Individual nutrients and serum klotho levels in adults aged 40–79 years

Abstract Several dietary factors (including adherence to the Mediterranean diet or higher nut intake) seem to positively affect circulating antiaging Klotho protein levels; yet, a description of possible relationships between individual nutrients and Klotho activity has not been evaluated. We analyzed the association of dietary intake of individual macro‐ and micronutrients and nonnutritive food components with circulating Klotho levels in a sample of 40‐ to 79‐year‐old US adults. Data from the 2015–2016 National Health and Nutrition Examination Survey were analyzed. Nutrient/food component intakes were calculated in relation to total energy intake using the nutrient density method, and available pristine serum samples were analyzed for serum Klotho concentrations. The final study sample consisted of 2637 participants (mean age 59.0 ± 10.7 years; 52% women). Higher Klotho concentrations were found with higher intake of carbohydrates (p < .001), total sugars (p < .001), dietary fibers (p < .001), vitamin D (p = .05), total folates (p = .015), and copper (p = .018). The results of the regression analysis with a crude model showed significant associations among five nutrients/food components (carbohydrates, alcohol, total sugars, dietary fibers, and niacin) and soluble Klotho levels across the sample. After adjusting the models for age and gender, the nutrient/food component–Klotho association remained significant for carbohydrates, total sugars, and alcohol (p < .05). Dietary exposure to individual nutrients and nonnutritive food components appears to be associated with Klotho activity; however, additional research is needed to investigate the relationship between cause and effect in diet composition–Klotho interplay.

several pathogenic processes (Dërmaku-Sopjani et al., 2013). The prevention of Klotho decline can, therefore, be a novel therapeutic strategy for many age-related diseases (Kuro-O, 2019). Klotho exists in two forms, as a membrane-bound enzyme/transporter and as a soluble hormone-like carrier, with the circulating form often used as a proxy for the turnover of Klotho proteins (Olauson et al., 2017) and a biomarker of aging (Veronesi et al., 2021). Klotho expression appears to be negatively affected by age, bone loss, and alcohol consumption (Chalhoub et al., 2016;González-Reimers et al., 2018). Several dietary factors (including adherence to the Mediterranean diet and/ or nut intake) seem to positively affect Klotho levels Jurado-Fasoli, Amaro-Gahete, De-la-O, Martinez-Tellez, et al., 2019). Klotho activated by nutrition or other factors might further affect several metabolic processes, including bioenergetics balance (Ostojic, 2021). Previous studies linking Klotho with diet have originated from animal research or recruited a small number of human participants, without a description of possible relationships between specific nutrients and Klotho activity. Therefore, the main aim of the present cross-sectional study was to analyze the asso- NHANES is selected using a complex four-stage sample design, in which sample weights were used to produce estimates of healthrelated statistics that would have been obtained if the entire sampling frame had been surveyed (Chen et al., 2020

| Demographics
NHANES 2015-2016 Demographics Data and Body Measures components were explored to acquire data on the general characteristics of participants, including individual, family, and household-level information, and weight, height, and body mass index data.

| Statistical analyses
Descriptive statistics were used to explore the distribution of Klotho levels in the sample, sample characteristics, and dietary intake. Oneway ANOVA was used to compare serum Klotho values across quartiles of dietary intake for each nutrient/food component (except for alcohol where we compare data below or above the median), with post hoc pairwise comparison tests employed to identify differences between individual sample pairs. The Kruskal-Wallis H test was used for testing trends across quartile categories. Bivariate analyses were used to identify relevant covariates, with the regression models adjusted for a posteriori recognized set of covariates. To avoid confounding of the nutrient-Klotho association due to variation in absolute food intake related to physical activity level and/or body size, we applied nutrient intake relative to energy intake as an independent variable in all models. Simple (crude) and multivariate linear regression analyses were performed to test the association between energy-adjusted individual nutrients/food components and serum In models where the interaction term was not significant, the interaction term was omitted and the nutrient/food component-Klotho association for both genders was presented combined. Data were analyzed using SPSS Statistics for Mac (Version 24.0; IBM), with the significance level set at p < .05.

| RE SULTS
The final study sample consisted of 2637 participants (52% women) who provided dietary data and soluble Klotho levels (Flowchart, Supplementary material). The demographic and basic dietary characteristics of the study participants are shown in Table 1. The mean soluble serum Klotho level across all participants was 827.1 ± 324.9 pg/mL (95% confidence interval [CI] from 814.7 to 839.5) ( Figure 1). Significantly lower serum Klotho levels were found in men than in women (797.4 ± 302.8 pg/mL vs. 854.4 ± 341.9 pg/ mL; p < .001). Significant gender differences were found for energyadjusted intakes of several nutrients/food components (p < .05), including carbohydrates, alcohol, total sugars, dietary fibers, polyunsaturated fatty acids, alcohol, and several micronutrients and other food components (Table S1). For most nutrients/food components, women had a higher nutrient density than men.
Mean serum Klotho levels varied across quartiles of dietary intake for several nutrients/food components, including carbohydrates, total sugars, dietary fibers, vitamin D, and calcium ( Figure 2). A significant trend for higher Klotho levels with higher intake of a specific nutrient was found for carbohydrates (p < .001), total sugars (p < .001), dietary fibers (p < .001), vitamin D (p = .05), total folates (p = .015), and copper (p = .018) ( Table S2). The participants who consumed over 1.39% of the total energy intake from alcohol per day (median intake) had significantly lower serum Klotho concentrations as compared to participants who consumed less alcohol (770.2 ± 313.2 pg/mL vs. 848.0 ± 331.8 pg/mL; p < .001). Serum Klotho levels across quartiles for other nutrients/food components are presented in Table S2.
A bivariate analysis identified age (r = −.077; p < .001) and gender (r = .081; p < .001) as relevant covariates for a regression model, while other variables (e.g., race, educational level, family income, body mass index) did not provide a significant contribution to the model and were excluded for regression analyses (p > .05).
The results of the regression analysis with a crude model showed a significant association among five nutrients/food components (carbohydrates, alcohol, total sugars, dietary fibers, and niacin) and soluble Klotho levels across the sample ( Figure 3); the crude regression models for all nutrients and food components are presented in Table S3. We identified a significant interaction by gender on the nutrient-Klotho association for dietary fibers, phosphorus, and potassium. The specific regression coefficients stratified by gender are TA B L E 1 Demographic and basic dietary characteristics of the study sample. presented in Table 2, with significant associations found only in men (p < .05). In the other multivariable models, the interaction term was not significant, and the interaction was omitted from the models, with the nutrient/food component-Klotho association analyzed for both genders combined. After adjusting the models, we found that the nutrient/food component-Klotho association remained significant for carbohydrates, total sugars, and alcohol (p < .05) (Table S3).

| DISCUSS ION
The present study is, to our knowledge, the first cross-sectional report that demonstrated significant associations between several individual nutrients or food components and serum Klotho levels at a populational level. We found that middle-aged U.S. adults who report having a higher intake of carbohydrates, total sugars, and less alcohol have higher levels of circulating Klotho, compared to individuals with lower intakes of carbohydrates and total sugars, and more alcohol, after controlling for age and gender. Significantly lower women. The study also demonstrated no significant associations between total energy intake (also macronutrients) with Klotho plasma levels, except for a direct association found between alcohol intake F I G U R E 1 Histogram of serum Klotho levels in U.S. adults aged 40-79 years (n = 2637). In terms of macronutrients, we found a significant positive association between carbohydrates and total sugars and serum Klotho  sample . Gender-specific differences in diet-health outcomes are, however, not implausible (Vinke et al., 2020). Male participants in our sample had lower mean Klotho levels combined with lower nutrient density for dietary fiber, phosphorus, and potassium compared with women. Given the small effect sizes for phosphorus and potassium, these associations should be interpreted with caution. The positive association with dietary fiber was, however, stronger than the associations observed for carbohydrates and total sugars in the whole sample. We have no explanation for this but it may be that the association between dietary fiber intake and Klotho might plateau for dietary intake above a certain level in line with higher energy-adjusted fiber intake among women. Potential gender differences in the observed diet-Klotho associations should be confirmed in future studies.

F I G U R E 2
As opposed to the more robust findings observed for carbohydrates, including total sugars, and for alcohol in this study, the links between Klotho levels and the intake of dietary fibers, vitamin D, total folates, copper, and niacin must be interpreted with great caution.
Our findings on dietary fiber remain unclear, as previously discussed.
This is also the case for vitamin D, although Klotho's suggested role in calcium and phosphate homeostasis, through suppressing serum vitamin D in the form of 1,25(OH)(2)D has been known for more than a decade (Martin et al., 2012), also confirmed by the negative association between serum 1,25(OH)(2D) and Klotho levels recently Although we analyzed a relatively large and heterogeneous sample of middle-aged US adults using an extensive list of independent variables (dietary exposures) and covariates, the present study is not without limitations. First, the cross-sectional design of our study prohibits any assertion of causality. Second, a 24-hour dietary recall used in this study to collect dietary information is based on selfreports and prone to bias, and has low reliability at the individual level, due to large day-to-day variations in diet, and hence a two-day intake does not capture the usual diet. This is especially a limitation for episodically consumed foods and beverages, like alcohol. On the contrary, the intake of carbohydrates, including total sugars, fluctuates less than alcohol on a day-to-day basis. Hence, the findings for carbohydrates and total sugars may be better grounded than the observations for alcohol. Furthermore, we analyzed a sample of middle-aged adults which limits the interpretation of our findings to this population; additional studies are highly warranted to extrapolate our results to younger cohorts, elderly over 80 years, and clinical populations. Finally, well-designed longitudinal observational studies including advanced biomarkers of Klotho activity and expression, in addition to detailed and rigorous dietary assessment, are necessary to advance our understanding of the possible effects of nutrients and food components on serum Klotho levels and health.

| CON CLUS ION
In conclusion, energy-adjusted intake of carbohydrates, total sug-

ACK N OWLED G EM ENT
None.

FU N D I N G I N FO R M ATI O N
None.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data analyzed during this study are included in this article. Further enquiries can be directed to the corresponding author.

E TH I C S S TATEM ENT
The ethical approval to conduct the current round of NHANES 2015-2016 was granted by the NCHS Research Ethics Review Board (Continuation of Protocol #2011-17).

CO N S E NT TO PA RTI CI PATE S TATE M E NT
Informed consent was obtained from all respondents to participate in the study. The research was conducted ethically in accordance with the World Medical Association Declaration of Helsinki.