Background

Aging is a complex and multifaceted process involving degenerative changes in physiological functions and the loss of regeneration potential in multiple tissues and organs1,2. Organismal aging is an inevitable process and represents a significant risk factor for various diseases, such as cardiovascular3,4, hepatic5, nephritic6, and nervous system diseases7. However, the aging process demonstrates heterogeneity, as individuals with the same chronological age exhibit greatly varied susceptibility to age-related disorders and disparate rates of aging due to differences in genetics, lifestyles, and environmental exposures8,9. As a comprehensive indicator for assessing the degree of biological aging in individuals, phenotypic age can distinguish mortality risk among individuals of identical chronological age by utilizing clinical chemical biomarker data from multiple systems. It not only identifies individuals at risk of multiple diseases and mortality but also stratifies populations into varying risk levels, encompassing the healthiest and unhealthiest10. Effectively identifying influencing factors associated with aging and ascertaining potential interventions capable of slowing down the aging process could help mitigate disease incidence and extend lifespan.

Emerging research has identified measurable genetic effects on the aging rate, with genetic influences accounting for only 1–27% of the variation in population longevity11,12, while lifestyle factors like income, nutrition, education and especially hygiene, medical therapy and health care play a more significant role in determining the pace of aging13. Notably, interventions such as experimental data from preclinical models, the administration of metformin, exercise and calorie restriction have been shown to decrease the risk of aging14. Among these interventions, diet has been identified as a key factor in regulating the aging process15. Specifically, foods abundant in dietary selenium16, dietary fiber17, dietary copper18, flavonoids19, dietary antioxidant-rich components20, high-quality carbohydrates, and plant proteins21 could decelerate the aging process. Currently, research on the relationship between dietary nutrient intake and aging remains limited, with existing studies focusing on the intake of a small number of individual dietary nutrients.

The consumption of methyl donor nutrients (MDNs) such as protein, folate, choline, betaine, vitamin B2 (VB2), vitamin B6 (VB6), vitamin B12 (VB12), and zinc plays a crucial role in one-carbon metabolism (OCM), DNA methylation, regulation of gene expression, and cellular metabolism22. Changes in DNA and histone methylation are hallmarks of aging23. Meanwhile, impairment in the OCM pathway could lead to a loss of epigenetic marks and potentially increase the rate of aging24. However, to date, no study has conducted a comprehensive evaluation of the association between dietary intake of MDNs and phenotypic age. Therefore, this study aimed to investigate the associations between MDNs and phenotypic age, utilizing data from the National Health and Nutrition Examination Survey (NHANES) from 2005 to 2018. Considering some baseline characteristics, such as age, gender, race, lifestyle and disease, may influence the rate of aging in humans to some extent8,9,13. This study also explored the interaction effects between these baseline characteristics and MDNs on phenotypic aging.

Methods

Study population

The data for this cross-sectional study was extracted from the NHANES spanning seven cycles from 2005 to 2018. NHANES is a project overseen by the National Center for Health Statistics (NCHS) and is designed to be a comprehensive cross-sectional survey representing the nation as a whole. Data from NHANES were gathered through interviews, as well as laboratory and physical examinations, with all participants providing written informed consent. Initially, 70,190 participants were considered, but 42,979 subjects with missing information were excluded for the calculation of biological aging (n = 36,892). Additionally, participants who lacked data necessary for methyl-donor nutritional quality index (MNQI) calculations, as well as individuals with total energy intake levels exceeding 500–5000 kcal/day, were also removed from the analysis (n = 6,087)25, leaving a final sample of 27,211 adults. Figure 1 illustrates the specific procedure. A comparative analysis of the main characteristics of the included and excluded subjects revealed the potential presence of selection bias, which may exert a certain influence on the research results and consequently limit the applicability of the study conclusions (Table S1).

Fig. 1
figure 1

Study population selection (N = 27,211) by Figdraw.

Full size image

Assessment of MDNs dietary intake

The food intake data from NHANES were obtained through interviews and a 24-h dietary review survey. The nutrients/food components from each food/beverage were calculated using U.S. Department of Agriculture’s Food and Nutrient Database. Based on previous research, we calculated the MNQI by assessing the intake of MDNs: protein, folate, choline, riboflavin, vitamin B6, vitamin B12, and zinc26. The optimal intake of each nutrient was defined as falling between two-thirds of the recommended intake and the upper limit value. Intakes within this range were scored as 1 (for riboflavin, vitamin B6, vitamin B12, and zinc) or 2 (for protein, folate, and choline), while intakes below or above this range were scored as 0. The methodology used was adapted from a previous study that demonstrated a significant correlation between MNQI and OCM, validating MNQI as a reliable tool for assessing the overall nutritional quality of dietary MDNs26,27. In accordance with this research, a score of 6 or higher indicates high methyl-donor nutritional quality index (HMNQI), whereas a score below 6 signifies low methyl-donor nutritional quality index (LMNQI)26,28.

Biological aging measure

PhenoAge, a novel measure of biological aging, was calculated using an algorithm derived from multivariate analyses of mortality hazards to estimate the risk of death. This study calculated individual phenoAge based on previous representative research on phenotypic age29,30,31. Briefly, PhenoAge is calculated using chronological age and 9 biomarkers (albumin, creatinine, glucose, C-reactive protein, lymphocyte percent, mean cell volume, red blood cell distribution width, alkaline phosphatase, and white blood cell count) that were selected using a Cox proportional hazard elastic net model for mortality based on ten-fold cross-validation. The algorithm for calculating PhenoAge is based on parametrization of 2 Gompertz proportional hazard models-one fit using all 10 selected variables, and the other fit using only chronological age. These biological markers were selected after a ten-fold cross-validation of mortality using a Cox proportional hazards elastic net model. PhenoAge was calculated using the BioAge R package. PhenoAge Acceleration (PhenoAge.Accel) was calculated as the residual from a linear regression model of PhenoAge against chronological age to quantify phenotypic aging30,31. Specifically, individuals were considered phenotypically older (or younger) if their PhenoAge.Accel values were greater (or less) than 0, relative to their chronological age32.

Covariates

We developed a theoretical model to estimate both the total and direct associations between MNQI and PhenoAge.Accel, based on a directed acyclic graph (DAG) (Fig. S1)33,34. The direct association refers to the relationship between the exposure (MNQI) and the outcome (PhenoAge.Accel), excluding mediation effects from other variables, including age (< 45 or ≥ 45 years), gender (male or female), race (non-Hispanic White or others), educational level (< high school, high school, or > high school), marital status (never married, married/living as married, or separated/divorced/widowed), annual household income (< US$20,000 or ≥ US$20,000), and smoking status (non-smokers or smokers). The total association, on the other hand, includes mediation effects from obesity (non-obese: BMI < 30 or obese: BMI ≥ 30) and health conditions (diabetes mellitus, hypertension, cardiovascular disease [CVD], and cancer). Additionally, total energy intake was included as a covariate in this study.

Statistical analysis

The weights in our research were calculated based on the recommendations found in the NHANES analysis guide. The new 14-year weights were calculated using 1/7 of the 2-day dietary weights. Continuous variables were described as means and standard errors (SEs), while categorical variables were presented as percentages. Weighted linear regression models were used to examine the relationship between MNQI and PhenoAge.Accel. Model I (direct association) was adjusted for age, gender, race, marital status, educational level, annual household income, smoking status. Model II (total association) was further adjusted for obesity, diabetes mellitus, hypertension, CVD, and total energy intake. In addition, stratified and interaction analyses were performed to assess whether the association between MNQI and PhenoAgeAccel varied across the variables presented in Table 1. A complete case analysis approach was employed in this study due to the minimal amount of missing. All analyses were conducted using R software (version 4.1.2), with statistical significance considered when the p-value was below 0.05.

Table 1 Weighted characteristics of the participants (N = 27,211).
Full size table

Results

Baseline characteristics

The basic characteristics of the study participants are shown in Table 1. The mean age of the 27,211 adults was 47.60 ± 0.27 years. The age range was from 20 to 85 years, and 52.37% of the participants were female. Significant differences were observed in total energy intake and PhenoAge.Accel, with the HMNQI group having higher total energy intake and being phenotypically younger relative to their chronological age. Participants with higher scores of methyl donor intake were more likely to be female, Non-Hispanic White, aged ≥ 45 years, in higher-income groups, married or living as married, non-smokers, diabetes mellitus, hypertension, or CVD, and individuals with higher education (Table 1). All differences were statistically significant (P < 0.05).

Association between MNQI and PhenoAge.Accel

The association between MNQI and PhenoAge.Accel was examined using weighted linear regression in Table 2. High MDNs intake was negatively associated with PhenoAge.Accel in the crude model and adjusted models: Crude model (β = -0.70; 95% CI − 1.01, -0.39; P<0.0001), Model I (β = − 0.71; 95% CI − 0.99, − 0.42; P<0.0001) and Model II (β = − 0.66; 95% CI − 0.91, − 0.40; P < 0.0001).

Table 2 Association between MNQI and phenoage acceleration (N = 27,211).
Full size table

Subgroup analyses

Figure 2 illustrates the subgroup analyses examining the relationship between MNQI and PhenoAge.Accel. The analyses consistently demonstrated a negative association between MNQI and PhenoAge.Accel across all subgroups. Notably, our findings suggest that gender may play a role in moderating this relationship, as evidenced by a significant interaction P value of 0.02. Specifically, the effect of MDNs intake on PhenoAge.Accel was more pronounced in males, with a β coefficient of − 0.94 and a 95% confidence interval of − 1.36 to − 0.52. However, due to the consistent directionality of all the associations, these results may lack significant clinical implications.

Fig. 2
figure 2

Subgroup analyses of the associations between MNQI and PhenoAge acceleration by age, gender, race, income, education level, marital status, smoking status, obesity, diabetes mellitus, hypertension, CVD, and cancer (N = 27,211).

Full size image

Discussion

In this study, we discovered a significant association between dietary MDNs intake and a deceleration in phenotypic aging, particularly in males. These relationships were observed in a sample representative of the U.S. population and remained significant even after controlling for multiple traditional risk factors such as smoking, hypertension, and diabetes.

The aging trajectory is malleable and can be decelerated by lifestyle factors such as good nutrition, sufficient physical activity, and avoidance of smoking35,36. Nutritional and dietary interventions play a crucial role in determining an individual’s health status and lifespan37,38,39. Dietary antioxidants, such as vitamins A, anthocyanins, lycopene, β-carotene, and polyphenols, have the ability to neutralize free radicals, thereby reducing oxidative stress and inflammation, protecting telomeres, and influencing aging-related signaling pathways. They also play a significant role in cancer prevention, cardiovascular system protection, and slowing down the aging process36,37,39. On the contrary, the chronic ingestion of one or more nutrients and energy far in excess of the body’s consumption has been demonstrated to hasten the aging process, leading to tissue damage and medullary inflammation, both of which contribute to accelerated organ aging40,41. In recent years, MDNs have gained significant attention due to their essential roles in providing methyl groups, participating in various physiological metabolisms, and influencing methylation homeostasis. Additionally, deficiencies in these MDNs have been associated with a range of health issues, including hyperhomocysteinemia, lipid metabolism disorders, histamine intolerance, and cardiovascular diseases38,42,43,44.

MDNs play a crucial role in maintaining DNA methylation levels, regulating gene expression, and cellular metabolism22. The global DNA methylation level of mammalian genomes decreases with age, and alterations to DNA methylation can impact the regulation of chromosome structure and gene expression, potentially influencing the phenotypic aging process significantly38. Consumption of MDNs has been shown to slow down the aging process by preserving DNA methylation levels and shielding cells from damage45,46. Animal studies indicate that MDNs modulate metabolism, immune response, and epigenetic events through interactions within the one-carbon metabolism pathway35. Furthermore, MDNs play a role in maintaining cardiovascular health and delaying aging by participating in the metabolism of homocysteine, converting it into harmless metabolites, and thereby reducing its levels in the blood38,42. Additionally, the antioxidant properties of MDNs may contribute to slowing down the aging process by scavenging free radicals in the body and reducing cellular damage caused by oxidative stress47. These processes are essential for maintaining cellular function and delaying the aging process. Thus, it can be inferred that increased consumption of MDNs may lead to a delay in phenotypic aging by promoting the normalization of these processes48,49.

Better dietary quality was associated with a reduced risk of accelerated aging50, some population-based studies examining the associations between individual nutrients and aging have likewise corroborated this correlation. One study has further corroborated the correlation between a higher intake of folate, particularly natural folate, and a lower intake of folic acid supplements, and lower biological age indicators51. Zinc, another essential nutrient, plays a crucial role in DNA repair, immune function, and antioxidant defense. Studies have demonstrated that zinc supplementation can enhance immune function in older adults, reduce the risk of infections, and decelerate phenotypic aging52. The subgroup analyses that explored the potential modifying impacts of baseline characteristics significantly highlighted decelerated phenotypic aging among male participants when analyzing the association between MNQI and phenotypic aging. This finding aligns with those reported by Yue Chen et al. This phenomenon can be attributed to the observation that plant-based diets have been shown to increase frailty in elderly males, meanwhile the males are more susceptible to the influence of unhealthy lifestyles53,54. Even though aging may be affected by diabetes, hypertension and CVD18,55,56, the fundamental connection between MNQI and phenotypic aging remained unchanged even when considering these chronic conditions.

The present study’s findings are important in clinical research and have notable implications for public health. Increasing evidence indicates that DNA methylation plays a critical role in the regulation of gene transcription and function in various biological processes. Aberrant changes in DNA methylation are closely associated with aging and the development of diseases57,58,59. Dietary MDNs and acceptors interact through complex biochemical pathways to maintain methylation equilibrium and regulate the functioning of normal organs45,60. It is crucial to consider the overall characteristics of dietary MDNs to accurately assess the potential of a diet to influence DNA methylation. This study utilized the MNQI to investigate the association between overall dietary MDNs intake and phenotypic aging in adults, offering new ideas and methods for the prevention and treatment of age-related diseases. The findings serve as a basis for developing evidence-based dietary guidelines and nutritional intervention strategies. Additionally, this study contributes to a better understanding of the pathogenesis of aging, providing insights for the development of novel anti-aging pharmaceuticals and therapeutic approaches.

This study is limited by several factors. Firstly, our study utilized a cross-sectional design and therefore could not establish a causal relationship between MDNs and phenotypic aging. Future prospective or causal extrapolation studies are necessary for validation. Secondly, the food intake data were obtained through interviews and a 24-h dietary review survey. However, self-reported data in the study may be inaccurate due to factors such as respondent subjectivity, memory errors, recall bias, and others. Additionally, conducting only 24-h diet recall interviews may not provide a comprehensive reflection of long-term dietary intake. Thirdly, the selection of study participants in this study introduced some selection bias due to the absence of data on biological aging and MNQI, which may compromise the accuracy and reliability of the conclusions, consequently limiting the generalizability of the study findings. In future studies, it is recommended that more rigorous inclusion and exclusion criteria be adopted to enhance sample representativeness and reduce selection bias. Finally, we were unable to completely eliminate all potential residual confounders arising from unmeasured factors.

Conclusions

In conclusion, our research suggests a link between MDNs consumption and phenotypic aging, indicating that moderate MDNs intake may help prevent aging and inform nutritional strategies for healthy longevity. However, further prospective studies are needed to clarify our findings.