Vitamin D3 for Life

Live Enlightened

Research libraryVitamin D3 For Life
Study design and methods

Secondary relevance to vitamin D This paper mainly addresses another topic and also reports a vitamin D analysis, result or scientific synthesis.

This paper mainly addresses another topic and also reports a vitamin D analysis, result or scientific synthesis.

Why this paper is included

Supporting source evidence:

Conversely, vitamin B6 (OR = 0.91; 95% CI: 0.86 to 0.96; p = 1.05 × 10−3) and vitamin D (OR = 0.94; 95% CI: 0.88 to 1.00; p = 0.03) showed inverse associations with chronotype, indicative of a preference for eveningness.

Source: Abstract. View original paper

Relevance classification does not establish study quality, causation or editorial approval.

Study design Genetic instrumental-variable study.

Genetic instrumental-variable study

Uses genetic variation as an instrument to investigate relationships; this is not a randomised treatment trial.

Meta-analysis unclear · Blinding unclear / not reported · Research setting awaiting verification

How this paper was classified

Automatic source-based classification; awaiting editorial verification. This is not a quality score.

Genetically Supported Causality Between Micronutrients and Sleep Behaviors: A Two‐Sample Mendelian Randomization Study

Source: Article title.

View supporting evidence in the original paper

What the study types mean
Systematic review
Uses explicit methods to identify, assess and summarise research on a defined question.
Scoping review
Maps the extent and characteristics of research on a topic.
Narrative review
Discusses existing research; systematic-review methods have not been established by this classification.
Review / evidence synthesis
Reviews existing research; the exact review methods need further assessment.
Randomised controlled trial
Assigns participants or groups to comparison interventions by chance.
Non-randomised intervention study
Examines an intervention without random allocation.
Cohort study
Examines exposures and outcomes over time within a defined group.
Case-control study
Compares people with an outcome against people without it, examining exposures.
Cross-sectional study
Examines a population at a particular point in time.
Observational study
Observes exposures and outcomes without assigning the exposure; the subtype is not established here.
Case report
Describes an individual case.
Case series
Describes a group of individual cases.
Secondary analysis
Analyses existing study data; this paper is not automatically another independent trial.
Laboratory study
Investigates biological processes in a laboratory setting.
Qualitative study
Examines experiences or meanings using qualitative methods.
Genetic instrumental-variable study
Uses genetic variation as an instrument to investigate relationships; this is not a randomised treatment trial.
Study protocol
Describes planned methods and is not evidence that the planned study has reported results.
Multiple reported designs
More than one design is explicitly reported; examine the supporting methods.
Study design awaiting verification
The available source does not establish a sufficiently clear study design.

Meta-analysis combines numerical results from studies. Blinding describes who was unaware of treatment allocation. Neither is a separate quality grade. Study design, vitamin D relevance and document format are independent filters.

Full studies

Genetically Supported Causality Between Micronutrients and Sleep Behaviors: A Two‐Sample Mendelian Randomization Study

Genetic instrumental-variable study · Brain and Behavior · 2025

Original publication title

Genetically Supported Causality Between Micronutrients and Sleep Behaviors: A Two‐Sample Mendelian Randomization Study

Ruijie Zhang; Liyan Luo; Lu Zhang; Xinao Lin; Chuyan Wu; Feng Jiang; Jimei Wang

© 2025 The Author(s). Brain and Behavior published by Wiley Periodicals LLC.

Original publication · Publisher record · CC BY 4.0

Source wording retained. Formatting, section navigation and display-title punctuation adapted for this research record.

Authors’ conclusion

Our two‐sample Mendelian randomization analysis suggests that individuals with higher folate levels tend to be morning‐oriented, which may be beneficial for enhancing sleep quality. Conversely, those with elevated levels of vitamin B6 and vitamin D tend to be evening‐oriented. Future research should include diverse populations and also consider conducting controlled, randomized trials to precisely determine micronutrient intake levels, providing a robust framework to validate the conclusions derived from Mendelian randomization.

An extract from the source publication. Read it in context.

At a glance

Study background

What background does the paper provide?

Sleep patterns are defined by their length, often calculated as the total amount of sleep within a 24‐h cycle, and by chronotype. Chronotype refers to an individual's natural predisposition to be active at particular times during a 24‐h cycle (Günal 2023). “Morning types,” commonly known as “larks,” tend to be more energetic and productive in the early morning and typically prefer to go to bed and wake up early. “Evening types,” often referred to as “owls,” are more active and focused in the evening, usually preferring to go to bed and rise later (Günal 2023; Wu et al. 2024).

Numerous factors, including age, gender, and ethnicity, influence sleep behavior (Hirshkowitz et al. 2015; Grandner 2019; Walch, Cochran, and Forger 2016). Notably, the duration and quality of sleep are significantly determined by genetic factors. Meta‐analyses have demonstrated that approximately 46% of the variance in sleep duration among individuals is attributable to genetic differences, and 44% of the variance in sleep quality is determined by genetics (Breitenstein, Doane, and Lemery‐Chalfant 2021). This finding is corroborated by genetic studies that employed genome‐wide association studies (GWAS) to analyze self‐reported sleep duration data from the UK Biobank.

The results were validated for consistency in separate research involving both adults and children/adolescents (Kocevska et al. 2021; Dashti et al. 2019).

The impact of micronutrient intake on sleep has received less attention than that of macronutrients. However, emerging studies suggest that shorter sleep durations in adults are associated with deficiencies in several micronutrients, specifically calcium, magnesium, and vitamins D and K (Ikonte et al. 2019; Al Hinai et al. 2024). Furthermore, numerous clinical trials have provided evidence supporting the assertion that adequate concentrations of zinc and iron can enhance sleep quality in infants and adolescents (Innocenti et al. 2023; Ji and Liu 2015; Kordas et al. 2009).

Traditional observational studies may be influenced by various confounding factors, which could introduce biases. To date, there has not been a comprehensive review of the literature examining the connection between micronutrients and sleep behaviors from a developmental standpoint.

Mendelian randomization (MR) utilizes single nucleotide polymorphism (SNP) statistics from GWAS to infer possible causal connections among various complex traits (Gupta, Walia, and Sachdeva 2017; Yao et al. 2022). SNPs, due to their random assignment during meiosis and fertilization prior to birth, are less susceptible to alterations by disease conditions, environmental factors, or other confounders (Davies, Holmes, and Davey Smith 2018). Therefore, they may offer a more precise understanding of true causal relationships compared to conventional observational studies.

In our research, we investigated the genetic underpinnings and potential causal links between micronutrients and sleep behaviors. This was achieved by examining genetic correlations and polygenic overlaps using GWAS summary statistics. Additionally, we employed two‐sample and multivariable MR (MVMR) analyses to explore the possible causal effects of micronutrients on sleep behaviors.

Source: Introduction Source: Introduction Source: Introduction Source: Introduction

Study methods

What methods does the paper describe?

2.1

Study Design

A schematic overview of the research design is depicted in Figure 1. We conducted an MR investigation utilizing data from 18 publicly accessible GWAS to acquire summary statistics. In the two‐sample MR analysis, data concerning sleep behaviors were sourced from three independent GWAS consortia. These data were employed for initial analyses and replication studies, and ultimately integrated into a meta‐analysis to consolidate findings. For the MVMR analysis, we investigated the causal relationships between various micronutrients and sleep behaviors. To increase the statistical robustness of our findings, we combined estimates from various data sources.

All studies involved in the GWAS had previously been approved by the appropriate review boards, and no further ethical approval or participant consent was necessary for this analysis.

FIGURE 1

A flow chart of the study design.

Figure image at the original source: open the original paper. This file is not included in this text import.
2.2

GWAS Data of Sleep‐Related Behaviors

Participants recorded their own sleep duration by noting the total hours slept every 24 h, including naps, measured in hourly increments. We carried out distinct GWAS for individuals of European descent, categorizing them based on their sleep duration. This included a group with short sleep duration (less than 7 h per night, n = 106,192 cases) and another with long sleep duration (more than 9 h per night, n = 34,184 cases). These were compared to a control group that reported sleeping between 7 and 9 h per night (n = 305,742) (Dashti et al. 2019).

We excluded responses that reported extremely low (less than 3 h) or extremely high (more than 18 h) sleep durations. Responses marked as “Do not know” or “Prefer not to answer” were considered missing data. We can obtain the GWAS summary statistics from the Sleep Disorder Knowledge Portal at https://sleep.hugeamp.org/downloads.html.

To assess chronotype, subjects responded to the question, “Do you consider yourself to be?” with choices spanning from “Definitely a ‘morning’ person” to “Definitely an ‘evening’ person.” Responses were scored, with “Definitely a ‘morning’ person” and “More a ‘morning’ than ‘evening’ person” classified as cases, assigned two and one points, respectively. Conversely, those identifying as “Definitely an ‘evening’ person” or “More an ‘evening’ than ‘morning’ person” were designated as controls, with scoring of −2 and −1 points, respectively. The study by Jones et al. (2019) encompassed 252,287 cases and 150,908 controls of European ancestry.

The GWAS summary statistics for this analysis are also available on the Sleep Disorder Knowledge Portal.

In the assessment of insomnia, participants answered the question: “Do you have trouble falling asleep at night or do you wake up in the middle of the night?” The response options provided were “never/rarely,” “sometimes,” “usually,” or “prefer not to answer.” Participants who chose “usually” were identified as cases of insomnia, while those who answered “never/rarely” were designated as controls. The GWAS, carried out by Watanabe et al. (2022), involved 593,724 cases and 1,771,286 controls of European descent and identified 554 loci associated with insomnia.

The GWAS summary statistics are available from https://ctg.cncr.nl/software/summary_statistics/.

In our research, we conducted a comprehensive search through PubMed and the IEU OpenGWAS project (https://gwas.mrcieu.ac.uk/) to identify the most recent large‐scale GWAS studies on micronutrients, with the final search conducted in April 2024. To avoid overlap between the exposures and outcomes in our study, we excluded certain micronutrients that were obtained from these specific databases during our data collection. We observed a notable absence of genomic studies concerning chloride; fluoride; phosphorus; sulfur; vitamins B1, B2, B3, B5, and B7; and iodine.

Furthermore, studies on vitamin K, cobalt, chromium, sodium, and molybdenum were not included due to their inconclusive genomic results (Dashti et al. 2014; Ng et al. 2015). This selection process led us to preliminarily identify 15 micronutrients for further investigation: copper (Evans et al. 2013), calcium (O'Seaghdha et al. 2013), carotene (Ferrucci et al. 2009), folate (Grarup et al. 2013), iron (Bell et al. 2021), magnesium (Meyer et al. 2010), potassium (Meyer et al. 2010), selenium (Evans et al. 2013), vitamin A (Mondul et al. 2011), vitamin B12 (Grarup et al. 2013), vitamin B6 (Hazra et al. 2009), vitamin C (Zheng et al. 2020), vitamin D (Revez et al. 2020), vitamin E (Major et al. 2011), and zinc (Evans et al. 2013).

2.3

Two‐Sample MR Analysis

In our research, we utilized the two‐sample MR approach, employing the Two Sample MR (version 0.5.11) package (Hemani, Tilling, and Smith 2017; Hemani et al. 2018). Our principal technique for MR analysis was the inverse variance weighting (IVW) method. This meta‐analytic approach combines Wald estimates from each SNP to compute overall effect estimates of the exposure on the outcome (Burgess, Butterworth, and Thompson 2013). This method operates under the assumption that all included variants are valid instruments, or that any horizontal pleiotropy is balanced, thus not violating the fundamental assumptions of MR.

We supplemented our primary analysis with several auxiliary methods to further assess causality. The MR‐Egger method (Bowden, Davey Smith, and Burgess 2015) uses weighted linear regression to address small study biases by analyzing the slope of outcome coefficients against exposure coefficients. It offers reliable causal estimates even when all genetic variants may be invalid instruments, albeit with less efficiency compared to IVW and median‐based methods. The weighted median approach integrates data from multiple genetic variants to produce a single causal estimate and tolerates up to 50% of the data originating from potentially invalid instruments.

SNPs were classified with mode‐based methods (Hartwig, Davey Smith, and Bowden 2017) into clusters by similarity in causal effects, estimating the causal effect based on the most populous clusters. This approach minimizes bias and reduces type‐I errors, particularly when the SNPs in the largest cluster are valid instruments.

To improve the trustworthiness of our findings, we undertook an extensive series of sensitivity checks. First, we evaluated the heterogeneity of SNP effects, which could bias the IVW estimates, by computing Cochran's Q statistics (Bowden et al. 2019). A p‐value above 0.05 for this test suggested minimal heterogeneity impact. Where heterogeneity was detected, we employed the random effects model of IVW to assess the causal effects more appropriately. Second, to address the risk of horizontal pleiotropy, which could violate the exclusion restriction assumption of Mendel's Third Law instrumental variables (IVs) should not influence the outcome except through the exposure (Didelez and Sheehan 2007; Angrist, Imbens, and Rubin 1996), we analyzed the MR‐Egger regression intercept.

An intercept not significantly differing from zero (p > 0.05) indicated an absence of horizontal pleiotropy (Burgess and Thompson 2017). Third, the MR Steiger directional test was utilized to verify the assumption that the exposure influences the outcome, rather than the reverse (Hemani, Tilling, and Smith 2017). This test compares the R 2 values for the outcome and exposure explained by the IVs. A significantly lower R 2 for the outcome than the exposure (MR Steiger test p < 0.05) suggested the absence of reverse causality between the two.

Lastly, we implemented a leave‐one‐out analysis to ascertain the robustness of the causal effect. This approach involves sequentially removing each IV and recalculating the IVW using the remaining IVs, thus assessing the influence of individual variants on the overall causal estimate. These methods collectively ensure our analysis is both thorough and reliable in detecting true causal relationships.

2.4

Multivariable MR Analysis

In this study, we acknowledged the potential for confounding among various vitamins and minerals when using two‐sample or standard MR due to the pleiotropic effects that one micronutrient may exert on others. To address this issue, we employed MVMR analysis. This approach integrates a pleiotropy correction by including all pertinent sleep behaviors in one model, which helps reduce bias (Sanderson 2021). For the MVMR analysis, we utilized all the variants from the two‐sample MR of micronutrients as IVs. This analysis was conducted using the “Mendelian Randomization” R package (version 0.10.0).

Within this framework, both the IVW and MR‐Egger methods were employed to determine the causal relationships between one sleep behavior and the other two.

To further assess heterogeneity in the outcomes of SNP, we calculated an adjusted Cochran's Q statistic using summary data (Sanderson et al. 2019), which helps identify variations in SNP effects that could impact the validity of our MR estimates. Additionally, the presence of pleiotropic effects was assessed by evaluating the MR‐Egger intercept term to determine whether the IVs had any effects on the outcome that were not mediated through the exposure. As our primary method, IVW was used under the assumption of no pleiotropic effects.

However, when directional pleiotropy was suspected, we employed the multivariable MR‐Egger method, which is specifically designed to accommodate such pleiotropy (Rees, Wood, and Burgess 2017). This comprehensive approach allows us to more accurately discern the direct and indirect effects of sleep behaviors on lifespan within a multivariable framework.

Source: Methods Source: Methods Source: Methods Source: Methods

Results: Two‐Sample MR Analysis

To explore the connections between micronutrients and various sleep behaviors, we utilized a two‐sample MR approach. We identified significant associations with chronotype for folate and vitamins B6 and D (p < 0.05) (Table S1; Figure 2). However, no significant correlations were observed between chronotype and the levels of calcium; carotene; copper; iron; magnesium; phosphorus; potassium; selenium; vitamins A, B12, C, and E; and zinc in the bloodstream (Table S1; Figure 2). Additionally, the MR analysis did not establish causal relationships between circulating micronutrients and the three sleep behaviors of short sleep duration, long sleep duration, and insomnia (Tables S2: S4).

FIGURE 2

Circos image of the associations of the 15 micronutrients with the chronotype from five statistic method. Red indicates a p value of less than 0.05.

Figure image at the original source: open the original paper. This file is not included in this text import.

Specifically, for each standard deviation (SD) increase in folate levels, the odds ratio (OR) is 1.09, with a 95% confidence interval (CI) ranging from 1.01 to 1.17 (p = 0.02). The MR‐Egger and weighted median approaches yielded congruent results, reinforcing the relationship. Specifically, the MR‐Egger method reported an OR of 1.22 with a 95% CI ranging from 1.03 to 1.44 and a p‐value of 0.04. The weighted median analysis showed an OR of 1.10 with a 95% CI of 1.10 to 1.19 and a p‐value of 0.03, as detailed in Table S1.

Additionally, vitamins B6 and D were negatively correlated with chronotype. For each standard deviation increase, the ORs were 0.91 with a 95% CI of 0.86 to 0.96 (p = 0.001) and 0.94 with a 95% CI of 0.88 to 1.00 (p = 0.03), respectively (Figures 2 and 3).

FIGURE 3

Two‐sample MR analysis on the association between folate, vitamin B6, and vitamin D levels and chronotype. Abbreviations: CI, confidence interval; IVW, inverse‐variance weighted; nsnp, number of SNP; OR, odds ratio.

Figure image at the original source: open the original paper. This file is not included in this text import.

MR results are presented in scatter plots with different color‐coded trend lines to showcase estimates calculated from the various methods, as seen in Figure 4. Additionally, forest plots depict the influence of each IV of folate, vitamin B6, and vitamin D levels on chronotype (Figure 5). To mitigate potential biases related to pleiotropy and the impact of individual IVs, additional analyses were performed. The Cochran's Q statistic revealed no significant heterogeneity among the IVs, with results for folate showing Q = 19.54, differences (df) = 21, p = 0.08; for vitamin B6, Q = 18.65, df = 15, p = 0.23; and for vitamin D, Q = 8.54, df = 11, p = 0.66.

Furthermore, no significant evidence of horizontal pleiotropy was found by the MR‐Egger regression, as demonstrated in Table S2. The robustness of these findings is further supported by the leave‐one‐out analysis, as shown in Figure 6, confirming that no single IV disproportionately influenced the overall causal assessment.

FIGURE 4

Scatter plot illustrating the two‐sample MR results for the effects of folate (A), vitamin B6 (B), and vitamin D (C) on chronotype. The slope of various colorful lines illustrates the estimated MR effect derived from different MR methods.

Figure image at the original source: open the original paper. This file is not included in this text import.
FIGURE 5

Forest plot of folate (A), vitamin B6 (B), and vitamin D (C) on chronotype. Each line represents the effect of an IV.

Figure image at the original source: open the original paper. This file is not included in this text import.
FIGURE 6

Forest plot of leave‐one‐out result of folate (A), vitamin B6 (B), and vitamin D (C). Each line represents the IVW estimate of the impact of short sleep duration on lifespan after excluding this specific SNP. The absence of any line crossing zero suggests that the result is robust.

Figure image at the original source: open the original paper. This file is not included in this text import.
Source: Two‐Sample MR Analysis Source: Two‐Sample MR Analysis Source: Two‐Sample MR Analysis Source: Two‐Sample MR Analysis Source: Two‐Sample MR Analysis Source: Two‐Sample MR Analysis Source: Two‐Sample MR Analysis Source: Two‐Sample MR Analysis

Results: Multivariable MR

Given the interrelationships among folate, vitamin B6, and vitamin D, the two‐sample MR analysis could not exclude potential confounding because of the other micronutrients. Thus, we conducted an MVMR analysis, estimating the direct effect of each exposure conditional on the others included in the model (Sanderson 2021). In this more comprehensive analysis, folate showed a positive causal effect on chronotype (OR = 1.286, 95% CI = 1.124 to 1.472, p < 0.001), while vitamin B6 demonstrated a negative causal effect (OR = 0.750, 95% CI = 0.648 to 0.868, p < 0.001).

The impact of vitamin D on chronotype remains unclear (Table S5; Figure 7).

FIGURE 7

MVMR results for sleep behaviors, conditioned on chronotype.

Figure image at the original source: open the original paper. This file is not included in this text import.
Source: Multivariable MR Source: Multivariable MR

Discussion

Through our two‐sample MR analyses, we identified associations between specific circulating micronutrients: folate, vitamin B6, and vitamin D: and chronotype. Folate exhibited a positive correlation with chronotype, suggesting that adults possessing elevated levels of folate tend to be morning‐oriented individuals (“early birds”). Conversely, higher concentrations of vitamins B6 and D were associated with a greater likelihood of being evening‐oriented (“night owls”). However, no causal relationships were established between these micronutrients and sleep disorders such as short sleep duration, long sleep duration, or insomnia.

Our MVMR analysis found a positive correlation between folate levels and chronotype, whereas vitamin B6 was negatively correlated with chronotype. The impact of vitamin D on chronotype remains unclear. Further analysis confirmed that individuals with higher levels of folate are more likely to exhibit an early‐to‐bed, early‐to‐rise pattern, while those with elevated levels of vitamin B6 tend to have a late‐to‐bed, late‐to‐rise sleep pattern.

Sleep habits are interconnected with the composition and timing of food intake. Morning chronotypes tend to have higher intakes of energy, protein, and fats, with lower carbohydrate consumption (Günal 2023). Posterior results show that a healthier dietary structure, characterized by closer alignment with the Mediterranean diet and reduced consumption of sugary drinks and juices, correlates with a decreased incidence of poor sleep (Huang et al. 2024). Although meta‐analyses suggest that no specific foods explicitly impact sleep, the timing of food intake does affect the chronotype (Netzer, Strohl, and Pramsohler 2024).

The results demonstrate that eating dinner after 10 p.m. leads to diminished lipase activity. Furthermore, the sleep: wake cycle influences the effect of lipolytic enzymes on metabolism (Arredondo‐Amador et al. 2020).

Extensive research highlights the protective effects of vitamin D on sleep quality, though its influence on chronotype is not well‐defined (Ji, Grandner, and Liu 2017). For instance, a study of postmenopausal women indicated that those with higher dietary intake of vitamin D experienced a delay in sleep phase onset (Grandner et al. 2010). Conversely, research in Japan involving young women aged 18 to 20 years revealed a negative correlation between the midpoint of sleep and the intake of several nutrients, including potassium, calcium, magnesium, iron, zinc, thiamine, riboflavin, and vitamins A, D, and B (Sato‐Mito et al. 2011).

The findings suggest that higher physiological levels of folate and vitamins D and B6 are linked to earlier sleep initiation and wake times (Jones et al. 2019).

Consistent with other studies, our research did not identify a significant association between sleep duration and the levels of folate, vitamin B6, or zinc (Ikonte et al. 2019). Nevertheless, these micronutrients, which are known to enhance melatonin synthesis, contribute to regulating sleep and circadian rhythms (Peuhkuri, Sihvola, and Korpela 2012). Iron and zinc are particularly crucial in enhancing the sleep quality of children and adolescents, which can significantly impact cognitive development due to issues related to sleep quality (Ji et al. 2021; Liu et al. 2024).

A study conducted in Turkey revealed that individuals with an evening‐type (E‐type) chronotype consume more vitamin B6 compared to those with a morning‐type (M‐type) chronotype (Toktaş, Erman, and Mert 2018). This increased intake may lead to a compensatory accumulation of vitamin B6 in the body, resulting in higher concentrations in the bloodstream.

It is crucial to highlight some limitations of our study. In the first place, our findings, derived from a population of European descent, may not be generalizable to other racial groups. Second, the classification of short sleep duration in our study was based on self‐reports rather than objective measurements, which could potentially introduce bias into the GWAS results. With the increasing prevalence of smart wearable devices, future research could potentially achieve more precise measurements of sleep conditions.

Source: Discussion Source: Discussion Source: Discussion Source: Discussion Source: Discussion Source: Discussion

: Peer Review

The peer review history for this article is available at https://publons.com/publon/10.1002/brb3.70237.

Source: Peer Review

Results

3.1

Two‐Sample MR Analysis

To explore the connections between micronutrients and various sleep behaviors, we utilized a two‐sample MR approach. We identified significant associations with chronotype for folate and vitamins B6 and D (p < 0.05) (Table S1; Figure 2). However, no significant correlations were observed between chronotype and the levels of calcium; carotene; copper; iron; magnesium; phosphorus; potassium; selenium; vitamins A, B12, C, and E; and zinc in the bloodstream (Table S1; Figure 2). Additionally, the MR analysis did not establish causal relationships between circulating micronutrients and the three sleep behaviors of short sleep duration, long sleep duration, and insomnia (Tables S2: S4).

FIGURE 2

Circos image of the associations of the 15 micronutrients with the chronotype from five statistic method. Red indicates a p value of less than 0.05.

Figure image at the original source: open the original paper. This file is not included in this text import.

Specifically, for each standard deviation (SD) increase in folate levels, the odds ratio (OR) is 1.09, with a 95% confidence interval (CI) ranging from 1.01 to 1.17 (p = 0.02). The MR‐Egger and weighted median approaches yielded congruent results, reinforcing the relationship. Specifically, the MR‐Egger method reported an OR of 1.22 with a 95% CI ranging from 1.03 to 1.44 and a p‐value of 0.04. The weighted median analysis showed an OR of 1.10 with a 95% CI of 1.10 to 1.19 and a p‐value of 0.03, as detailed in Table S1.

Additionally, vitamins B6 and D were negatively correlated with chronotype. For each standard deviation increase, the ORs were 0.91 with a 95% CI of 0.86 to 0.96 (p = 0.001) and 0.94 with a 95% CI of 0.88 to 1.00 (p = 0.03), respectively (Figures 2 and 3).

FIGURE 3

Two‐sample MR analysis on the association between folate, vitamin B6, and vitamin D levels and chronotype. Abbreviations: CI, confidence interval; IVW, inverse‐variance weighted; nsnp, number of SNP; OR, odds ratio.

Figure image at the original source: open the original paper. This file is not included in this text import.

MR results are presented in scatter plots with different color‐coded trend lines to showcase estimates calculated from the various methods, as seen in Figure 4. Additionally, forest plots depict the influence of each IV of folate, vitamin B6, and vitamin D levels on chronotype (Figure 5). To mitigate potential biases related to pleiotropy and the impact of individual IVs, additional analyses were performed. The Cochran's Q statistic revealed no significant heterogeneity among the IVs, with results for folate showing Q = 19.54, differences (df) = 21, p = 0.08; for vitamin B6, Q = 18.65, df = 15, p = 0.23; and for vitamin D, Q = 8.54, df = 11, p = 0.66.

Furthermore, no significant evidence of horizontal pleiotropy was found by the MR‐Egger regression, as demonstrated in Table S2. The robustness of these findings is further supported by the leave‐one‐out analysis, as shown in Figure 6, confirming that no single IV disproportionately influenced the overall causal assessment.

FIGURE 4

Scatter plot illustrating the two‐sample MR results for the effects of folate (A), vitamin B6 (B), and vitamin D (C) on chronotype. The slope of various colorful lines illustrates the estimated MR effect derived from different MR methods.

Figure image at the original source: open the original paper. This file is not included in this text import.
FIGURE 5

Forest plot of folate (A), vitamin B6 (B), and vitamin D (C) on chronotype. Each line represents the effect of an IV.

Figure image at the original source: open the original paper. This file is not included in this text import.
FIGURE 6

Forest plot of leave‐one‐out result of folate (A), vitamin B6 (B), and vitamin D (C). Each line represents the IVW estimate of the impact of short sleep duration on lifespan after excluding this specific SNP. The absence of any line crossing zero suggests that the result is robust.

Figure image at the original source: open the original paper. This file is not included in this text import.
3.2

Multivariable MR

Given the interrelationships among folate, vitamin B6, and vitamin D, the two‐sample MR analysis could not exclude potential confounding because of the other micronutrients. Thus, we conducted an MVMR analysis, estimating the direct effect of each exposure conditional on the others included in the model (Sanderson 2021). In this more comprehensive analysis, folate showed a positive causal effect on chronotype (OR = 1.286, 95% CI = 1.124 to 1.472, p < 0.001), while vitamin B6 demonstrated a negative causal effect (OR = 0.750, 95% CI = 0.648 to 0.868, p < 0.001).

The impact of vitamin D on chronotype remains unclear (Table S5; Figure 7).

FIGURE 7

MVMR results for sleep behaviors, conditioned on chronotype.

Figure image at the original source: open the original paper. This file is not included in this text import.
Source: Results Source: Results

Conclusion

Our two‐sample Mendelian randomization analysis suggests that individuals with higher folate levels tend to be morning‐oriented, which may be beneficial for enhancing sleep quality. Conversely, those with elevated levels of vitamin B6 and vitamin D tend to be evening‐oriented. Future research should include diverse populations and also consider conducting controlled, randomized trials to precisely determine micronutrient intake levels, providing a robust framework to validate the conclusions derived from Mendelian randomization.

Source: Conclusion
Authors’ abstract

ABSTRACT

Background

Sleep behaviors, defined by the total duration of sleep and chronotype, significantly influence overall health. Compromised sleep quality, which is often manifested through reduced sleep duration and the prevalence of insomnia, has been found to be associated with micronutrient deficiencies. Nonetheless, the existence of a causal relationship between micronutrient levels and sleep behaviors remains to be established.

Methods

A two‐sample Mendelian randomization (MR) analysis, utilizing data from genome‐wide association studies (GWAS), was employed to examine the associations between 15 micronutrients (copper; calcium; carotene; folate; iron; magnesium; potassium; selenium; vitamins A, B12, B6, C, D, and E; and zinc) and various sleep behaviors, including short and long sleep durations, insomnia, and chronotype. Furthermore, multivariable MR (MVMR) analysis was performed to address potential confounding due to the interrelationships among micronutrients and to discern potential causal relationships.

Results

The MR analysis identified a causal association between folate levels and chronotype (odds ratio [OR] = 1.09; 95% confidence interval [CI]: 1.01 to 1.17; p = 0.02), indicating a tendency toward morningness. Conversely, vitamin B6 (OR = 0.91; 95% CI: 0.86 to 0.96; p = 1.05 × 10−3) and vitamin D (OR = 0.94; 95% CI: 0.88 to 1.00; p = 0.03) showed inverse associations with chronotype, indicative of a preference for eveningness. MVMR analysis confirmed the positive association between folate (OR = 1.286, 95% CI = 1.124 to 1.472, p < 0.001) and chronotype and a negative association with vitamin B6 (OR = 0.750, 95% CI = 0.648 to 0.868, p < 0.001).

No causal relationships were established between micronutrient levels and either sleep duration or insomnia.

Conclusions

Elevated folate levels correlate with morning‐type preferences (“morning birds”), while higher concentrations of vitamin B6 are associated with evening‐type preferences (“evening owls”).

Read the full paper
1

Introduction

Sleep patterns are defined by their length, often calculated as the total amount of sleep within a 24‐h cycle, and by chronotype. Chronotype refers to an individual's natural predisposition to be active at particular times during a 24‐h cycle (Günal 2023). “Morning types,” commonly known as “larks,” tend to be more energetic and productive in the early morning and typically prefer to go to bed and wake up early. “Evening types,” often referred to as “owls,” are more active and focused in the evening, usually preferring to go to bed and rise later (Günal 2023; Wu et al. 2024).

Numerous factors, including age, gender, and ethnicity, influence sleep behavior (Hirshkowitz et al. 2015; Grandner 2019; Walch, Cochran, and Forger 2016). Notably, the duration and quality of sleep are significantly determined by genetic factors. Meta‐analyses have demonstrated that approximately 46% of the variance in sleep duration among individuals is attributable to genetic differences, and 44% of the variance in sleep quality is determined by genetics (Breitenstein, Doane, and Lemery‐Chalfant 2021). This finding is corroborated by genetic studies that employed genome‐wide association studies (GWAS) to analyze self‐reported sleep duration data from the UK Biobank.

The results were validated for consistency in separate research involving both adults and children/adolescents (Kocevska et al. 2021; Dashti et al. 2019).

The impact of micronutrient intake on sleep has received less attention than that of macronutrients. However, emerging studies suggest that shorter sleep durations in adults are associated with deficiencies in several micronutrients, specifically calcium, magnesium, and vitamins D and K (Ikonte et al. 2019; Al Hinai et al. 2024). Furthermore, numerous clinical trials have provided evidence supporting the assertion that adequate concentrations of zinc and iron can enhance sleep quality in infants and adolescents (Innocenti et al. 2023; Ji and Liu 2015; Kordas et al. 2009).

Traditional observational studies may be influenced by various confounding factors, which could introduce biases. To date, there has not been a comprehensive review of the literature examining the connection between micronutrients and sleep behaviors from a developmental standpoint.

Mendelian randomization (MR) utilizes single nucleotide polymorphism (SNP) statistics from GWAS to infer possible causal connections among various complex traits (Gupta, Walia, and Sachdeva 2017; Yao et al. 2022). SNPs, due to their random assignment during meiosis and fertilization prior to birth, are less susceptible to alterations by disease conditions, environmental factors, or other confounders (Davies, Holmes, and Davey Smith 2018). Therefore, they may offer a more precise understanding of true causal relationships compared to conventional observational studies.

In our research, we investigated the genetic underpinnings and potential causal links between micronutrients and sleep behaviors. This was achieved by examining genetic correlations and polygenic overlaps using GWAS summary statistics. Additionally, we employed two‐sample and multivariable MR (MVMR) analyses to explore the possible causal effects of micronutrients on sleep behaviors.

2

Methods

2.1

Study Design

A schematic overview of the research design is depicted in Figure 1. We conducted an MR investigation utilizing data from 18 publicly accessible GWAS to acquire summary statistics. In the two‐sample MR analysis, data concerning sleep behaviors were sourced from three independent GWAS consortia. These data were employed for initial analyses and replication studies, and ultimately integrated into a meta‐analysis to consolidate findings. For the MVMR analysis, we investigated the causal relationships between various micronutrients and sleep behaviors. To increase the statistical robustness of our findings, we combined estimates from various data sources.

All studies involved in the GWAS had previously been approved by the appropriate review boards, and no further ethical approval or participant consent was necessary for this analysis.

FIGURE 1

A flow chart of the study design.

Figure image at the original source: open the original paper. This file is not included in this text import.
2.2

GWAS Data of Sleep‐Related Behaviors

Participants recorded their own sleep duration by noting the total hours slept every 24 h, including naps, measured in hourly increments. We carried out distinct GWAS for individuals of European descent, categorizing them based on their sleep duration. This included a group with short sleep duration (less than 7 h per night, n = 106,192 cases) and another with long sleep duration (more than 9 h per night, n = 34,184 cases). These were compared to a control group that reported sleeping between 7 and 9 h per night (n = 305,742) (Dashti et al. 2019).

We excluded responses that reported extremely low (less than 3 h) or extremely high (more than 18 h) sleep durations. Responses marked as “Do not know” or “Prefer not to answer” were considered missing data. We can obtain the GWAS summary statistics from the Sleep Disorder Knowledge Portal at https://sleep.hugeamp.org/downloads.html.

To assess chronotype, subjects responded to the question, “Do you consider yourself to be?” with choices spanning from “Definitely a ‘morning’ person” to “Definitely an ‘evening’ person.” Responses were scored, with “Definitely a ‘morning’ person” and “More a ‘morning’ than ‘evening’ person” classified as cases, assigned two and one points, respectively. Conversely, those identifying as “Definitely an ‘evening’ person” or “More an ‘evening’ than ‘morning’ person” were designated as controls, with scoring of −2 and −1 points, respectively. The study by Jones et al. (2019) encompassed 252,287 cases and 150,908 controls of European ancestry.

The GWAS summary statistics for this analysis are also available on the Sleep Disorder Knowledge Portal.

In the assessment of insomnia, participants answered the question: “Do you have trouble falling asleep at night or do you wake up in the middle of the night?” The response options provided were “never/rarely,” “sometimes,” “usually,” or “prefer not to answer.” Participants who chose “usually” were identified as cases of insomnia, while those who answered “never/rarely” were designated as controls. The GWAS, carried out by Watanabe et al. (2022), involved 593,724 cases and 1,771,286 controls of European descent and identified 554 loci associated with insomnia.

The GWAS summary statistics are available from https://ctg.cncr.nl/software/summary_statistics/.

In our research, we conducted a comprehensive search through PubMed and the IEU OpenGWAS project (https://gwas.mrcieu.ac.uk/) to identify the most recent large‐scale GWAS studies on micronutrients, with the final search conducted in April 2024. To avoid overlap between the exposures and outcomes in our study, we excluded certain micronutrients that were obtained from these specific databases during our data collection. We observed a notable absence of genomic studies concerning chloride; fluoride; phosphorus; sulfur; vitamins B1, B2, B3, B5, and B7; and iodine.

Furthermore, studies on vitamin K, cobalt, chromium, sodium, and molybdenum were not included due to their inconclusive genomic results (Dashti et al. 2014; Ng et al. 2015). This selection process led us to preliminarily identify 15 micronutrients for further investigation: copper (Evans et al. 2013), calcium (O'Seaghdha et al. 2013), carotene (Ferrucci et al. 2009), folate (Grarup et al. 2013), iron (Bell et al. 2021), magnesium (Meyer et al. 2010), potassium (Meyer et al. 2010), selenium (Evans et al. 2013), vitamin A (Mondul et al. 2011), vitamin B12 (Grarup et al. 2013), vitamin B6 (Hazra et al. 2009), vitamin C (Zheng et al. 2020), vitamin D (Revez et al. 2020), vitamin E (Major et al. 2011), and zinc (Evans et al. 2013).

2.3

Two‐Sample MR Analysis

In our research, we utilized the two‐sample MR approach, employing the Two Sample MR (version 0.5.11) package (Hemani, Tilling, and Smith 2017; Hemani et al. 2018). Our principal technique for MR analysis was the inverse variance weighting (IVW) method. This meta‐analytic approach combines Wald estimates from each SNP to compute overall effect estimates of the exposure on the outcome (Burgess, Butterworth, and Thompson 2013). This method operates under the assumption that all included variants are valid instruments, or that any horizontal pleiotropy is balanced, thus not violating the fundamental assumptions of MR.

We supplemented our primary analysis with several auxiliary methods to further assess causality. The MR‐Egger method (Bowden, Davey Smith, and Burgess 2015) uses weighted linear regression to address small study biases by analyzing the slope of outcome coefficients against exposure coefficients. It offers reliable causal estimates even when all genetic variants may be invalid instruments, albeit with less efficiency compared to IVW and median‐based methods. The weighted median approach integrates data from multiple genetic variants to produce a single causal estimate and tolerates up to 50% of the data originating from potentially invalid instruments.

SNPs were classified with mode‐based methods (Hartwig, Davey Smith, and Bowden 2017) into clusters by similarity in causal effects, estimating the causal effect based on the most populous clusters. This approach minimizes bias and reduces type‐I errors, particularly when the SNPs in the largest cluster are valid instruments.

To improve the trustworthiness of our findings, we undertook an extensive series of sensitivity checks. First, we evaluated the heterogeneity of SNP effects, which could bias the IVW estimates, by computing Cochran's Q statistics (Bowden et al. 2019). A p‐value above 0.05 for this test suggested minimal heterogeneity impact. Where heterogeneity was detected, we employed the random effects model of IVW to assess the causal effects more appropriately. Second, to address the risk of horizontal pleiotropy, which could violate the exclusion restriction assumption of Mendel's Third Law instrumental variables (IVs) should not influence the outcome except through the exposure (Didelez and Sheehan 2007; Angrist, Imbens, and Rubin 1996), we analyzed the MR‐Egger regression intercept.

An intercept not significantly differing from zero (p > 0.05) indicated an absence of horizontal pleiotropy (Burgess and Thompson 2017). Third, the MR Steiger directional test was utilized to verify the assumption that the exposure influences the outcome, rather than the reverse (Hemani, Tilling, and Smith 2017). This test compares the R 2 values for the outcome and exposure explained by the IVs. A significantly lower R 2 for the outcome than the exposure (MR Steiger test p < 0.05) suggested the absence of reverse causality between the two.

Lastly, we implemented a leave‐one‐out analysis to ascertain the robustness of the causal effect. This approach involves sequentially removing each IV and recalculating the IVW using the remaining IVs, thus assessing the influence of individual variants on the overall causal estimate. These methods collectively ensure our analysis is both thorough and reliable in detecting true causal relationships.

2.4

Multivariable MR Analysis

In this study, we acknowledged the potential for confounding among various vitamins and minerals when using two‐sample or standard MR due to the pleiotropic effects that one micronutrient may exert on others. To address this issue, we employed MVMR analysis. This approach integrates a pleiotropy correction by including all pertinent sleep behaviors in one model, which helps reduce bias (Sanderson 2021). For the MVMR analysis, we utilized all the variants from the two‐sample MR of micronutrients as IVs. This analysis was conducted using the “Mendelian Randomization” R package (version 0.10.0).

Within this framework, both the IVW and MR‐Egger methods were employed to determine the causal relationships between one sleep behavior and the other two.

To further assess heterogeneity in the outcomes of SNP, we calculated an adjusted Cochran's Q statistic using summary data (Sanderson et al. 2019), which helps identify variations in SNP effects that could impact the validity of our MR estimates. Additionally, the presence of pleiotropic effects was assessed by evaluating the MR‐Egger intercept term to determine whether the IVs had any effects on the outcome that were not mediated through the exposure. As our primary method, IVW was used under the assumption of no pleiotropic effects.

However, when directional pleiotropy was suspected, we employed the multivariable MR‐Egger method, which is specifically designed to accommodate such pleiotropy (Rees, Wood, and Burgess 2017). This comprehensive approach allows us to more accurately discern the direct and indirect effects of sleep behaviors on lifespan within a multivariable framework.

3

Results

3.1

Two‐Sample MR Analysis

To explore the connections between micronutrients and various sleep behaviors, we utilized a two‐sample MR approach. We identified significant associations with chronotype for folate and vitamins B6 and D (p < 0.05) (Table S1; Figure 2). However, no significant correlations were observed between chronotype and the levels of calcium; carotene; copper; iron; magnesium; phosphorus; potassium; selenium; vitamins A, B12, C, and E; and zinc in the bloodstream (Table S1; Figure 2). Additionally, the MR analysis did not establish causal relationships between circulating micronutrients and the three sleep behaviors of short sleep duration, long sleep duration, and insomnia (Tables S2: S4).

FIGURE 2

Circos image of the associations of the 15 micronutrients with the chronotype from five statistic method. Red indicates a p value of less than 0.05.

Figure image at the original source: open the original paper. This file is not included in this text import.

Specifically, for each standard deviation (SD) increase in folate levels, the odds ratio (OR) is 1.09, with a 95% confidence interval (CI) ranging from 1.01 to 1.17 (p = 0.02). The MR‐Egger and weighted median approaches yielded congruent results, reinforcing the relationship. Specifically, the MR‐Egger method reported an OR of 1.22 with a 95% CI ranging from 1.03 to 1.44 and a p‐value of 0.04. The weighted median analysis showed an OR of 1.10 with a 95% CI of 1.10 to 1.19 and a p‐value of 0.03, as detailed in Table S1.

Additionally, vitamins B6 and D were negatively correlated with chronotype. For each standard deviation increase, the ORs were 0.91 with a 95% CI of 0.86 to 0.96 (p = 0.001) and 0.94 with a 95% CI of 0.88 to 1.00 (p = 0.03), respectively (Figures 2 and 3).

FIGURE 3

Two‐sample MR analysis on the association between folate, vitamin B6, and vitamin D levels and chronotype. Abbreviations: CI, confidence interval; IVW, inverse‐variance weighted; nsnp, number of SNP; OR, odds ratio.

Figure image at the original source: open the original paper. This file is not included in this text import.

MR results are presented in scatter plots with different color‐coded trend lines to showcase estimates calculated from the various methods, as seen in Figure 4. Additionally, forest plots depict the influence of each IV of folate, vitamin B6, and vitamin D levels on chronotype (Figure 5). To mitigate potential biases related to pleiotropy and the impact of individual IVs, additional analyses were performed. The Cochran's Q statistic revealed no significant heterogeneity among the IVs, with results for folate showing Q = 19.54, differences (df) = 21, p = 0.08; for vitamin B6, Q = 18.65, df = 15, p = 0.23; and for vitamin D, Q = 8.54, df = 11, p = 0.66.

Furthermore, no significant evidence of horizontal pleiotropy was found by the MR‐Egger regression, as demonstrated in Table S2. The robustness of these findings is further supported by the leave‐one‐out analysis, as shown in Figure 6, confirming that no single IV disproportionately influenced the overall causal assessment.

FIGURE 4

Scatter plot illustrating the two‐sample MR results for the effects of folate (A), vitamin B6 (B), and vitamin D (C) on chronotype. The slope of various colorful lines illustrates the estimated MR effect derived from different MR methods.

Figure image at the original source: open the original paper. This file is not included in this text import.
FIGURE 5

Forest plot of folate (A), vitamin B6 (B), and vitamin D (C) on chronotype. Each line represents the effect of an IV.

Figure image at the original source: open the original paper. This file is not included in this text import.
FIGURE 6

Forest plot of leave‐one‐out result of folate (A), vitamin B6 (B), and vitamin D (C). Each line represents the IVW estimate of the impact of short sleep duration on lifespan after excluding this specific SNP. The absence of any line crossing zero suggests that the result is robust.

Figure image at the original source: open the original paper. This file is not included in this text import.
3.2

Multivariable MR

Given the interrelationships among folate, vitamin B6, and vitamin D, the two‐sample MR analysis could not exclude potential confounding because of the other micronutrients. Thus, we conducted an MVMR analysis, estimating the direct effect of each exposure conditional on the others included in the model (Sanderson 2021). In this more comprehensive analysis, folate showed a positive causal effect on chronotype (OR = 1.286, 95% CI = 1.124 to 1.472, p < 0.001), while vitamin B6 demonstrated a negative causal effect (OR = 0.750, 95% CI = 0.648 to 0.868, p < 0.001).

The impact of vitamin D on chronotype remains unclear (Table S5; Figure 7).

FIGURE 7

MVMR results for sleep behaviors, conditioned on chronotype.

Figure image at the original source: open the original paper. This file is not included in this text import.
4

Discussion

Through our two‐sample MR analyses, we identified associations between specific circulating micronutrients: folate, vitamin B6, and vitamin D: and chronotype. Folate exhibited a positive correlation with chronotype, suggesting that adults possessing elevated levels of folate tend to be morning‐oriented individuals (“early birds”). Conversely, higher concentrations of vitamins B6 and D were associated with a greater likelihood of being evening‐oriented (“night owls”). However, no causal relationships were established between these micronutrients and sleep disorders such as short sleep duration, long sleep duration, or insomnia.

Our MVMR analysis found a positive correlation between folate levels and chronotype, whereas vitamin B6 was negatively correlated with chronotype. The impact of vitamin D on chronotype remains unclear. Further analysis confirmed that individuals with higher levels of folate are more likely to exhibit an early‐to‐bed, early‐to‐rise pattern, while those with elevated levels of vitamin B6 tend to have a late‐to‐bed, late‐to‐rise sleep pattern.

Sleep habits are interconnected with the composition and timing of food intake. Morning chronotypes tend to have higher intakes of energy, protein, and fats, with lower carbohydrate consumption (Günal 2023). Posterior results show that a healthier dietary structure, characterized by closer alignment with the Mediterranean diet and reduced consumption of sugary drinks and juices, correlates with a decreased incidence of poor sleep (Huang et al. 2024). Although meta‐analyses suggest that no specific foods explicitly impact sleep, the timing of food intake does affect the chronotype (Netzer, Strohl, and Pramsohler 2024).

The results demonstrate that eating dinner after 10 p.m. leads to diminished lipase activity. Furthermore, the sleep: wake cycle influences the effect of lipolytic enzymes on metabolism (Arredondo‐Amador et al. 2020).

Extensive research highlights the protective effects of vitamin D on sleep quality, though its influence on chronotype is not well‐defined (Ji, Grandner, and Liu 2017). For instance, a study of postmenopausal women indicated that those with higher dietary intake of vitamin D experienced a delay in sleep phase onset (Grandner et al. 2010). Conversely, research in Japan involving young women aged 18 to 20 years revealed a negative correlation between the midpoint of sleep and the intake of several nutrients, including potassium, calcium, magnesium, iron, zinc, thiamine, riboflavin, and vitamins A, D, and B (Sato‐Mito et al. 2011).

The findings suggest that higher physiological levels of folate and vitamins D and B6 are linked to earlier sleep initiation and wake times (Jones et al. 2019).

Consistent with other studies, our research did not identify a significant association between sleep duration and the levels of folate, vitamin B6, or zinc (Ikonte et al. 2019). Nevertheless, these micronutrients, which are known to enhance melatonin synthesis, contribute to regulating sleep and circadian rhythms (Peuhkuri, Sihvola, and Korpela 2012). Iron and zinc are particularly crucial in enhancing the sleep quality of children and adolescents, which can significantly impact cognitive development due to issues related to sleep quality (Ji et al. 2021; Liu et al. 2024).

A study conducted in Turkey revealed that individuals with an evening‐type (E‐type) chronotype consume more vitamin B6 compared to those with a morning‐type (M‐type) chronotype (Toktaş, Erman, and Mert 2018). This increased intake may lead to a compensatory accumulation of vitamin B6 in the body, resulting in higher concentrations in the bloodstream.

It is crucial to highlight some limitations of our study. In the first place, our findings, derived from a population of European descent, may not be generalizable to other racial groups. Second, the classification of short sleep duration in our study was based on self‐reports rather than objective measurements, which could potentially introduce bias into the GWAS results. With the increasing prevalence of smart wearable devices, future research could potentially achieve more precise measurements of sleep conditions.

5

Conclusion

Our two‐sample Mendelian randomization analysis suggests that individuals with higher folate levels tend to be morning‐oriented, which may be beneficial for enhancing sleep quality. Conversely, those with elevated levels of vitamin B6 and vitamin D tend to be evening‐oriented. Future research should include diverse populations and also consider conducting controlled, randomized trials to precisely determine micronutrient intake levels, providing a robust framework to validate the conclusions derived from Mendelian randomization.

Author Contributions

Ruijie Zhang: writing: original draft, methodology, investigation. Liyan Luo: supervision. Lu Zhang: writing: review and editing, methodology, investigation. Xinao Lin: writing: review and editing, software. Chuyan Wu: supervision. Feng Jiang: conceptualization, methodology, writing: review and editing. Jimei Wang: funding acquisition, supervision, project administration.

Ethics Statement

All the GWAS data utilized in this research were sourced from publicly accessible databases, and no original data were collected for this study. Each of the studies included had received approval from their respective institutional ethics review committees. Additionally, informed consents, both for participation and publication, were obtained from all participants involved.

Consent

All authors consent to publication.

Conflicts of Interest

The authors declare no conflicts of interest.

Peer Review

The peer review history for this article is available at https://publons.com/publon/10.1002/brb3.70237.

Supporting information

Supplement file at the original source: open the original paper. This file is not included in this text import.

Supplementary Materials.

Supplement file at the original source: open the original paper. This file is not included in this text import.

Supplementary Materials.

Supplement file at the original source: open the original paper. This file is not included in this text import.

Supplementary Materials.

Supplement file at the original source: open the original paper. This file is not included in this text import.

Supplementary Materials.

Supplement file at the original source: open the original paper. This file is not included in this text import.

Supplementary Materials.

Supplement file at the original source: open the original paper. This file is not included in this text import.

Acknowledgments

The authors have nothing to report.

Data Availability Statement

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

References

Al Hinai , M. , E. C. Jansen , P. X. Song , K. E. Peterson , and A Baylin . 2024 . “Iron Deficiency and Vitamin D Deficiency Are Associated With Sleep in Females of Reproductive Age: An Analysis of NHANES 2005‐2018 Data .” Journal of Nutrition 154 , no. 2 : 648Not reported657 . 38042351 10.1016/j.tjnut.2023.11.030 PMC10997906
Angrist , J. D. , G. W. Imbens , and D. B Rubin . 1996 . “Identification of Causal Effects Using Instrumental Variables .” Journal of the American Statistical Association 91 , no. 434 : 444Not reported455 .
Arredondo‐Amador , M. , C. Zambrano , A. Kulyté , et al. 2020 . “Circadian Rhythms in Hormone‐sensitive Lipase in Human Adipose Tissue: Relationship to Meal Timing and Fasting Duration .” Journal of Clinical Endocrinology and Metabolism 105 , no. 12 : e4407Not reportede4416 . 32725188 10.1210/clinem/dgaa492 PMC7538104
Bell , S. , A. S. Rigas , M. K. Magnusson , et al. 2021 . “A Genome‐Wide Meta‐Analysis Yields 46 New Loci Associating With Biomarkers of Iron Homeostasis .” Communications Biology 4 , no. 1 : 156 . 33536631 10.1038/s42003-020-01575-z PMC7859200
Bowden , J. , G. Davey Smith , and S Burgess . 2015 . “Mendelian Randomization With Invalid Instruments: Effect Estimation and Bias Detection Through Egger Regression .” International Journal of Epidemiology 44 , no. 2 : 512Not reported525 . 26050253 10.1093/ije/dyv080 PMC4469799
Bowden , J. , F. Del Greco M , C. Minelli , et al. 2019 . “Improving the Accuracy of Two‐Sample Summary Data Mendelian Randomization: Moving Beyond the NOME Assumption .” International Journal of Epidemiology 48 , no. 3 : 728Not reported742 . 10.1093/ije/dyy258 . 30561657 PMC6659376
Breitenstein , R. S. , L. D. Doane , and K Lemery‐Chalfant . 2021 . “Children's Objective Sleep Assessed With Wrist‐Based Accelerometers: Strong Heritability of Objective Quantity and Quality Unique From Parent‐Reported Sleep .” Sleep 44 , no. 1 : zsaa142. 32729904 10.1093/sleep/zsaa142 PMC7819846
Burgess , S. , A. Butterworth ., and S. G Thompson . 2013 . “Mendelian Randomization Analysis With Multiple Genetic Variants Using Summarized Data .” Genetic Epidemiology 37 , no. 7 : 658Not reported665 . 24114802 10.1002/gepi.21758 PMC4377079
Burgess , S. , and S. G Thompson . 2017 . “Interpreting Findings From Mendelian Randomization Using the MR‐Egger Method .” European Journal of Epidemiology 32 , no. 5 : 377Not reported389 . 28527048 10.1007/s10654-017-0255-x PMC5506233
Dashti , H. S. , S. E. Jones , A. R. Wood , et al. 2019 . “Genome‐Wide Association Study Identifies Genetic Loci for Self‐Reported Habitual Sleep Duration Supported by Accelerometer‐Derived Estimates .” Nature Communications 10 , no. 1 : 1100 . 10.1038/s41467-019-08917-4 PMC6405943 30846698
Dashti , H. S. , M. K. Shea , C. E. Smith , et al. 2014 . “Meta‐Analysis of Genome‐Wide Association Studies for Circulating Phylloquinone Concentrations .” American Journal of Clinical Nutrition 100 , no. 6 : 1462Not reported1469 . 25411281 10.3945/ajcn.114.093146 PMC4232014
Davies , N. M. , M. V. Holmes , and G Davey Smith . 2018 . “Reading Mendelian Randomisation Studies: A Guide, Glossary, and Checklist for Clinicians .” BMJ 362 : k601 . 30002074 10.1136/bmj.k601 PMC6041728
Didelez , V. , and N Sheehan . 2007 . “Mendelian Randomization as an Instrumental Variable Approach to Causal Inference .” Statistical Methods in Medical Research 16 , no. 4 : 309Not reported330 . 17715159 10.1177/0962280206077743
Evans , D. M. , G. Zhu , V. Dy , et al. 2013 . “Genome‐Wide Association Study Identifies Loci Affecting Blood Copper, Selenium and Zinc .” Human Molecular Genetics 22 , no. 19 : 3998Not reported4006 . 23720494 10.1093/hmg/ddt239 PMC3766178
Ferrucci , L. , J. R. Perry , A. Matteini , et al. 2009 . “Common Variation in the Beta‐carotene 15,15'‐monooxygenase 1 Gene Affects Circulating Levels of Carotenoids: A Genome‐Wide Association Study .” American Journal of Human Genetics 84 , no. 2 : 123Not reported133 . 19185284 10.1016/j.ajhg.2008.12.019 PMC2668002
Grandner , M. A. 2019 . “Chapter 2 ‐ Epidemiology of Insufficient Sleep and Poor Sleep Quality.” In Sleep and Health , edited by M. A. Grandner , 11Not reported20 . Cambridge, MA: Academic Press.
Grandner , M. A. , D. F. Kripke , N. Naidoo , and R. D Langer . 2010 . “Relationships Among Dietary Nutrients and Subjective Sleep, Objective Sleep, and Napping in Women .” Sleep Medicine 11 , no. 2 : 180Not reported184 . 20005774 10.1016/j.sleep.2009.07.014 PMC2819566
Grarup , N. , P. Sulem , C. H. Sandholt , et al. 2013 . “Genetic Architecture of Vitamin B12 and Folate Levels Uncovered Applying Deeply Sequenced Large Datasets .” PLoS Genetics 9 , no. 6 : e1003530. 23754956 10.1371/journal.pgen.1003530 PMC3674994
Günal , A. M. 2023 . “Sleep, Activity, and Diet in Harmony: Unveiling the Relationships of Chronotype, Sleep Quality, Physical Activity, and Dietary Intake .” Frontiers in Nutrition 10 : 1301818. 38162523 10.3389/fnut.2023.1301818 PMC10757331
Gupta , V. , G. K. Walia , and M. P Sachdeva . 2017 . “‘Mendelian Randomization’: An Approach for Exploring Causal Relations in Epidemiology .” Public Health 145 : 113Not reported119 . 28359378 10.1016/j.puhe.2016.12.033
Hartwig , F. P. , G. Davey Smith , and J Bowden . 2017 . “Robust Inference in Summary Data Mendelian Randomization via the Zero Modal Pleiotropy Assumption .” International Journal of Epidemiology 46 , no. 6 : 1985Not reported1998 . 29040600 10.1093/ije/dyx102 PMC5837715
Hazra , A. , P. Kraft , R. Lazarus , et al. 2009 . “Genome‐Wide Significant Predictors of Metabolites in the One‐Carbon Metabolism Pathway .” Human Molecular Genetics 18 , no. 23 : 4677Not reported4687 . 19744961 10.1093/hmg/ddp428 PMC2773275
Hemani , G. , K., D. Tilling , and G Smith . 2017 . “Orienting the Causal Relationship Between Imprecisely Measured Traits Using GWAS Summary Data .” PLoS Genetics 13 , no. 11 : e1007081. 29149188 10.1371/journal.pgen.1007081 PMC5711033
Hemani , G. , J. Zheng , B. Elsworth , et al. 2018 . “The MR‐Base Platform Supports Systematic Causal Inference Across the Human Phenome .” Elife 7 : e34408. 29846171 10.7554/eLife.34408 PMC5976434
Hirshkowitz , M. , K. Whiton , S. M. Albert , et al. 2015 . “National Sleep Foundation's Sleep Time Duration Recommendations: Methodology and Results Summary .” Sleep Health 1 , no. 1 : 40Not reported43 . 29073412 10.1016/j.sleh.2014.12.010
Huang , L. , Y. Jiang , Z. Sun , et al. 2024 . “Healthier Dietary Patterns Are Associated With Better Sleep Quality Among Shanghai Suburban Adults: A Cross‐Sectional Study .” Nutrients 16 , no. 8 : 1165 . 38674856 10.3390/nu16081165 PMC11054136
Ikonte , C. J. , J. G. Mun , C. A. Reider , R. W. Grant , and S. H Mitmesser . 2019 . “Micronutrient Inadequacy in Short Sleep: Analysis of the NHANES 2005 to 2016 .” Nutrients 11 , no. 10 : 2335 . 31581561 10.3390/nu11102335 PMC6835726
Innocenti , A. , G. Lentini , S. Rapacchietta , et al. 2023 . “The Role of Supplements and Over‐the‐Counter Products to Improve Sleep in Children: A Systematic Review .” International Journal of Molecular Sciences 24 , no. 9 : 7821 . 37175525 10.3390/ijms24097821 PMC10178725
Ji , X. , C. W. Compher , S. Y. Irving , J. Kim , D. F. Dinges , and J Liu . 2021 . “Serum Micronutrient Status, Sleep Quality and Neurobehavioural Function Among Early Adolescents .” Public Health Nutrition 24 , no. 17 : 5815Not reported5825 . 34034852 10.1017/S1368980021002329 PMC10117421
Ji , X. , M. A. Grandner , and J Liu . 2017 . “The Relationship Between Micronutrient Status and Sleep Patterns: A Systematic Review .” Public Health Nutrition 20 , no. 4 : 687Not reported701 . 27702409 10.1017/S1368980016002603 PMC5675071
Ji , X. , and J Liu . 2015 . “Associations Between Blood Zinc Concentrations and Sleep Quality in Childhood: A Cohort Study .” Nutrients 7 , no. 7 : 5684Not reported5696 . 26184300 10.3390/nu7075247 PMC4517024
Jones , S. E. , J. M. Lane , A. R. Wood , et al. 2019 . “Genome‐Wide Association Analyses of Chronotype in 697,828 Individuals Provides Insights Into Circadian Rhythms .” Nature Communications 10 , no. 1 : 343 . 10.1038/s41467-018-08259-7 PMC6351539 30696823
Kocevska , D. , N. L. Barclay , W. M. Bramer , P. R. Gehrman , and E. J. W Van Someren . 2021 . “Heritability of Sleep Duration and Quality: A Systematic Review and Meta‐Analysis .” Sleep Medicine Reviews 59 : 101448. 33636423 10.1016/j.smrv.2021.101448
Kordas , K. , E. H. Siegel , D. K. Olney , et al. 2009 . “The Effects of Iron and/or Zinc Supplementation on Maternal Reports of Sleep in Infants From Nepal and Zanzibar .” Journal of Developmental and Behavioral Pediatrics 30 , no. 2 : 131Not reported139 . 19322104 10.1097/DBP.0b013e31819e6a48 PMC2771202
Liu , J. , X. Ji , E. Rovit , S. Pitt , and T Lipman . 2024 . “Childhood Sleep: Assessments, Risk Factors, and Potential Mechanisms .” World Journal of Pediatrics 20 , no. 2 : 105Not reported121 . 36441394 10.1007/s12519-022-00628-z PMC9702880
Major , J. M. , K. Yu , W. Wheeler , et al. 2011 . “Genome‐Wide Association Study Identifies Common Variants Associated With Circulating Vitamin E Levels .” Human Molecular Genetics 20 , no. 19 : 3876Not reported3883 . 21729881 10.1093/hmg/ddr296 PMC3168288
Meyer , T. E. , G. C. Verwoert , S.‐J. Hwang , et al. 2010 . “Genome‐Wide Association Studies of Serum Magnesium, Potassium, and Sodium Concentrations Identify Six Loci Influencing Serum Magnesium Levels .” PLOS Genetics 6 , no. 8 : e1001045. 20700443 10.1371/journal.pgen.1001045 PMC2916845
Mondul , A. M. , K. Yu , W. Wheeler , et al. 2011 . “Genome‐Wide Association Study of Circulating Retinol Levels .” Human Molecular Genetics 20 , no. 23 : 4724Not reported4731 . 21878437 10.1093/hmg/ddr387 PMC3209826
Netzer , N. C. , K. P. Strohl , and S. Pramsohler . 2024 . “Influence of Nutrition and Food on Sleep: Is There Evidence? ” Sleep Breath 28 , no. 1 : 61Not reported68 . 10.1007/s11325-023-02921-1 . 37740061 PMC10954981
Ng , E. , P. M. Lind , C. Lindgren , et al. 2015 . “Genome‐Wide Association Study of Toxic Metals and Trace Elements Reveals Novel Associations .” Human Molecular Genetics 24 , no. 16 : 4739Not reported4745 . 26025379 10.1093/hmg/ddv190 PMC4512629
O'Seaghdha , C. M. , H. Wu , Q. Yang , et al. 2013 . “Meta‐Analysis of Genome‐Wide Association Studies Identifies Six New Loci for Serum Calcium Concentrations .” PLos Genetics 9 , no. 9 : e1003796. 24068962 10.1371/journal.pgen.1003796 PMC3778004
Peuhkuri , K. , N. Sihvola , and R Korpela . 2012 . “Dietary Factors and Fluctuating Levels of Melatonin .” Food and Nutrition Research 56 , no. 1 : 17252. 10.3402/fnr.v56i0.17252 PMC3402070 22826693
Rees , J. M. B. , A. M. Wood , and S Burgess . 2017 . “Extending the MR‐Egger Method for Multivariable Mendelian Randomization to Correct for Both Measured and Unmeasured Pleiotropy .” Statistics in Medicine 36 , no. 29 : 4705Not reported4718 . 28960498 10.1002/sim.7492 PMC5725762
Revez , J. A. , T. Lin , Z. Qiao , et al. 2020 . “Genome‐wide association study identifies 143 loci associated with 25 hydroxyvitamin D concentration .” Nature Communications 1 , no. 11 . 10.1038/s41467-020-15421-7 . PMC7118120 32242144
Sanderson , E. 2021 . “Multivariable Mendelian Randomization and Mediation .” Cold Spring Harbor Perspectives in Medicine 11 , no. 2 : a038984. 32341063 10.1101/cshperspect.a038984 PMC7849347
Sanderson , E. , G. Davey Smith , F. Windmeijer , and J Bowden . 2019 . “An Examination of Multivariable Mendelian Randomization in the Single‐Sample and Two‐Sample Summary Data Settings .” International Journal of Epidemiology 48 , no. 3 : 713Not reported727 . 30535378 10.1093/ije/dyy262 PMC6734942
Sato‐Mito , N. , S. Sasaki , K. Murakami , et al. 2011 . “The Midpoint of Sleep Is Associated With Dietary Intake and Dietary Behavior Among Young Japanese Women .” Sleep Medicine 12 , no. 3 : 289Not reported294 . 21296614 10.1016/j.sleep.2010.09.012
Toktaş , N. , K. A. Erman , and Z Mert . 2018 . “Nutritional Habits According to Human Chronotype and Nutritional Status of Morningness and Eveningness .” Journal of Education and Training Studies 6 : 61Not reported67 .
Walch , O. J. , A. Cochran , and D. B Forger . 2016 . “A Global Quantification of “Normal” Sleep Schedules Using Smartphone Data .” Science Advances 2 , no. 5 : e1501705. 27386531 10.1126/sciadv.1501705 PMC4928979
Watanabe , K. , P. R. Jansen , J. E. Savage , et al. 2022 . “Genome‐Wide Meta‐Analysis of Insomnia Prioritizes Genes Associated With Metabolic and Psychiatric Pathways .” Nature Genetics 54 , no. 8 : 1125‐+. 35835914 10.1038/s41588-022-01124-w
Wu , Y. , C. Y. Zhang , X. Liu , et al. 2024 . “Shared Genetic Architecture and Causal Relationship Between Sleep Behaviors and Lifespan .” Translational Psychiatry 14 , no. 1 : 108 . 38388528 10.1038/s41398-024-02826-x PMC10883970
Yao , S. , M. Zhang , S. S. Dong , et al. 2022 . “Bidirectional Two‐Sample Mendelian Randomization Analysis Identifies Causal Associations Between Relative Carbohydrate Intake and Depression .” Nature Human Behaviour 6 , no. 11 : 1569Not reported1576 . 10.1038/s41562-022-01412-9 35851841
Zheng , J.‐S. , J. Luan , E. Sofianopoulou , et al. 2020 . “Plasma Vitamin C and Type 2 Diabetes: Genome‐Wide Association Study and Mendelian Randomization Analysis in European Populations .” Diabetes Care 44 , no. 1 : 98Not reported106 . 10.2337/dc20-1328 . 33203707 PMC7783939
Publication and source details

Original title: Genetically Supported Causality Between Micronutrients and Sleep Behaviors: A Two‐Sample Mendelian Randomization Study

Ruijie Zhang; Liyan Luo; Lu Zhang; Xinao Lin; Chuyan Wu; Feng Jiang; Jimei Wang

Brain and Behavior · 2025

* Correspondence: Jimei Wang (wjm821920@163.com) | Feng Jiang (dxyjiang@163.com)
#

Ruijie Zhang and Lian Luo contributed equally to this study.

© 2025 The Author(s). Brain and Behavior published by Wiley Periodicals LLC.https://creativecommons.org/licenses/by/4.0/This is an open access article under the terms of the http://creativecommons.org/licenses/by/4.0/ License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.Science and Technology Innovation Program of Shanghai20Z11900603

Repository JATS snapshot; version as supplied

Source retrieved: 2026-11-11T02:48:17.464083+00:00

DOI: 10.1002/brb3.70237 · PMC11799067