Folate and vitamin B12 are essential for a variety of metabolic processes. Both micronutrients have been shown to reduce oxidative stress significantly. The present cross-sectional investigation evaluated the association between serum and dietary folate and vitamin B12 levels and leukocyte telomere length, an index of cellular aging influenced by oxidative stress. The study included 5581 adults from the National Health and Nutrition Examination Survey (NHANES). Because participants were randomly selected, results are generalizable to all civilian, noninstitutionalized U.S. adults. A blood draw provided DNA and serum folate and B12 information. The quantitative polymerase chain reaction method was used to measure telomere length. The Bio-Rad Quantaphase II folate and vitamin B12 radioassay kit was used to quantify levels of folate and vitamin B12. Dietary folate and vitamin B12 were assessed using a multipass 24 h recall. In some models, age, race, smoking pack-years, alcohol use, body mass index, total physical activity, hours fasted before the blood draw, and diabetes status were employed as covariates to minimize their influence. Findings showed that for each additional year of chronological age, telomeres were 15.6 base pairs shorter, on average (
A good measure of DNA stability, cellular integrity, and biologic aging is the length of telomeres. Telomeres are nucleoprotein structures that cap the ends of chromosomes. Because the ends of telomeres cannot be completely replicated when cells divide, telomeric DNA is used to safeguard genetic code, so it is not lost from the ends of chromosomes. Telomeres naturally shorten during each cell division. The loss of telomere base pairs is sometimes called “the end replication problem,” but other factors can cause telomere shortening. Eventually, telomeres become critically short and senescence and cell death occur. Consequently, morbidity and mortality associated with several disorders are higher in adults with shorter telomeres [
Oxidative stress plays a major role in the length of telomeres and in cell aging [
A number of dietary factors can influence oxidative stress levels in the body. For example, research indicates that whole foods, such as fruits and vegetables, are significantly related to lower levels of oxidative stress [
Several investigations show that dietary factors that reduce inflammation and oxidative stress also tend to influence cellular aging favorably. There are a number of examples. Studies by Lian et al. [
On the other hand, a 2018 meta-analysis by Pérez et al. that included five randomized controlled trials covering nine different diets investigated the effects of diet on biologic aging. Collectively, the meta-analysis showed no effect of diet on telomere length [
Folate and vitamin B12 are micronutrients required for essential metabolic functions [
Because folate and vitamin B12 are major players in the control of oxidative stress [
Hence, the purpose of the present investigation was to evaluate the relationship between folate and vitamin B12 levels, derived from the blood and also from the diet, on telomere length in 5581 women and men, representative of the U.S. adult population. Another objective was to determine the effect of age, sex, race, hours of fasting, smoking, alcohol use, body mass index, total physical activity, diabetes status, and pregnancy status, on the vitamin and cellular aging relationships.
NHANES, also known as the National Health and Nutrition Examination Survey, is an ongoing assessment of the health and lifestyles of women, men, and children living in the United States. All NHANES data are cross-sectional. NHANES randomly samples individuals using a multilevel design so that external validity is strong and results can be generalized to all civilian, noninstitutionalized people in the U.S.
Although NHANES has been collecting national data for many decades, only two 2-year survey cycles include information on telomere length, 1999–2000 and 2001–2002. The public did not have access to telomere data until November 2014. All NHANES data sets are free and available online to the general public [
Collection of blood samples by NHANES allowing access to the DNA of individuals was delimited to participants who were ≥20 years old. A sample of 10,291 men and women was eligible and 76% (7827) agreed to give a sample and actually provided a useable DNA sample. To ensure absolute confidentiality, all individuals who were age ≥85 were given the age of 85 by NHANES. Therefore, men and women age 85 or older were omitted from the sample used in this study.
Women and men were required to have complete data to be included in the investigation. A total of 5581 adults (2648 men and 2933 women) had data for all the variables used in this study. The National Center for Health Statistics Ethics Review Board sanctioned collection of the NHANES data and written informed consent was acquired from each subject [
Leukocyte telomere length was the outcome measure of the present investigation. Serum and dietary folate and vitamin B12 concentrations were the exposure variables. Covariates included age, race, hours of fasting before the blood draw, body mass index (BMI), pack-years of smoking, total MET-minutes of physical activity, alcohol use, diabetes status, and pregnancy status (for women). Sex was controlled by stratifying the sample, focusing on women and men separately.
The techniques utilized to calculate telomere length have been reported by NHANES in detail [
According to NHANES [
NHANES mobile examination centers (MECs) were used for the collection of blood specimens. Serum concentrations of folate and vitamin B12 were evaluated using the Bio-Rad Laboratories Quantaphase II folate/vitamin B12 radioassay kit [
Exclusion criteria included hemophiliacs, individuals who had received chemotherapy during the previous 4 weeks, and participants with rashes, gauze dressings, casts, edema, open sores or wounds, etc. The laboratory staff included certified medical technologists and phlebotomists. Members of the laboratory staff each completed comprehensive training before working in the MEC.
To measure the dietary levels of folate and vitamin B12, a 24 h dietary recall interview was conducted by NHANES by employing a computer-based system [
A number of covariates were used to control for differences in a variety of potential mediating factors. The demographic covariates were age and race, and subjects were stratified according to sex. Five categories were used by NHANES to distinguish among races and ethnicities: non-Hispanic White, non-Hispanic Black, Mexican American, Other race or Multiracial, and Other Hispanic. Several additional variables were measured to allow differences to be controlled, including hours of fasting before the blood draw, alcohol use, body mass index, smoking pack-years, total METs of physical activity, diabetes status, and, for females, pregnancy status.
Food intake can influence serum concentrations of vitamins, so hours of fasting prior to the blood draw were measured by NHANES. To encourage an overnight fast, blood draws were administered in the morning. Average fasting time was
BMI was indexed using the formula
Long-term exposure to the effects of smoking was indexed by quantifying pack-years of cigarette smoking. Pack-years was estimated as the number of years the individual reported smoking times the average number of cigarettes reported smoking per day, divided by 20.
Participation in leisure time physical activity was used as a covariate. A total of 48 different physical activities were listed, and subjects identified the ones they participated in, if any, in the past month. Subjects were also required to identify if the intensity of the activity was moderate or vigorous, using NHANES explanations to guide them. Additionally, subjects reported the average duration of the activity. NHANES did not count participation in an activity if it lasted less than 10 minutes. A MET score was calculated for each individual, and total MET-minutes per week were estimated for each participant using the compendium of physical activities [
Alcohol use was also employed as a covariate. Subjects were separated into three categories based on their self-reported alcohol drinking habits: heavy drinkers, moderate drinkers, and abstainers. Men who indicated they drank three or more alcoholic drinks per day over the past 12 months were labeled heavy drinkers, whereas women were labeled heavy drinkers if they consumed two or more alcoholic beverages per day. Men who drank more than zero and less than three alcoholic beverages per day were defined as moderate drinkers, whereas women were moderate drinkers if they drank more than zero but less than two drinks per day over the past 12 months. Adults who indicated they did not drink alcohol in the past 12 months were labeled abstainers.
Participants who had fasting blood glucose levels of 126 mg/dL or higher were defined as diabetic. Additionally, adults who indicated they had been told by their physician that they were diabetic, or they were taking insulin or a prescription medication for their diabetes, were considered diabetic. The diabetes status covariate was treated as a dichotomous variable: diabetic or normal.
Female participants reported whether they were pregnant. Pregnancy status was treated as a dichotomous covariate: pregnant or not pregnant.
NHANES used a multilevel random sampling strategy to produce findings generalizable to the noninstitutionalized, civilian, adult population of the United States. In general, counties within the United States were randomly selected initially, then roads, then houses, and then individuals. Individual sampling weights, which were assigned to each participant by NHANES, were included in each analysis resulting in unbiased national estimates. SAS SurveyFreq was employed to produce weighted frequencies to describe the categorical data. SAS SurveyMeans was used to generate weighted means (±SE) to describe continuous variables.
Serum and dietary levels of folate and vitamin B12 were the exposure variables. Their distributions deviated from normal, so they were log-transformed before inclusion in the analyses. Leukocyte telomere length was the outcome variable.
The SAS SurveyReg procedure, taking into account strata, clusters, and individual sampling weights, was utilized to calculate the extent of the linear associations between folate and vitamin B12 and telomere length, including the linear relationships between serum and dietary levels of the vitamins. In the present investigation, with age and race controlled, women had significantly longer telomeres than men. Women and men also differed in their levels of folate and vitamin B12. Hence, the relationships between folate and vitamin B12 and telomere length were evaluated separately for women and men.
Although a sample of more than 5500 individuals was used in the present investigation, statistical power was not extreme. In particular, the multistage, nested sampling strategy of NHANES produced 29 degrees of freedom (df) in the denominator for all analyses, including those delimited to women only and men only. The 29 degrees of freedom were derived by subtracting the 28 NHANES strata from the 57 NHANES clusters.
Differences in the demographic variables (age and race) were controlled statistically using partial correlation and the SAS SurveyReg procedure. Adjustments were also made for additional potential mediating variables using partial correlation, including the number of hours fasted before the blood draw, BMI, smoking, physical activity, alcohol use, diabetes status, and pregnancy status (for females) to test their effect on the vitamin and telomere length associations. Sex was controlled by stratifying the sample into two categories, women and men.
Alpha was set at the 0.05 level and all
Individual sample weights were used to produce findings generalizable to the noninstitutionalized, civilian adult population of the United States. Using the total sample, average age (±SE) was
The relationship between serum folate and dietary folate was 0.20 (
Median levels of the folate and vitamin B12 variables are displayed in Table
Percentiles for the exposure and outcome variables representing U.S. women and men.
Variable | Percentile (±SE) | ||||
---|---|---|---|---|---|
5th | 25th | 50th | 75th | 95th | |
Women ( |
|||||
Men ( |
|||||
Combined ( |
|||||
Women ( |
|||||
Men ( |
|||||
Combined ( |
|||||
Women ( |
|||||
Men ( |
|||||
Combined ( |
|||||
Women ( |
|||||
Men ( |
|||||
Combined ( |
|||||
Women ( |
|||||
Men ( |
|||||
Combined ( |
Note: SE: standard error. Table values include person-level weighted adjustments based on the sampling methods of NHANES so that values represent those of the U.S. adult population.
The association between chronological age and telomere length was linear and inverse. Telomeres were 15.6 base pairs shorter for each additional year of chronological age within the sample (
In the present sample, men and women differed significantly in the length of their telomeres. Specifically, after adjusting for differences in age and race, average telomere length was significantly shorter in men than women, as shown in Table
Mean differences in telomere length (base pairs) between U.S. men and women (
Men | Women | |||
---|---|---|---|---|
Telomere length | ||||
Model 1 | 4.6 | 0.0402 | ||
Model 2 | 6.8 | 0.0146 |
SE: standard error. Means have been adjusted for the covariates, as defined by each model. Model 1 included adjustment for differences in age and race. Model 2 included adjustment for differences in age and race, as well as pack-years of smoking, alcohol use, BMI, total physical activity, hours of fasting, and diabetes status. Mean differences between women and men for Models 1 and 2 are significant (
As displayed in Table
Relationship between serum and dietary levels of folate and vitamin B12 (per 10% increase) and telomere length (base pairs) in U.S. women and men.
Telomere length (base pairs) | ||||
---|---|---|---|---|
Exposure variable | Regression coefficient | |||
Serum folate (ng/mL) | ||||
Model 1 | 9.57 | 2.75 | 10.5 | 0.0030 |
Model 2 | 8.75 | 2.94 | 8.7 | 0.0062 |
Dietary folate (mcg) | ||||
Model 1 | 5.41 | 2.41 | 5.0 | 0.0325 |
Model 2 | 5.98 | 2.89 | 4.3 | 0.0479 |
Serum vitamin B12 (pg/mL) | ||||
Model 1 | 3.48 | 2.01 | 2.8 | 0.1056 |
Model 2 | 5.59 | 3.15 | 3.1 | 0.0867 |
Dietary vitamin B12 (mcg) | ||||
Model 1 | 2.12 | 1.02 | 4.3 | 0.0468 |
Model 2 | 1.94 | 1.17 | 2.8 | 0.1080 |
Serum folate (ng/mL) | ||||
Model 1 | 5.81 | 3.39 | 2.9 | 0.0972 |
Model 2 | 4.56 | 3.35 | 1.9 | 0.1838 |
Dietary folate (mcg) | ||||
Model 1 | 2.27 | 2.86 | 0.6 | 0.4341 |
Model 2 | 0.71 | 2.99 | 0.1 | 0.8149 |
Serum vitamin B12 (pg/mL) | ||||
Model 1 | -1.92 | 3.58 | 0.3 | 0.5948 |
Model 2 | -2.32 | 3.59 | 0.4 | 0.5227 |
Dietary vitamin B12 (mcg) | ||||
Model 1 | 0.72 | 1.66 | 0.2 | 0.6665 |
Model 2 | 0.16 | 1.66 | 0.0 | 0.9239 |
SE: standard error. For Model 1, the covariates were age and race. For Model 2, in additional to age and race, the model was adjusted for differences in pack-years of smoking, alcohol use, BMI, total physical activity, hours fasted, diabetes status, and pregnancy status (for women). Interpretation of the regression coefficients is as follows for the first row (Model 1) regarding serum folate (ng/mL) in women: after adjusting for differences in age and race, for each 10% increase in serum folate, telomeres were 9.57 base pairs longer, on average.
As displayed in Table
In women, after adjusting for differences in age, race, smoking, alcohol use, BMI, total physical activity, hours fasted, diabetes status, and pregnancy status, for each 10% increase in serum vitamin B12, telomeres were 5.59 base pairs longer, on average (
After controlling for age and race, vitamin B12 intake was linearly and positively related to telomere length in women (
The primary purpose of this investigation was to determine the relationship between serum and dietary concentrations of folate and vitamin B12 and leukocyte telomere length, an objective marker of cellular aging, within a large, random, national sample of U.S. women and men, ages 20-84. Findings differed between women and men (Table
As expected, telomeres were progressively shorter for each additional year of chronological age. The relationship between age and telomere length was inverse, linear, and substantial. Specifically, for each additional year of age, telomeres were 15.6 base pairs shorter, on average. Interpretation of these results is straightforward. For example, over any 10-year span of chronological age, the average difference in telomere length was roughly 156 base pairs in the present sample (
Given the robust, linear associations between folate, serum and dietary, and telomere length in women, it follows that cellular aging was less in women who had high levels of folate compared to their counterparts. Specifically, with all the covariates controlled, for each 10% increase in serum folate, telomeres were 8.75 base pairs longer, on average (Table
The estimated cellular aging difference based on dietary folate levels in women was less than those based on serum concentrations. Multiple factors likely account for this finding. Dietary estimates were based on a self-reported dietary recall compared to the blood folate values, which were objectively measured. In short, serum concentrations represent the real-time status of folate in the body. Serum levels reflect the net outcome of bioavailability and metabolism of folate. However, as indicated in a recent review by Thurnham and Northrop-Clewes, nutrient biomarkers are not foolproof. They can be thrown off by infection and inflammation. Moreover, in another review, Elmadfa and Meyer suggest that nutrient biomarkers and dietary assessments complement each other. Both have their strengths and weaknesses. Consequently, both were used in the present study.
Although the relationships between folate and vitamin B12 and telomere length did not reach statistical significance in men, most of the associations were meaningful in women. Women in the highest quartile of dietary folate reported more than twice the folate intake compared to those in the lowest quartile. Given the relationships in women tended to be linear, estimates of the cellular aging advantages were substantial. Few other dietary factors are associated with greater aging benefits. For example, a recent publication focused on the consumption of nuts and seeds and telomere length. For each 10% increment in energy derived from nuts and seeds, biologic aging decreased by 3.3 years, on average [
The U.S. Recommended Daily Allowance (RDA) for folate in the U.S. is 400 mcg for adults. According to the dietary findings in the present national sample, median (±SE) levels were low in women and very close to the RDA in men. Median intake was
U.S. men and women seem to be doing better regarding vitamin B12 consumption. The RDA for vitamin B12 in the U.S. is 2.4 mcg for adults. Median (±SE) intake in the present sample was
Mean dietary folate and vitamin B12 levels were higher in men than in women, and U.S. women were much more likely to have intakes below the RDA for folate and vitamin B12. This may partially account for the demonstrated stronger relationships between folate and vitamin B12 and telomere length in women in the present study. In short, women could be more susceptible to DNA damage resulting from lower dietary levels of folate or vitamin B12. The obvious difference in the link between folate and vitamin B12 and telomere length between women and men warrants further research.
Although U.S. women reported substantially lower dietary intakes of folate and vitamin B12 than men, women had higher serum concentrations of folate than men. Regarding serum levels of vitamin B12, men and women had similar median levels, but women had higher highs and lower lows, more extreme levels. How can this be? Specifically, how could U.S. women report lower dietary consumption of folate yet have higher serum concentrations of folate? Although there are multiple possibilities, one reason stands out: underreporting. When reporting dietary intake, both men and women tend to report eating less food than they actually consume [
The correlation between dietary folate and serum concentrations of folate was stronger in women than in men. This would suggest that dietary intake of folate, at least self-reported consumption, plays a more significant role in serum concentrations among women than men. The correlation was weak between dietary vitamin B12 and serum levels of B12 in both men and women, indicating that there is not much overlap between the two.
A number of investigations indicate that diets with significant antioxidant levels tend to promote cell longevity, including vegetables [
Dietary folate and vitamin B12 both seem to possess antioxidant qualities. Consequently, each has the potential to prevent the shortening of telomeres and the reduction of premature biologic aging. For example, Cagnacci et al. [
To date, few investigations have studied directly the relationship between folate and vitamin B12 levels and telomere length, and none have used a U.S. national sample. Moreover, results have been inconsistent. For example, using a subsample of 1715 women from the Nurses’ Health Study, Liu et al. concluded that plasma and also dietary concentrations of folate and vitamin B12 were not related to telomere length [
On the other hand, Richards et al. studied a sample of 1319 healthy adults from the United Kingdom. A significant relationship was reported between serum folate levels and telomere length, after adjusting for differences in age [
An investigation by Paul et al. included 195 men from Italy. The researchers found a nonlinear relationship between plasma concentrations of folate and telomere length. When folate levels were above the median, the relationship with telomere length was positive, but when folate levels were below the median, the association was inverse [
Lastly, using 1044 older men and women from the Framingham Offspring Study (mean age: 59 years), the scientists reported that the relationship was inverse and linear when delimited to the 2nd–5th quintiles of plasma folate levels. However, the association was not significant when the 1st quintile was included. Findings were not reported for men and women separately, and the association between vitamin B12 and telomere length was not reported [
In the present study, as levels of folate and vitamin B12 decreased in women, telomeres were progressively shorter. Multiple mechanisms could account for this finding. For example, oxidative stress plays a key role in the length of telomeres and in cell aging [
The same mechanism can also be considered using a reversed perspective. Specifically, the association between oxidative stress and lower concentrations of serum folate and vitamin B12 could be a result of folate and B12 serving as antioxidants. Once oxidized, folate and B12 are no longer biologically active. In short, rather than low folate or B12 concentrations leading to oxidative stress, oxidative stress could lead to lower levels of serum folate and B12. Since women may consume significantly less folate and vitamin B12 than men, sometimes, these micronutrients may not be maintained at adequate concentrations in women. This could explain why low levels of folate and vitamin B12 were linked to short telomeres in women but not in men.
Other factors accounting for the folate, vitamin B12, and telomere associations are also possible. From a molecular perspective, Moore et al. offer multiple related mechanisms to explain the association between folate concentrations and telomere length. They indicate that low folate levels could lead to shorter telomeres by removal of increased uracil in the telomere hexamer repeat (TTAGGG), which has been shown to produce DNA breaks and abasic sites (AP sites). Additionally, the association could be partly due to incompetent binding of shelterin proteins to telomeric DNA due to decreased affinity with uracil and AP sites because of uracil repairs in the telomere repeat [
The relationship between folate and vitamin B12 and biologic aging was consistent and meaningful in women, but not in men. Given oxidative stress is a key factor associated with shorter telomeres, and low levels of folate and vitamin B12 each tend to be linked to higher levels of oxidative stress, it follows that the relationship between folate and vitamin B12 and telomere length would be expected to be stronger in men. Undoubtedly, other factors are involved in the folate, B12, and telomere associations, especially in men.
The current study had several limitations. First, causal conclusions are not valid because the investigation used a cross-sectional design. Additionally, the significant findings could be partly a result of residual confounding. In other words, high levels of folate and vitamin B12 could be biomarkers of a healthy lifestyle, which could be the underlying factor accounting for the longer telomeres found in the women of this investigation. Similarly, although multiple potential confounders were controlled in the present study, there could be other factors that might explain the positive association between folate and B12 and telomere length.
The present investigation also had a number of strengths and novel characteristics: (1) Participants were randomly selected. Hence, the results are generalizable to all noninstitutionalized, civilian adults of the United States. (2) The random sample included 5581 adults. As a result, findings are stable. (3) Chronological age was strongly related to telomere length, attesting to the precision and validity of the telomere measurement protocol by NHANES and the Blackburn Laboratory at the University of California, San Francisco. (4) Using telomere base pairs, biological aging differences were presented in chronological years, facilitating a straightforward understanding of the strength of the relationship between folate and vitamin B12 and cell aging. (5) The sample incorporated all races and adults 20-84 years old. (6) A total of 10 potential confounding variables were controlled statistically, providing evidence that the key results of the present study are robust. (7) Both serum concentrations and dietary levels of folate and vitamin B12 were studied.
In conclusion, results from a nationally representative sample of U.S. adults indicate that cellular aging tends to increase linearly as levels of folate and vitamin B12 decrease in women, but not in men. The observed relationship in women could be a function of low concentrations of folate and vitamin B12 resulting in higher oxidative stress levels, leading to accelerated cell aging and shorter telomeres. The reverse is also logical—oxidative stress leading to lower levels of folate and vitamin B12, resulting in telomere attrition, particularly in women. The relationship could also be due to these vitamins influencing the structure and function of DNA and the epigenetic regulation of telomeres through DNA methylation. Whatever the mechanism, it appears that the micronutrients folate and vitamin B12, especially the former, play important roles in the biologic aging of women. The role of these vitamins in connection with telomere length and aging in men is less apparent. Clearly, additional research focusing on folate and vitamin B12 and biologic aging is warranted, particularly randomized controlled trials, so that causality can be deciphered.
The data used to support the findings of this study were supplied online by the U.S. Centers for Disease Control and Prevention, as part of the National Health and Nutrition Examination Survey (NHANES). The NHANES data and related documentation are free to the public and available at
The research was performed as part of employment of the author at Brigham Young University, Provo, Utah.
The author declares no conflict of interest.
A special thanks is due to all the women and men who sacrificed their time to participate in the collection of the NHANES data.