, He GW. Imbalance of Homocysteine and H(2)S: Significance, Mechanisms, and Therapeutic Promise in Vascular Injury. Oxid Med Cell Longev. 2019:7629673. [DOI] [PMC free article] [PubMed]
- 26.Paganelli F, Mottola G, Fromonot J, Marlinge M, Deharo P, Guieu R, Ruf J. Hyperhomocysteinemia and Cardiovascular Disease: is the Adenosinergic System the Missing Link? Int J Mol Sci. 2021; 22(4). [DOI] [PMC free article] [PubMed]
- 27.Tinelli C, Di Pino A, Ficulle E, Marcelli S, Feligioni M. Hyperhomocysteinemia as a risk factor and potential nutraceutical target for certain pathologies. Front Nutr. 2019;6:49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Liu Y, Geng T, Wan Z, Lu Q, Zhang X, Qiu Z, Li L, Zhu K, Liu L, Pan A, et al. Associations of serum folate and vitamin B12 levels with Cardiovascular Disease Mortality among patients with type 2 diabetes. JAMA Netw Open. 2022;5(1):e2146124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Verhaar MC, Stroes E, Rabelink TJ. Folates and Cardiovascular Disease. ATVB. 2022;22(1):6–13. [DOI] [PubMed]
- 30.Undas A, Brozek J, Szczeklik A. Homocysteine and thrombosis: from basic science to clinical evidence. Thromb Haemost. 2005;94(5):907–15. [DOI] [PubMed] [Google Scholar]
- 31.Eldibany MM, Caprini JA. Hyperhomocysteinemia and thrombosis: an overview. Arch Pathol Lab Med. 2007;131(6):872–84. [DOI] [PubMed]
- 32.Xu R, Huang F, Wang Y, Liu Q, Lv Y, Zhang Q. Gender- and age-related differences in homocysteine concentration: a c ross-sectional study of the general population of China. Sci Rep. 2020;10(1):17401. [DOI] [PMC free article] [PubMed]
- 33.Zhao J, Li Z, Hou C, Sun F, Dong J, Chu X, Guo Y. Gender differences in risk factors for high plasma homocysteine levels based on a retrospective checkup cohort using a generalized estimating equation analysis. Lipids Health Dis. 2021;20(1):31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Sun J, Han W, Wu S, Jia S, Yan Z, Guo Y, Zhao Y, Zhou Y, Liu X. Combined effect of hyperhomocysteinemia and smoking on the severity of coronary artery disease in young adults ≤ 35 years of age: a hospital-based observational study. BMC Cardiovasc Disord. 2021;21(1):484. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Yuan D, Chu J, Lin H, Zhu G, Qian J, Yu Y, Yao T, Ping F, Chen F, Liu X. Mechanism of homocysteine-mediated endothelial injury and its conseque nces for atherosclerosis. Front Cardiovasc Med. 2023;9. [DOI] [PMC free article] [PubMed]
- 36.O’Callaghan P, Meleady R, Fitzgerald T, Graham I. Smoking and plasma homocysteine. Eur Heart J. 2002;23(20):1580–6. [DOI] [PubMed] [Google Scholar]
- 37.Xu R, Huang F, Wang Y, Liu Q, Lv Y, Zhang Q. Gender- and age-related differences in homocysteine concentration: a cross-sectional study of the general population of China. Sci Rep. 2020;10(1):17401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Ostrakhovitch EA, Tabibzadeh S. Homocysteine and age-associated disorders. Ageing Res Rev. 2019;49:144–64. [DOI] [PubMed] [Google Scholar]
- 39.Kuo H-K, Sorond FA, Chen J-H, Hashmi A, Milberg WP, Lipsitz LA. The role of homocysteine in multisystem age-related problems: a system atic review. J Gerontol Biol Sci Med Sci. 2005;60(9):1190–201. [DOI] [PubMed]
- 40.O’Callaghan P. Smoking and plasma homocysteine. Eur Heart J. 2022;23(20):1580–6. [DOI] [PubMed]
- 41.Stehouwer CD, van Guldener C. Does homocysteine cause hypertension? Clin Chem Lab Med. 2003;41(11):1408–11. [DOI] [PubMed] [Google Scholar]
- 42.Elias MF, Brown CJ. New evidence for Homocysteine lowering for management of treatment-res istant hypertension. Am J Hypertens. 2022;35(4):303–5. [DOI] [PMC free article] [PubMed]
- 43.Platt DE, Hariri E, Salameh P, Merhi M, Sabbah N, Helou M, Mouzaya F, Nemer R, Al-Sarraj Y, El-Shanti H et al. Type II diabetes mellitus and hyperhomocysteinemia: a complex interact ion. Diabetol Metab Syndr. 2017;9:19. [DOI] [PMC free article] [PubMed]
- 44.Huang T, Ren J, Huang J, Li D. Association of homocysteine with type 2 diabetes: a meta-analysis implementing mendelian randomization approach. BMC Genomics. 2013;14:867. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Lei X, Zeng G, Zhang Y, Li Q, Zhang J, Bai Z, Yang K. Association between homocysteine level and the risk of diabetic retinopathy: a systematic review and meta-analysis. Diabetol Metab Syndr. 2018;10:61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Fratoni V, Brandi ML. B vitamins, homocysteine and bone health. Nutrients. 2015;7(4):2176–92. [DOI] [PMC free article] [PubMed]
- 47.Stabler SP. Clinical practice. Vitamin B12 deficiency. N Engl J Med. 2013;368(2):149–60. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The NHANES data is publically available and can be downloaded from the following sites (https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?BeginYear=1999; https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?BeginYear=2001; https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?BeginYear=2003; https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?BeginYear=2005;)