Background
Homocysteine levels are closely associated with overall health. An elevation in homocysteine can lead to a variety of health issues and is influenced by vitamins B12, B6, and folic acid. However, the dose-response relationship between homocysteine and vitamin A remains unclear among the middle-aged and elderly population in rural China.
Objective
This study investigates the dose-response relationship and threshold effect between serum homocysteine and vitamin A levels in vivo. Utilizing large-scale data from middle-aged and elderly individuals in China, it also evaluates these relationships across gender and age subgroups, providing a scientific basis for preventing and treating hyperhomocysteinemia in this population.
Methods
To examine the dose-response relationship between vitamin A and homocysteine, we used multiple linear regression, multinomial logistic regression, generalized additive models, and restricted cubic splines. This analysis involved 28,860 middle-aged to elderly participants from rural China.
Result
The distribution of vitamin A and homocysteine levels is similar across different genders and age groups. However, homocysteine levels are elevated in men and individuals over the age of 65. A U-shaped dose-response relationship exists between these two variables, independent of gender or age. In China, the vitamin A levels among middle-aged and elderly populations are categorized into quintiles. As the concentration of vitamin A increases, the β coefficient exhibits a trend that initially decreases before subsequently increasing.The serum levels of both vitamin A and homocysteine across various genders and age groups demonstrate a U-shaped relationship. Specifically, as serum vitamin A levels rise, homocysteine levels first decline before rising again. This U-shaped nonlinear association between vitamin A and homocysteine was also identified within the high-homocysteine group. Restricted spline analysis revealed that with increasing serum vitamin A concentrations, the odds ratio (OR) for hyperhomocysteinemia initially decreased before increasing.Notably, when serum vitamin A concentration reaches 0.514 µg/ml, the OR for hyperhomocysteinemia exceeds 1.
Conclusion
This study found a U-shaped relationship between serum vitamin A and homocysteine levels in middle-aged and elderly Chinese people. Proper vitamin A levels may help reduce homocysteine in those with hyperhomocysteinemia. The research provides a scientific basis for understanding this correlation and supports precise prevention and management of hyperhomocysteinemia.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12937-026-01284-z.
Keywords: Homocysteine, Vitamin a dose response, Prevention
Highlights
There was a dose-response relationship between serum vitamin A and serum homocysteine levels in the Chinese population.
There exists a U-shaped dose-response relationship between serum vitamin A levels and homocysteine levels in the Chinese population, as well as among different subgroups, and in the population with hyperhomocysteinemia.
When the plasma vitamin A content exceeds 0.514 µg/mL, vitamin A is a risk factor for hyperhomocysteinemia.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12937-026-01284-z.
Introduction
Homocysteine serves as a critical intermediate in the methionine metabolic pathway within human metabolism. Its metabolism primarily depends on the catalytic actions of B vitamins, occurring via remethylation and transsulfuration pathways [1].
Clinically, the reference range for homocysteine is 5–15 µmol/L, and elevated levels in blood are diagnosed as hyperhomocysteinemia [2]. Elevated homocysteine levels have been positively correlated with an increased risk of vascular diseases. Specifically, high homocysteine levels induce endothelial oxidative stress, promote vascular inflammation, and elevate the risk of hypertension [3, 4]. Research indicates that hyperhomocysteinemia is an independent risk factor for hypertension, with approximately 75% of hypertensive patients exhibiting hyperhomocysteinemia. Furthermore, it is strongly associated with coronary artery disease, myocardial infarction, and stroke. Patients with both hypertension and hyperhomocysteinemia face a stroke risk over tenfold higher than the general population [5–7].
From 2020 to 2022, the “Cardiovascular Disease and Its Risk Factors Monitoring in Chinese Residents” project surveyed 298,438 individuals, revealing a hypertension prevalence rate of 31.6% among adult residents [8]. A prospective cohort study of 12,952 adults conducted by the China Health and Nutrition Survey (CHNS) demonstrated that the age-standardized incidence of hypertension increased from 40.8% during 1993–1997 to 48.6% during 2011–2015 [9]. Additionally, the CHS survey on hypertension in China estimated that the number of residents aged 18 years and older with high-normal blood pressure reached approximately 435 million between 2012 and 2015 [10]. Recruitment of 515,047 Chinese people in the 2023 Chinese prospective cohort study on nutrition and chronic diseases (CPNAS) documented that hyperhomocysteinemia is most common in this cohort [11].
The World Health Organization (WHO) has established clinical diagnostic criteria for serum vitamin A levels, defining a deficiency as a serum retinol concentration of less than 0.7 µmol/L. Marginal deficiency is characterized by concentrations ranging from 0.70 to 1.05 µmol/L, while concentrations greater than or equal to 1.05 µmol/L are considered normal. Vitamin A, in the form of retinol, is absorbed and transported via specific binding proteins. Retinol plays a critical role in various physiological functions, including maintaining normal vision, promoting growth and development, and preserving the health of skin and mucous membranes [12]. Deficiency in vitamin A can result in conditions such as night blindness, dry eyes, and keratomalacia, whereas excessive intake can lead to adverse effects like hepatic cirrhosis, increased fracture risk, headaches, and neurological symptoms such as vertigo [13, 14].
It has been demonstrated that the metabolism of homocysteine is dependent on specific B vitamins. While vitamin A does not directly participate in homocysteine metabolism, it may indirectly influence homocysteine levels by modulating the utilization or metabolic efficiency of these B vitamins. Additionally, there may be interactions between vitamin A and B vitamins in maintaining physiological function and health status [15, 16]. Research indicates that micronutrient and precision nutrition studies have significant potential in the prevention and treatment of chronic diseases [17]. Through targeted nutritional interventions, it is possible to enhance individual health, prevent disease, and promote chronic disease management as well as social development [18].
The objective of this study is to provide a scientific foundation for the prevention and management of hyperhomocysteinemia, as well as to inform the development of precise nutritional strategies tailored to the Chinese population.
Method
Participants
The participants in this study were sourced from five distinct survey databases: firstly, the 2017 National Nutrition and Hypertension Survey; secondly, the 2016–2018 Nested Case-Control Study (NCC) on Nutrition and Stroke in Lianyungang; thirdly, the 2017 Nested Case-Control Study (NCC) on Nutrition and Stroke in Rongcheng; fourthly, the Chinese Stroke Primary Prevention Trial (CSPPT) Nested Case-Control Study (NCC); and fifthly, the Chinese Precision Nutrition and Health Knowledge, Attitude, and Practice Real-World Study (CPNAS) - Nutrophin Study.
2017 National Nutrition and Hypertension Survey [19]. This cross-sectional study examined 2,609 middle-aged and elderly hypertensive patients across 14 provinces in China from February 2017 to May 2018.
Nested Case-Control Study on Nutrition and Stroke [20]: Utilizing data from the Chinese Type H Hypertension Registry, this nested case-control study included 3,327 first-time ischemic stroke cases and an equal number of controls matched by age, sex, and village. The study was conducted in Lianyungang (2016–2018) and Rongcheng (2017), with respective sample sizes of 4,144 and 2,510 participants.
Chinese Stroke Primary Prevention Trial (CSPPT) - Nested Case-Control Study (NCC) [21, 22]: Conducted from 2008 to 2013, this nested case-control study utilized data from the CSPPT, ultimately enrolling a total of 6,197 participants.
Chinese Precision Nutrition and Health KAP Real World Study (CPNAS) - Nutrophin Study [23] (ChiCTR2100051983||http://www.chictr.org.cn/;The registration date is August 10, 2021): This study was carried out in five provinces: Shandong, Jiangsu, Hubei, Guizhou, and Ningxia. A subset of 13,400 participants was selected from the total 396,379 CPNAS participants for the nutritional biomarker substudy [11].
Participants were pooled from five population-based datasets and merged into a unified nutrient database (n = 28,860). Individuals who reported folic acid supplementation were first excluded (n = 2,918). Participants with missing values for vitamin A (VA) (n = 48) or homocysteine (HCY) (n = 140) were subsequently removed. Individuals younger than 18 years and those with extreme outlier values of VA or HCY (|z| > 3) were also excluded (n = 771). Further exclusions were applied to participants with missing folic acid (FA) data (n = 6,686), missing vitamin B12 data (n = 19), or missing values for other covariates (n = 236). After all exclusion steps, a total of 18,042 participants were included in the final analysis.The process is shown in the Fig. 1.
Fig. 1.
subjet screening flow chart
Laboratory measurements
A fasting venous blood sample was collected from each participant. Biomarkers were measured in plasma or serum, and plasma samples were separated within 30 min of collection and subsequently stored at -80 ℃.All test indicators were carefully recorded, including folic acid, vitamin B12, vitamin D, total cholesterol (CHOL), low-density lipoprotein cholesterol (LDL-C), triglycerides (TG), uric acid (UA), homocysteine (HCy), creatinine (CREA), and glucose (GLU), all of which were measured based on serum levels. Notably, the adjustment factors in the analysis did not account for dietary intake.
Vitamin A
Vitamins A was analyzed using High Performance Liquid Chromatography-Tandem Mass Spectrometry (HPLC-MS/MS) with the Kinetex®C18 100 A (3 × 75 mm,2.6 μm) Liquid Chromatography Column, via multiple reaction monitoring (MRM) and isotope internal standard quantification [24].
Homocysteine
Homocysteine levels were measured using a fully automated biochemical analyzer (Beckman Coulter, Inc., 250 S. Kraemer Blvd., Brea, CA 92821, USA). For this study, homocysteine was detected in serum samples. According to clinical diagnostic criteria, a homocysteine level within the range of 5–15 µmol/L is considered normal, while levels exceeding 15 µmol/L are classified as hyperhomocysteinemia.
The following biomarkers were quantified in serum samples: alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (TBIL), uric acid (UA), creatinine (CREA), glucose (GLU), total cholesterol (CHOL), triglycerides (TG), and low-density lipoprotein cholesterol (LDL-C). All serum biomarkers were analyzed using a fully automated biochemical analyzer (Beckman Coulter, Inc., Brea, CA, USA).
Covariate
To more comprehensively elucidate the relationship between vitamin A and homocysteine, we conducted multivariate adjustments for potential confounders. In Model 1, adjustments were made for age (categorized as < 65 years and ≥ 65 years), sex (male or female), body mass index (BMI), smoking status (never, former, current), and alcohol consumption (never, former, current). Building upon Model 1, Model 2 further adjusted for folate (FA), vitamin B12 (supplemented or not), methylene tetrahydrofolate reductase (MTHFR) C677T genotype (presence or absence), hypertension (presence or absence), stroke (presence or absence), and coronary heart disease (presence or absence).
Statistical analysis
For baseline characteristics, continuous variables that follow a normal or approximately normal distribution are presented as mean ± standard deviation and compared using a T-test. Continuous variables with non-normal distributions are reported as median (interquartile range: 25th-75th percentile) and analyzed using the Wilcoxon Rank-Sum test for two groups or the Kruskal-Wallis test for more than two groups. Categorical variables are summarized as counts (percentages) and compared using the Chi-square test for two groups or ANOVA for more than two groups to examine the distribution of continuous variables within the study population, which is visualized using density plots.
Multivariate linear regression models, both unadjusted and adjusted, were employed to quantify the association between vitamin A and homocysteine levels. Smoothing curves were utilized to investigate dose-response relationships and to identify any potential nonlinear associations or confounding factors. In this study, we applied a generalized additive model (GAM) with cubic spline smoothing to simulate the relationship between the predictor and the response variable.
Multivariate logistic regression analysis was conducted to evaluate the association between vitamin A and serum homocysteine levels. Vitamin A was treated as a continuous variable, with the Q2 group serving as the reference category. This approach allowed for the fitting of the nonlinear relationship between vitamin A intake and the risk of hyperhomocysteinemia.All statistical analyses were conducted using R Software (version 4.1.1).
Result
It is concluded that the distribution of vitamin A and homocysteine levels exhibits similar morphological characteristics across different genders and age groups. Table 1 Basic characteristics of the study population based on different sexes and ages under vitamin A. Figure 2 illustrates the density distributions of vitamin A and homocysteine in the general population. Figure 3 presents the smooth fitting curves for vitamin A and homocysteine levels, adjusted for variables including age, sex, body mass index (BMI), smoking status, alcohol consumption, folic acid (FA), vitamin B12, methyltetrahydrofolate reductase (MTHFR) C677T genotype, hypertension, stroke, and coronary heart disease. The homocysteine levels were found to be higher in males compared to females. Additionally, individuals over 65 years of age exhibited higher homocysteine levels than those under 65. The dose-response relationship between vitamin A and homocysteine was U-shaped, irrespective of gender or age.
Table 1.
Basic characteristics of the study population based on different sexes and ages under vitamin A
| Variables | Level | Overall | Male | Age<40 | 40 ≤ Age<65 | Age ≥ 65 | p.value | Female | Age<40 | 40 ≤ Age<65 | Age ≥ 65 | p.value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n | 18,042 | 8155 | 130 | 3859 | 4166 | 9887 | 113 | 5460 | 4314 | |||
| BMI (mean (SD)) | 25.33 (3.81) | 24.84 (3.64) | 26.26 (3.79) | 25.50 (3.58) | 24.19 (3.57) | < 0.001 | 25.73 (3.89) | 23.98 (4.08) | 25.93 (3.88) | 25.54 (3.88) | < 0.001 | |
| Smoke.new (%) | Never | 12,743 (70.6) | 3103 (38.1) | 62 (47.7) | 1369 (35.5) | 1672 (40.1) | < 0.001 | 9640 (97.5) | 112 (99.1) | 5357 (98.1) | 4171 (96.7) | < 0.001 |
| Formerly | 1513 ( 8.4) | 1452 (17.8) | 6 ( 4.6) | 607 (15.7) | 839 (20.1) | 61 ( 0.6) | 1 ( 0.9) | 20 ( 0.4) | 40 ( 0.9) | |||
| Currently | 3786 (21.0) | 3600 (44.1) | 62 (47.7) | 1883 (48.8) | 1655 (39.7) | 186 ( 1.9) | 0 ( 0.0) | 83 ( 1.5) | 103 ( 2.4) | |||
| alcohol.new (%) | Never | 12,761 (70.7) | 3400 (41.7) | 56 (43.1) | 1480 (38.4) | 1864 (44.7) | < 0.001 | 9361 (94.7) | 92 (81.4) | 5135 (94.0) | 4134 (95.8) | < 0.001 |
| Formerly | 922 ( 5.1) | 806 ( 9.9) | 5 ( 3.8) | 341 ( 8.8) | 460 (11.0) | 116 ( 1.2) | 1 ( 0.9) | 70 ( 1.3) | 45 ( 1.0) | |||
| Currently | 4359 (24.2) | 3949 (48.4) | 69 (53.1) | 2038 (52.8) | 1842 (44.2) | 410 ( 4.1) | 20 (17.7) | 255 ( 4.7) | 135 ( 3.1) | |||
| FA (mean (SD)) | 8.18 (5.04) | 7.61 (5.10) | 9.20 (6.58) | 7.88 (5.19) | 7.30 (4.95) | < 0.001 | 8.66 (4.94) | 9.77 (5.69) | 8.79 (4.81) | 8.47 (5.07) | < 0.001 | |
| B12 (mean (SD)) | 439.09 (185.66) | 426.05 (186.98) | 391.27 (145.08) | 427.82 (184.53) | 425.49 (190.30) | 0.087 | 449.85 (183.87) | 365.77 (142.73) | 456.82 (182.72) | 443.24 (185.52) | < 0.001 | |
| C677T (%) | CC | 4571 (25.3) | 2155 (26.4) | 37 (28.5) | 1018 (26.4) | 1100 (26.4) | 0.278 | 2416 (24.4) | 27 (23.9) | 1329 (24.3) | 1060 (24.6) | 0.551 |
| CT | 8592 (47.6) | 3934 (48.2) | 53 (40.8) | 1842 (47.7) | 2039 (48.9) | 4658 (47.1) | 46 (40.7) | 2583 (47.3) | 2029 (47.0) | |||
| TT | 4879 (27.0) | 2066 (25.3) | 40 (30.8) | 999 (25.9) | 1027 (24.7) | 2813 (28.5) | 40 (35.4) | 1548 (28.4) | 1225 (28.4) | |||
| hypertension (%) | No | 5690 (31.5) | 2441 (29.9) | 50 (38.5) | 1102 (28.6) | 1289 (30.9) | 0.007 | 3249 (32.9) | 81 (71.7) | 2089 (38.3) | 1079 (25.0) | < 0.001 |
| Yes | 12,352 (68.5) | 5714 (70.1) | 80 (61.5) | 2757 (71.4) | 2877 (69.1) | 6638 (67.1) | 32 (28.3) | 3371 (61.7) | 3235 (75.0) | |||
| Stroke (%) | No | 17,456 (96.8) | 7850 (96.3) | 127 (97.7) | 3735 (96.8) | 3988 (95.7) | 0.03 | 9606 (97.2) | 112 (99.1) | 5360 (98.2) | 4134 (95.8) | < 0.001 |
| Yes | 586 ( 3.2) | 305 ( 3.7) | 3 ( 2.3) | 124 ( 3.2) | 178 ( 4.3) | 281 ( 2.8) | 1 ( 0.9) | 100 ( 1.8) | 180 ( 4.2) | |||
| CHD (%) | No | 16,864 (93.5) | 7620 (93.4) | 125 (96.2) | 3667 (95.0) | 3828 (91.9) | < 0.001 | 9244 (93.5) | 112 (99.1) | 5280 (96.7) | 3852 (89.3) | < 0.001 |
| Yes | 1178 ( 6.5) | 535 ( 6.6) | 5 ( 3.8) | 192 ( 5.0) | 338 ( 8.1) | 643 ( 6.5) | 1 ( 0.9) | 180 ( 3.3) | 462 (10.7) | |||
| VA (mean (SD)) | 0.60 (0.17) | 0.64 (0.18) | 0.63 (0.15) | 0.67 (0.19) | 0.61 (0.18) | < 0.001 | 0.57 (0.15) | 0.47 (0.12) | 0.58 (0.16) | 0.56 (0.15) | < 0.001 | |
| Hcy (mean (SD)) | 13.32 (4.40) | 14.47 (4.73) | 15.08 (6.34) | 14.00 (4.69) | 14.88 (4.67) | < 0.001 | 12.37 (3.86) | 10.79 (3.90) | 11.69 (3.69) | 13.28 (3.88) | < 0.001 |
For continuous variables, values are presented as mean ± SD. BMI, body mass index (kg/m2) Hcy homocysteine(µmol/L), CHD Coronary Heart Disease, C677T Methylenetetrahydrofolate Reductase
Fig. 2.
Density profiles of vitamin A and homogeneity
Fig. 3.
Smooth fitting curve between plasma VA and HCY (Adjustedforage, sex, BMI, smoking, alcohol, FA, VB12, MTHFRC677T, hypertension, stroke, coronary heart disease)
The association between serum vitamin A levels and homocysteine (Hcy) concentrations was assessed using linear regression models, with adjustment for the aforementioned covariates in both Model 1 and Model 2. Table 2 displays the results of hierarchical linear regression analyses examining the relationship between serum vitamin A and Hcy, stratified by age groups (under 40, 40–65, and ≥ 65 years) and gender. Due to the small sample sizes among men and women under 40 years of age, subgroup analyses by sex in this age group were not statistically reliable; therefore, data for this group were not stratified by gender. The findings indicated significant gender-specific differences when analyses were stratified by age at 40 and 65 years: women exhibited greater sensitivity to vitamin A levels, with both low (< 0.4 µg/mL) and high concentrations associated with elevated Hcy levels. In contrast, men demonstrated a more pronounced increase in Hcy at higher vitamin A levels. With respect to age, the strongest association was observed in individuals aged 40–65 years, whereas among older adults (≥ 65 years), a typical U-shaped relationship was evident. Table 3 summarizes the results of logistic regression analyses evaluating the association between serum vitamin A levels and the risk of hyperhomocysteinemia (defined as Hcy ≥ 15 µmol/L). The optimal range of serum vitamin A appears to lie within the second and third quartiles (Q2–Q3, 0.5–0.6 µg/mL), where the risk of hyperhomocysteinemia is lowest. Additional results from linear regression analyses of the association between serum vitamin A and Hcy concentrations, stratified by gender and age (≥ 65 years), are presented in Table 4.
Table 2.
Linear regression analysis of plasma VA in relation to Hcy (Stratified by sex and age 40,65)
| VA, µg/mL | N | Median.IQR. | Unadjusted | Model 1 | Model 2 | |||
|---|---|---|---|---|---|---|---|---|
| β (95%CI) | P.crude | β (95%CI) | P | β (95%CI) | P | |||
| Age < 40 | 243 | 11.6(9.2–14.6) | 5.63(1.12,10.14) | 0.015 | -1.99(-6.90,2.92) | 0.428 | -2.07(-6.58,2.44) | 0.369 |
| Quartile1 | ||||||||
| Q1 (< 0.4) | 49 | 10.3(8.2–13.2) | -1.84(-4.09,0.41) | 0.11 | 0.12(-2.12,2.36) | 0.917 | 1.12(-0.99,3.23) | 0.301 |
| Q2 (0.4 ~ 0.5) | 48 | 10.6(9.1–12.8) | -2.33(-4.60,-0.07) | 0.044 | -1.20(-3.38,0.98) | 0.281 | -0.40(-2.45,1.66) | 0.705 |
| Q3 (0.5 ~ 0.6) | 49 | 12(9.7–15.8) | reference | reference | reference | |||
| Q4 (0.6 ~ 0.7) | 48 | 12(9.7–16) | -0.08(-2.34,2.18) | 0.945 | -0.82(-3.02,1.38) | 0.467 | 0.11(-1.93,2.16) | 0.914 |
| Q5 (≥ 0.7) | 49 | 13.1(10.5–16) | 0.57(-1.68,2.82) | 0.62 | -0.85(-3.03,1.32) | 0.442 | -0.15(-2.18,1.87) | 0.883 |
| P for trend | 0.266 | 0.292 | 0.622 | |||||
| Age 40–65 | 9319 | 11.7(10.3–13.8) | 2.95(2.46,3.45) | < 0.001 | 1.36(0.86,1.87) | < 0.001 | 0.83(0.36,1.30) | < 0.001 |
| Quartile2 | ||||||||
| Q1 (< 0.5) | 1852 | 11.1(9.6–13.3) | -0.29(-0.57,-0.02) | 0.036 | -0.01(-0.28,0.25) | 0.921 | -0.05(-0.30,0.19) | 0.674 |
| Q2 (0.5 ~ 0.6) | 1859 | 11.3(10.1–13.3) | -0.22(-0.49,0.05) | 0.115 | -0.12(-0.38,0.14) | 0.373 | -0.10(-0.35,0.14) | 0.422 |
| Q3 (0.6 ~ 0.6 | 1878 | 11.5(10.2–13.6) | reference | reference | reference | |||
| Q4 (0.6 ~ 0.8) | 1860 | 11.9(10.5–14.1) | 0.45(0.17,0.72) | 0.001 | 0.28(0.01,0.54) | 0.041 | 0.14(-0.10,0.39) | 0.257 |
| Q5 (≥ 0.8) | 1870 | 12.5(10.9–14.8) | 1.16(0.88,1.43) | < 0.001 | 0.62(0.35,0.89) | < 0.001 | 0.32(0.07,0.58) | 0.012 |
| P for trend | < 0.001 | < 0.001 | 0.006 | |||||
| Male Age < 65 | 3989 | 12.8(11.1–15.4) | 0.86(0.06,1.65) | 0.035 | 1.15(0.33,1.97) | 0.006 | 0.53(-0.23,1.30) | 0.17 |
| Quartile3 | ||||||||
| Q1 (< 0.5) | 798 | 12.6(10.9–15.4) | 0.29(-0.17,0.76) | 0.218 | 0.20(-0.27,0.66) | 0.413 | 0.08(-0.35,0.51) | 0.728 |
| Q2 (0.5 ~ 0.6) | 795 | 12.6(11–15) | 0.11(-0.36,0.57) | 0.651 | 0.07(-0.40,0.54) | 0.769 | 0.06(-0.37,0.48) | 0.798 |
| Q3 (0.6 ~ 0.7) | 800 | 12.7(11–15) | reference | reference | reference | |||
| Q4 (0.7 ~ 0.8) | 795 | 12.8(11.3–15.1) | 0.26(-0.21,0.72) | 0.28 | 0.30(-0.17,0.76) | 0.213 | 0.15(-0.28,0.58) | 0.495 |
| Q5 (≥ 0.8) | 801 | 13.4(11.5–16.3) | 0.91(0.44,1.37) | < 0.001 | 0.96(0.50,1.43) | < 0.001 | 0.55(0.11,0.98) | 0.014 |
| P for trend3 | < 0.001 | < 0.001 | 0.022 | |||||
| Female Age < 65 | 4166 | 13.7(11.8–16.6) | 0.65(-0.15,1.44) | 0.114 | 1.27(0.45,2.09) | 0.002 | 0.78(-0.00,1.55) | 0.05 |
| Quartile4 | ||||||||
| Q1 (< 0.5) | 828 | 13.8(11.7–17.1) | 0.72(0.27,1.17) | 0.002 | 0.57(0.12,1.02) | 0.013 | 0.49(0.07,0.91) | 0.023 |
| Q2 (0.5 ~ 0.5) | 836 | 13.7(11.7–16.4) | 0.40(-0.05,0.84) | 0.082 | 0.32(-0.13,0.76) | 0.164 | 0.22(-0.19,0.64) | 0.297 |
| Q3 (0.5 ~ 0.6) | 831 | 13.2(11.6–15.9) | reference | reference | reference | |||
| Q4 (0.6 ~ 0.7) | 835 | 13.7(11.8–16.3) | 0.31(-0.14,0.76) | 0.177 | 0.39(-0.05,0.84) | 0.084 | 0.19(-0.22,0.61) | 0.361 |
| Q5 (≥ 0.7) | 836 | 14.4(12.2–17.3) | 1.00(0.56,1.45) | < 0.001 | 1.16(0.71,1.61) | < 0.001 | 0.80(0.37,1.22) | < 0.001 |
| P for trend | 0.002 | < 0.001 | 0.007 | |||||
| Male Age ≥ 65 | 5573 | 11(9.8–12.8) | 1.63(1.01,2.25) | < 0.001 | 1.54(0.91,2.17) | < 0.001 | 1.01(0.42,1.61) | < 0.001 |
| Quartile5 | ||||||||
| Q1 (< 0.4) | 1112 | 10.7(9-12.6) | -0.07(-0.38,0.24) | 0.653 | -0.03(-0.34,0.27) | 0.83 | -0.06(-0.34,0.23) | 0.694 |
| Q2 (0.4 ~ 0.5) | 1113 | 10.8(9.6–12.5) | -0.00(-0.31,0.30) | 0.978 | 0.01(-0.30,0.32) | 0.953 | 0.01(-0.28,0.29) | 0.956 |
| Q3 (0.5 ~ 0.6) | 1112 | 10.9(9.8–12.5) | reference | reference | reference | |||
| Q4 (0.6 ~ 0.7) | 1121 | 11.2(10.1–12.8) | 0.28(-0.02,0.59) | 0.072 | 0.28(-0.02,0.59) | 0.071 | 0.24(-0.04,0.53) | 0.092 |
| Q5 (≥ 0.7) | 1115 | 11.4(10.3–13.3) | 0.72(0.42,1.03) | < 0.001 | 0.71(0.41,1.02) | < 0.001 | 0.42(0.13,0.71) | 0.004 |
| P for trend | < 0.001 | < 0.001 | < 0.001 | |||||
| Female Age ≥ 65 | 4314 | 12.3(10.9–14.5) | 0.83(0.08,1.59) | 0.03 | 0.96(0.20,1.71) | 0.013 | 0.91(0.18,1.64) | 0.015 |
| Quartile6 | ||||||||
| Q1 (< 0.4) | 859 | 12.3(10.8–14.9) | 0.62(0.26,0.99) | < 0.001 | 0.56(0.19,0.92) | 0.003 | 0.52(0.17,0.87) | 0.004 |
| Q2 (0.4 ~ 0.5) | 855 | 12.1(10.8–14.2) | 0.15(-0.21,0.52) | 0.414 | 0.17(-0.20,0.53) | 0.369 | 0.22(-0.13,0.57) | 0.209 |
| Q3 (0.5 ~ 0.6) | 874 | 12.1(10.8–14.2) | reference | reference | reference | |||
| Q4 (0.6 ~ 0.7) | 862 | 12.2(10.8–14.1) | 0.13(-0.23,0.49) | 0.482 | 0.15(-0.21,0.52) | 0.41 | 0.23(-0.12,0.58) | 0.204 |
| Q5 (≥ 0.7) | 864 | 12.8(11.3–15.3) | 0.94(0.58,1.30) | < 0.001 | 0.93(0.57,1.29) | < 0.001 | 0.86(0.51,1.21) | < 0.001 |
| P for trend | < 0.001 | < 0.001 | < 0.001 | |||||
Due to the small sample size, the data for men and women under 40 years of age did not reach statistical significance; therefore, no gender-based differentiation was made for individuals aged 19–39. The primary focus of the study is on middle-aged and elderly populations, particularly those aged 65 and above
Table 3.
Logistic regression analysis of serum vitamin A levels and the risk of hyperhomocysteinemia (Hcy ≥ 15µmol/L)
| VA, µg/mL | N | Median.IQR. | Unadjusted | Model 1 | Model 2 | |||
|---|---|---|---|---|---|---|---|---|
| β (95%CI) | P.crude | β (95%CI) | P | β (95%CI) | P | |||
| Age < 40 | 243 | 57(23.5) | 7.59(1.24,46.49) | 0.028 | 1.02(0.12,9.04) | 0.984 | 1.10(0.08,14.34) | 0.94 |
| Quartile1 | ||||||||
| Q1 (< 0.4) | 49 | 9(18.4) | 1.57(0.51,4.83) | 0.427 | 2.22(0.68,7.31) | 0.189 | 2.14(0.63,7.32) | 0.225 |
| Q2 (0.4 ~ 0.5) | 48 | 6(12.5) | reference | reference | reference | |||
| Q3 (0.5 ~ 0.6) | 49 | 13(26.5) | 2.53(0.87,7.33) | 0.088 | 1.81(0.59,5.58) | 0.302 | 1.30(0.37,4.56) | 0.682 |
| Q4 (0.6 ~ 0.7) | 48 | 13(27.1) | 2.60(0.90,7.55) | 0.079 | 1.60(0.50,5.08) | 0.424 | 1.54(0.44,5.38) | 0.499 |
| Q5 (≥ 0.7) | 49 | 16(32.7) | 3.39(1.20,9.63) | 0.022 | 1.65(0.54,5.10) | 0.383 | 1.48(0.42,5.21) | 0.545 |
| P for trend | 0.012 | 0.599 | 0.675 | |||||
| Age 40–65 | 9319 | 1636(17.6) | 3.49(2.60,4.69) | < 0.001 | 1.72(1.25,2.36) | < 0.001 | 1.45(1.04,2.03) | 0.03 |
| Quartile2 | ||||||||
| Q1 (< 0.5) | 1852 | 282(15.2) | 1.13(0.94,1.36) | 0.176 | 1.24(1.03,1.50) | 0.024 | 1.20(0.99,1.47) | 0.066 |
| Q2 (0.5 ~ 0.6) | 1859 | 254(13.7) | reference | reference | reference | |||
| Q3 (0.6 ~ 0.6) | 1878 | 309(16.5) | 1.24(1.04,1.49) | 0.017 | 1.19(0.99,1.43) | 0.062 | 1.20(0.99,1.46) | 0.062 |
| Q4 (0.6 ~ 0.8) | 1860 | 349(18.8) | 1.46(1.22,1.74) | < 0.001 | 1.30(1.08,1.56) | 0.005 | 1.24(1.03,1.50) | 0.027 |
| Q5 (≥ 0.8) | 1870 | 442(23.6) | 1.96(1.65,2.32) | < 0.001 | 1.48(1.24,1.77) | < 0.001 | 1.34(1.11,1.62) | 0.002 |
| P for trend | < 0.001 | < 0.001 | 0.005 | |||||
| Male Age < 65 | 3989 | 1084(27.2) | 1.62(1.11,2.35) | 0.012 | 1.84(1.25,2.71) | 0.002 | 1.55(1.02,2.35) | 0.04 |
| Quartile3 | ||||||||
| Q1 (< 0.5) | 798 | 214(26.8) | 1.10(0.88,1.38) | 0.384 | 1.07(0.85,1.34) | 0.566 | 1.03(0.81,1.31) | 0.807 |
| Q2 (0.5 ~ 0.6) | 795 | 198(24.9) | reference | reference | reference | |||
| Q3 (0.6 ~ 0.7) | 800 | 201(25.1) | 1.01(0.81,1.27) | 0.919 | 1.02(0.82,1.29) | 0.841 | 1.04(0.82,1.33) | 0.731 |
| Q4 (0.7 ~ 0.8) | 795 | 202(25.4) | 1.03(0.82,1.29) | 0.817 | 1.06(0.84,1.33) | 0.626 | 1.02(0.79,1.30) | 0.899 |
| Q5 (≥ 0.8) | 801 | 269(33.6) | 1.52(1.23,1.90) | < 0.001 | 1.58(1.27,1.98) | < 0.001 | 1.42(1.12,1.81) | 0.004 |
| P for trend | 0.001 | < 0.001 | 0.011 | |||||
| Femal Age < 65 | 4166 | 1543(37.0) | 1.62(1.14,2.30) | 0.008 | 2.03(1.41,2.93) | < 0.001 | 1.82(1.24,2.68) | 0.002 |
| Quartile4 | ||||||||
| Q1 (< 0.5) | 828 | 311(37.6) | 1.05(0.86,1.28) | 0.649 | 1.02(0.84,1.25) | 0.828 | 1.02(0.83,1.26) | 0.826 |
| Q2 (0.5 ~ 0.5) | 836 | 305(36.5) | reference | reference | reference | |||
| Q3 (0.5 ~ 0.6) | 831 | 268(32.3) | 0.83(0.68,1.01) | 0.069 | 0.85(0.69,1.04) | 0.121 | 0.88(0.71,1.09) | 0.227 |
| Q4 (0.6 ~ 0.7) | 835 | 295(35.3) | 0.95(0.78,1.16) | 0.623 | 1.01(0.82,1.23) | 0.958 | 0.97(0.79,1.20) | 0.801 |
| Q5 (≥ 0.7) | 836 | 364(43.5) | 1.34(1.10,1.63) | 0.003 | 1.46(1.20,1.79) | < 0.001 | 1.36(1.10,1.69) | 0.004 |
| P for trend | 0.024 | 0.001 | 0.018 | |||||
| Male Age ≥ 65 | 5573 | 609(10.9) | 1.45(0.85,2.48) | 0.17 | 1.39(0.81,2.38) | 0.226 | 1.22(0.70,2.15) | 0.481 |
| Quartile5 | ||||||||
| Q1 (< 0.4) | 1112 | 125(11.2) | 1.31(0.99,1.73) | 0.056 | 1.32(1.00,1.74) | 0.054 | 1.28(0.96,1.72) | 0.093 |
| Q2 (0.4 ~ 0.5) | 1113 | 98(8.8) | reference | reference | reference | |||
| Q3 (0.5 ~ 0.6) | 1112 | 113(10.2) | 1.17(0.88,1.56) | 0.275 | 1.17(0.88,1.56) | 0.274 | 1.21(0.90,1.62) | 0.217 |
| Q4 (0.6 ~ 0.7) | 1121 | 130(11.6) | 1.36(1.03,1.79) | 0.03 | 1.36(1.03,1.79) | 0.029 | 1.35(1.01,1.80) | 0.044 |
| Q5 (≥ 0.7) | 1115 | 143(12.8) | 1.52(1.16,2.00) | 0.002 | 1.50(1.14,1.97) | 0.003 | 1.37(1.03,1.82) | 0.032 |
| P for trend | 0.005 | 0.007 | 0.043 | |||||
| Female Age ≥ 65 | 4314 | 928(21.5) | 1.61(1.01,2.57) | 0.046 | 1.82(1.14,2.91) | 0.013 | 1.79(1.11,2.90) | 0.018 |
| Quartile6 | ||||||||
| Q1 (< 0.4) | 859 | 212(24.7) | 1.39(1.10,1.75) | 0.005 | 1.31(1.04,1.65) | 0.023 | 1.26(0.99,1.61) | 0.056 |
| Q2 (0.4 ~ 0.5) | 855 | 163(19.1) | reference | reference | reference | |||
| Q3 (0.5 ~ 0.6) | 874 | 154(17.6) | 0.91(0.71,1.16) | 0.438 | 0.90(0.70,1.15) | 0.395 | 0.87(0.67,1.11) | 0.26 |
| Q4 (0.6 ~ 0.7) | 862 | 171(19.8) | 1.05(0.83,1.33) | 0.686 | 1.06(0.83,1.35) | 0.628 | 1.07(0.84,1.37) | 0.597 |
| Q5 (≥ 0.7) | 864 | 228(26.4) | 1.52(1.21,1.91) | < 0.001 | 1.51(1.20,1.91) | < 0.001 | 1.43(1.13,1.81) | 0.003 |
| P for trend | 0.027 | 0.015 | 0.035 | |||||
Due to the small sample size, the data for men and women under 40 years of age did not reach statistical significance; therefore, no gender-based differentiation was made for individuals aged 19–39. The primary focus of the study is on middle-aged and elderly populations, particularly those aged 65 and above
Table 4.
Threshold effect analyses of plasma VA on the prevalence of hyperhomocysteinemia
| VA, µg/mL | Crude OR 95%IC | Crude P | Adjust OR 95%IC | Adjust P |
|---|---|---|---|---|
| A straight line effect: | 2.239 (1.838,2.729) | < 0.001 | 1.855 (1.485,2.318) | < 0.001 |
| Fold points: | 0.514 µg/mL | |||
| < K segment effect 1: | 0.092 (0.049,0.172) | < 0.001 | 0.137 (0.069,0.272) | < 0.001 |
| >= Segment K effect 2: | 5.495 (4.242,7.119) | < 0.001 | 3.902 (2.921,5.213) | < 0.001 |
| Effect difference of 2 and 1: | 59.859 (27.591,129.866) | < 0.001 | 28.558 (12.284,66.392) | < 0.001 |
| Log-likelihood ratio test: | < 0.001 | < 0.001 |
Supplementary-Table-1 presents the logistic regression analysis of plasma vitamin A and hyperhomocysteinemia stratified by gender, while Supplementary-Table-2 displays the analysis for the combined group. To examine the association between serum vitamin A levels and serum homocysteine levels across subgroups, the study population was categorized into five quartiles according to serum vitamin A concentrations. Quartile 2 (Q2) was used as the reference group, and quartiles Q3–Q5 were merged for comparison. After adjusting for covariates in Model 1 and Model 2, a U-shaped dose-response relationship between vitamin A and homocysteine levels was observed, as previously indicated. Figure 4 presents a forest plot illustrating the adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for elevated homocysteine across categories of plasma vitamin A, stratified by relevant covariates. Notably, hypertension was identified as a potential effect modifier.
Fig. 4.
Restricted spline curve for associations of plasma VA with the prevalence of hyperhomocysteinemia by age. P for nonlinear < 0.001.
(Variables adjusted for: age, sex, BMI, smoking, alcohol consumption, folate, vitamin B12, methyltransferase MTHFR C677T, hypertension, stroke, coronary artery disease.)
Figure 5 depicts the restricted cubic spline curve showing the prevalence of hyperhomocysteinemia in relation to plasma vitamin A levels across age groups, with a significant nonlinear trend (P for nonlinearity < 0.001). Figure 6 displays the restricted cubic spline curve for the association between plasma vitamin A and the prevalence of hyperhomocysteinemia, also demonstrating significant nonlinearity (P < 0.001). Binary logistic regression analysis revealed that, as vitamin A quartiles increased, the prevalence of hyperhomocysteinemia initially decreased and subsequently increased. The β coefficients for each vitamin A quartile were as follows: Q1: 1.17 (95% CI: 1.03–1.32), Q2: 1.86 (95% CI: 1.49–2.32), Q3: 1.03 (95% CI: 0.91–1.16), Q4: 1.12 (95% CI: 0.99–1.26), and Q5: 1.48 (95% CI: 1.31–1.67). Furthermore, restricted cubic spline analysis confirmed that the odds ratio (OR) of hyperhomocysteinemia exhibited an initial decline followed by an increase with rising serum vitamin A levels. This pattern remained consistent across different age and sex subgroups.
Fig. 5.
Restricted spline curve for associations of plasma VA with the prevalence of hyperhomocysteinemia by sex. P for nonlinear < 0.001(Variables adjusted for: age, sex, BMI, smoking, alcohol consumption, folate, vitamin B12, methyltransferase MTHFR C677T, hypertension, stroke, coronary artery disease)
Fig. 6.
Forest plots on the association of plasma VA levels with the risk of hyperhomocysteinemia (adjusted OR and 95% CI), stratified by pertinent covariables(the second quintile of the dataset as the reference group)
Table 4 presents the threshold effect analysis of plasma vitamin A (VA) on the prevalence of hyperhomocysteinemia. The association between the threshold effect of vitamin A and elevated homocysteine levels was examined, with the breakpoint identified at 0.514 µg/mL. This finding further supports a U-shaped relationship between vitamin A status and homocysteine concentrations, with statistical significance confirmed through differential effect analysis. Figure 7 displays a forest plot illustrating the associations between potential effect modifiers of VA and homocysteine (HCY) across subgroups stratified by the VA threshold of 0.514 µg/mL. Using this concentration as the cutoff, the correlation between vitamin A and homocysteine was categorized, and potential effect modifiers were incorporated into the construction of the forest plot. The results indicate that when plasma vitamin A levels are below 0.514 µg/mL, body mass index (BMI) acts as an interaction factor. In contrast, when vitamin A levels are at or above 0.514 µg/mL, age and smoking status emerge as significant interaction factors. Notably, a statistically significant interaction between vitamin A and homocysteine levels was observed.
Fig. 7.
Forest plots on the association of VA and HCY by potential effect modifiers in various subgroups divided by VA concentration of 0.514 ug/mL
Discussion
Based on a well-designed large-sample epidemiological cohort study, this research provides robust evidence for a significant association between vitamin A and homocysteine through both cross-sectional and longitudinal analyses. The results indicate a U-shaped dose-response relationship between vitamin A and homocysteine (nonlinear P < 0.001) in the middle-aged and elderly Chinese population. This relationship persists across different age groups, genders, and high homocysteine levels. The incidence of hyperhomocysteinemia was lowest when serum vitamin A concentration was 0.514 µg/mL, providing substantial evidence for the relationship between serum vitamin A and homocysteine in the Chinese population. In participants with vitamin A levels at or above 0.514 µg/mL, an interaction between age and hypertensive stratification variables was observed. For individuals with vitamin A levels below 0.514 µg/mL, the prevalence of hyperhomocysteinemia significantly decreased as vitamin A levels increased. Conversely, for those with vitamin A levels at or above 0.514 µg/mL, the prevalence of hyperhomocysteinemia significantly increased. Serum vitamin A levels were assessed by quartiles; compared to the second quartile, the first quartile (Q1: OR, 1.17; 95% CI, 1.04–1.32) and the third to fifth quartiles (Q3-5: OR, 1.19; 95% CI, 1.08–1.32) showed higher prevalence of hyperhomocysteinemia.
Individuals with elevated homocysteine levels tend to be older, and the prevalence of hyperhomocysteinemia increases with age, being significantly higher in males compared to females. This observation aligns with previous studies, including a cross-sectional study on the prevalence of hyperhomocysteinemia in Hunan, China [25], and the first meta-analysis of hyperhomocysteinemia prevalence in China [26]. Based on our research hypothesis, we conducted a large-scale cross-sectional study to investigate the U-shaped dose-response relationship between vitamin A and homocysteine levels. Additionally, we explored the nonlinear relationship between vitamin A and homocysteine within subgroups, treating both as continuous variables. These findings provide valuable insights into the pathophysiological mechanisms underlying the association between vitamin A and hyperhomocysteinemia.
Homocysteine, a sulfur-containing amino acid generated during the conversion of methionine to cysteine, is influenced by metabolic processes regulated by various factors [27]. After adjusting for potential confounding variables, our analysis confirmed an independent association between vitamin A and homocysteine levels in middle-aged and elderly rural Chinese populations.
Vitamin A content is quantified using retinol equivalents (RE), where 1 RE corresponds to 0.001 mg of retinol, 0.006 mg of beta-carotene, or 3.3 international units (IU) of vitamin A [28]. Retinol can be sourced from all forms of vitamin A, while retinoids are naturally present in certain animal-based foods commonly consumed in daily diets [12]. Vitamin A deficiency can disrupt the activities of osteoclasts and osteoblasts, resulting in abnormal growth of the occipital bone and spine, accompanied by severe neurological complications, and also increases the risk of fractures [29–31]. Therefore, it is crucial to ensure appropriate intake levels of vitamin A from fortified foods or supplements, avoiding both insufficiency and excess. Our study indicates that vitamin A is a significant factor influencing hyperhomocysteinemia, providing valuable insights for future cohort studies to investigate the epidemiology and nutritional implications of vitamin A in relation to hyperhomocysteinemia.
Our findings elucidate the U-shaped relationship between vitamin A and homocysteine levels, offering valuable insights for predicting cardiovascular disease risk. By modulating micronutrient intake, it is possible to maintain optimal homocysteine levels. For instance, monitoring vitamin A and homocysteine levels in pregnant women can facilitate early screening and prevention of birth defects. Leveraging these insights, public health policies and nutritional guidelines can be formulated to mitigate disease risk within the population, thereby fostering interdisciplinary collaboration among nutritionists, medical professionals, epidemiologists, and biologists.
The strengths of this study include a large sample size and extensive geographic coverage, encompassing the majority of provinces in China. This is one of the few large-scale studies focusing on middle-aged and elderly individuals in rural areas of China. We have implemented a robust quality control system in our clinical laboratory to ensure the comparability and stability of various measurements. Our analysis comprehensively examines the relationship between vitamin A and homocysteine, including dose-response and threshold effects. Both the logistic regression model and the restricted cubic spline (RCS) model account for multiple covariates to yield accurate and reliable results. However, as a cross-sectional study, it has limitations in establishing causality. Further randomized controlled trials and real-world studies are necessary to validate the reliability of our findings.
Conclusion
A dose-response relationship was observed between serum vitamin A levels and homocysteine concentrations among middle-aged and elderly individuals in rural China, with variations noted across different subgroups. Maintaining an adequate level of vitamin A in patients with hyperhomocysteinemia may assist in reducing homocysteine levels. This study provides valuable scientific evidence for the correlation between vitamin A and homocysteine.
Supplementary Information
Acknowledgements
We thank all the participants in the study and members of the survey teams in all the study centers. Members of the CPNAS collaborative group: (a) China Nutrition and Health Food Association: Zhen jia Bian, Liang qiu Li, Ning ling Sun, Xiao shu Cheng, Han ping Shi, Jian ping Li, Wen hua Ling, Jin gang Yang, Gui fan Sun, Binyan Wang, Hui hui Bao, Chen Mao, Xianhui Qin. (b) Steering committee: Gangqiang Ding, Jun sheng Huo, Yong Huo, Da fang Chen, Yan Zhang, Ping Li, Guangyun Mao, Zeng ning Li, Xiao liang Shu, Xiang Gao, Ming Liu, Pinning Feng, Xin zheng Lu, Yong Duan, Yu Fu, Jian long Wu, Jia man Ou, Xuli Wu, Xiao Huang, Ziyi Zhou, Shu fang Xu, Mingli He, Hai Ma, Qing Dong. (c) Project coordination group: Hou xun Xing, Gen fu Tang, Zhi ping Li, Yun Song, Li shun Liu, Qiang qiang He, Ping Chen, Jia feng Xu, Chang rui Ou, He hao Zhu, Jia ping Huan, Dongmei Cheng, Yu Zhou Ding, Peng Xu, Ying Wang , Jia xiang Ding, Yuan yuan Xu, Jin Mei Zhou, Jun jie Bao, Yu Peng, Fei Guo, Yun qiu Xie, Xian hui Qin, Huan Zhou.
Author statement
This manuscript has not been published previously.
Authors’ contributions
We thank all the participants in the study and members of the survey teams in all the study centers. Members of the CPNAS collaborative group: (a) China Nutrition and Health Food Association: Zhen jia Bian, Liang qiu Li, Ning ling Sun, Xiao shu Cheng, Han ping Shi, Jian ping Li, Wen hua Ling, Jin gang Yang, Gui fan Sun, Binyan Wang, Hui hui Bao, Chen Mao, Xianhui Qin. (b) Steering committee: Gangqiang Ding, Jun sheng Huo, Yong Huo, Da fang Chen, Yan Zhang, Ping Li, Guangyun Mao, Zeng ning Li, Xiao liang Shu, Xiang Gao, Ming Liu, Pinning Feng, Xin zheng Lu, Yong Duan, Yu Fu, Jian long Wu, Jia man Ou, Xuli Wu, Xiao Huang, Ziyi Zhou, Shu fang Xu, Mingli He, Hai Ma, Qing Dong. (c) Project coordination group: Hou xun Xing, Gen fu Tang, Zhi ping Li, Yun Song, Li shun Liu, Qiang qiang He, Ping Chen, Jia feng Xu, Chang rui Ou, He hao Zhu, Jia ping Huan, Dongmei Cheng, Yu Zhou Ding, Peng Xu, Ying Wang , Jia xiang Ding, Yuan yuan Xu, Jin Mei Zhou, Jun jie Bao, Yu Peng, Fei Guo, Yun qiu Xie, Xian hui Qin, Huan Zhou.
Funding
1. Longhu Talent Project of Bengbu Medical University [grant no.LH250101002].
2. The study was supported by funding from the Development and Reform Commission of Shenzhen Municipality [grant no.XMHT20220104055],
[grant no.XMHT20240104002]
Data availability
No datasets were generated or analysed during the current study.
Declarations
Ethics approval and consent to participate
The protocol and procedures of this study have been reviewed and approved by the Ethics Committees of Beijing Shijitan Hospital, Capital Medical University. CPNAS is registered on https://www.chictr.org.cn at ChiCTR2100051983. The collection of biosamples was approved by the China Human Genetic Resources Management Office, with the approval number [2021] CJ2958. Informed consent was obtained from all individual participants included in the study.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Ning Chen, Dongmei Cheng and Yu Zhou Ding contributed equally to this work.
Contributor Information
Xian hui Qin, Email: pharmaqin@126.com.
Han ping Shi, Email: shihp@ccmu.edu.cn.
Huan Zhou, Email: zhouhuan@bbmc.edu.cn.
References
- 1.Welch GN, Loscalzo J. Homocysteine and atherothrombosis. N Engl J Med Apr. 1998;9(15):1042–50. 10.1056/nejm199804093381507. [DOI] [PubMed] [Google Scholar]
- 2.Wu DF, Yin RX, Deng JL. Homocysteine, hyperhomocysteinemia, and H-type hypertension. Eur J Prev Cardiol Jul. 2024;23(9):1092–103. 10.1093/eurjpc/zwae022. [DOI] [PubMed] [Google Scholar]
- 3.Boushey CJ, Beresford SA, Omenn GS, Motulsky AG. A quantitative assessment of plasma homocysteine as a risk factor for vascular disease. Probable benefits of increasing folic acid intakes. Jama Oct. 1995;4(13):1049–57. 10.1001/jama.1995.03530130055028. [DOI] [PubMed] [Google Scholar]
- 4.Toole JF, Malinow MR, Chambless LE, et al. Lowering homocysteine in patients with ischemic stroke to prevent recurrent stroke, myocardial infarction, and death: the vitamin intervention for stroke prevention (VISP) randomized controlled trial. Jama Feb. 2004;4(5):565–75. 10.1001/jama.291.5.565. [DOI] [PubMed] [Google Scholar]
- 5.Smith AD, Refsum H. Homocysteine - from disease biomarker to disease prevention. J Intern Med Oct. 2021;290(4):826–54. 10.1111/joim.13279. [DOI] [PubMed] [Google Scholar]
- 6.Ganguly P, Alam SF. Role of homocysteine in the development of cardiovascular disease. Nutr J Jan. 2015;10:14:6. 10.1186/1475-2891-14-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Huang X, Li Y, Li P, et al. Association between percent decline in serum total homocysteine and risk of first stroke. Neurology Nov. 2017;14(20):2101–7. 10.1212/wnl.0000000000004648. [DOI] [PubMed] [Google Scholar]
- 8.Center For Cardiovascular Diseases The Writing Committee Of The Report On, Cardiovascular H. Diseases in China N. Report on cardiovascular health and diseases in China 2023: an updated summary. Biomed Environ Sci Sep. 2024;20(9):949–92. 10.3967/bes2024.162. [DOI] [PubMed] [Google Scholar]
- 9.Luo Y, Xia F, Yu X, Li P, Huang W, Zhang W. Long-term trends and regional variations of hypertension incidence in china: a prospective cohort study from the China health and nutrition Survey, 1991–2015. BMJ Open Jan. 2021;13(1):e042053. 10.1136/bmjopen-2020-042053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wang Z, Chen Z, Zhang L, et al. Status of hypertension in china: results from the China hypertension Survey, 2012–2015. Circulation May. 2018;29(22):2344–56. 10.1161/circulationaha.117.032380. [DOI] [PubMed] [Google Scholar]
- 11.Li J, Mao C, Xu B, et al. China precision nutrition and Health-KAP (Knowledge, Attitude, and Practice) Real-World study (CPNAS): enrollment progress and baseline characteristics. Precision Nutr. 2025;4(1):e00095. 10.1097/pn9.0000000000000095. [Google Scholar]
- 12.Bates CJ, Vitamin A. Lancet. 1995;345(8941):31 – 5.10.1016/s0140-6736(95)91157-x [DOI] [PubMed] [Google Scholar]
- 13.Song P, Wang J, Wei W, Chang X, Wang M, An L. The prevalence of vitamin A deficiency in Chinese children: A systematic review and Bayesian meta-analysis. Nutrients. 2017;9(12):1285. 10.3390/nu9121285. [DOI] [PMC free article] [PubMed]
- 14.Mawson AR, Onor GI. Gout and vitamin A intoxication: is there a connection? Semin Arthritis Rheum Apr. 1991;20(5):297–304. 10.1016/0049-0172(91)90030-4. [DOI] [PubMed] [Google Scholar]
- 15.Moll S, Varga EA. Homocysteine and MTHFR mutations. Circulation. 2015;132(1):e6–e9. 10.1161/CIRCULATIONAHA.114.013311. [DOI] [PubMed]
- 16.Jakubowski H. Homocysteine modification in protein Structure/Function and human disease. Physiol Rev Jan. 2019;1(1):555–604. 10.1152/physrev.00003.2018. [DOI] [PubMed] [Google Scholar]
- 17.Xu BP, Shi H. Precision nutrition: concept, evolution, and future vision. Precis Nutr. 2022;1(1):e00002. 10.1097/PN9.0000000000000002.
- 18.Yun S, Ping C, ZP A, Jianping L, Hanping S. Precision nutrition: 8 stages and 5 dimensions. Precision Nutr. 2023;2(4):e00057. 10.1097/pn9.0000000000000057. [Google Scholar]
- 19.Tianyu C, Xiao H, Ping C, et al. Distribution and status of vitamin B12 in Chinese adults with hypertension: a comprehensive report across 14 provinces. Precision Nutr. 2023;2(4):e00060. 10.1097/pn9.0000000000000060. [Google Scholar]
- 20.Xie H, Zhang K, Wei Y, Ruan G, Zhang H, Li S, Song Y, Chen P, Liu L, Wang B, Shi H. The association of serum betaine concentrations with the risk of new-onset cancers: results from two independent nested case-control studies. Nutr Metab (Lond). 2023;20(1):46. 10.1186/s12986-023-00755-y. [DOI] [PMC free article] [PubMed]
- 21.Huo Y, Li J, Qin X, et al. Efficacy of folic acid therapy in primary prevention of stroke among adults with hypertension in china: the CSPPT randomized clinical trial. Jama Apr. 2015;7(13):1325–35. 10.1001/jama.2015.2274. [DOI] [PubMed] [Google Scholar]
- 22.Wei Y, He Q, Zhu H, et al. A negative association between plasma phylloquinone and All-Cause mortality in Chinese adults with hypertension: A nested Case-Control study. J Nutr Mar. 2024;154(3):978–84. 10.1016/j.tjnut.2023.12.009. [DOI] [PubMed] [Google Scholar]
- 23.Hanping S, Jianping L, Chen M, et al. China precision nutrition and Health-KAP real world study (CPN-KAPS): rationale, study design, and protocol. Precision Nutr. 2022;1(3):e00021. 10.1097/pn9.0000000000000021. [Google Scholar]
- 24.Chen P, Lin G, Lin L, Song Y, Huan J, Tian M. Simultaneous determination of vitamins A, D, E, K in human plasma. Precision Nutr. 2024;3(4):e00086. 10.1097/pn9.0000000000000086. [Google Scholar]
- 25.Yang B, Fan S, Zhi X, et al. Prevalence of hyperhomocysteinemia in china: a systematic review and meta-analysis. Nutrients Dec. 2014;29(1):74–90. 10.3390/nu7010074. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Yang Y, Zeng Y, Yuan S, et al. Prevalence and risk factors for hyperhomocysteinemia: a population-based cross-sectional study from Hunan, China. BMJ Open Dec. 2021;6(12):e048575. 10.1136/bmjopen-2020-048575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Al Hageh C, Alefishat E, Ghassibe-Sabbagh M, et al. Homocysteine levels, H-Hypertension, and the MTHFR C677T genotypes: A complex interaction. Heliyon Jun. 2023;9(6):e16444. 10.1016/j.heliyon.2023.e16444. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Mawson AR. Hypervitaminosis A toxicity and gout. Lancet May. 1984;26(8387):1181. 10.1016/s0140-6736(84)91424-7. [DOI] [PubMed] [Google Scholar]
- 29.Feskanich D, Singh V, Willett WC, Colditz GA. Vitamin A intake and hip fractures among postmenopausal women. Jama Jan 2. 2002;287(1):47–54. 10.1001/jama.287.1.47. [DOI] [PubMed] [Google Scholar]
- 30.Lim LS, Harnack LJ, Lazovich D, Folsom AR. Vitamin A intake and the risk of hip fracture in postmenopausal women: the Iowa women’s health study. Osteoporos Int Jul. 2004;15(7):552–9. 10.1007/s00198-003-1577-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Chandra RK. Serum retinol levels and fracture risk. N Engl J Med May. 2003;8(19):1927–8. 10.1056/NEJM200305083481917. [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
No datasets were generated or analysed during the current study.