Skip to main content
Download PDF

Causal associations between sleep traits and brain structure: a bidirectional Mendelian randomization study

Behavioral and Brain Functions volume 19, Article number: 17 (2023) Cite this article

Abstract

Background

Emerging evidence suggests bidirectional causal relationships between sleep disturbance and psychiatric disorders, but the underlying mechanisms remain unclear. Understanding the bidirectional causality between sleep traits and brain imaging-derived phenotypes (IDPs) will help elucidate the mechanisms. Although previous studies have identified a range of structural differences in the brains of individuals with sleep disorders, it is still uncertain whether grey matter (GM) volume alterations precede or rather follow from the development of sleep disorders.

Results

After Bonferroni correction, the forward MR analysis showed that insomnia complaint remained positively associated with the surface area (SA) of medial orbitofrontal cortex (β, 0.26; 95% CI, 0.15–0.37; P = 5.27 × 10–6). In the inverse MR analysis, higher global cortical SA predisposed individuals less prone to suffering insomnia complaint (OR, 0.89; 95%CI, 0.85–0.94; P = 1.51 × 10–5) and short sleep (≤ 6 h; OR, 0.98; 95%CI, 0.97–0.99; P = 1.51 × 10–5), while higher SA in posterior cingulate cortex resulted in a vulnerability to shorter sleep durations (β, − 0.09; 95%CI, − 0.13 to − 0.05; P = 1.21 × 10–5).

Conclusions

Sleep habits not only result from but also contribute to alterations in brain structure, which may shed light on the possible mechanisms linking sleep behaviours with neuropsychiatric disorders, and offer new strategies for prevention and intervention in psychiatric disorders and sleep disturbance.

Background

Sleep, as a modifiable lifestyle habit, is essential for sustaining human life. It is characterized by multiple dimensions, including sleep quantity, quality, and circadian rhythm. Accumulating evidence suggests a complex bidirectional causality between unhealthy sleep patterns and neuropsychiatric disorders [1, 2]. Additionally, various sleep traits have emerged as potential markers or treatment targets for psychiatric disorders [3, 4]. For example, insomnia often precedes the onset of depressive disorder (DD) [5] and significantly increases the risk of developing DD in the future [6]; Sleep disturbances are prominent in patients with schizophrenia (SCZ) (up to 80% [7]), compared to the general population (around 20% [8]). However, the mechanisms linking sleep behaviours with neuropsychiatric disorders remain unclear.

Psychiatric disorders refer to a group of mental disorders characterized by psychological or behavioral abnormalities. Different brain structures have specific functions in controlling behavior and performance. Notably, the dorsolateral prefrontal cortex (PFC) and orbital frontal cortex (OFC) have been reported to be implicated in SCZ and bipolar disorder (BD), respectively [9]. Interestingly, these two brain regions have also been frequently identified as exhibiting morphological alterations in sleep-related MRI studies [10,11,12,13,14,15,16]. Considering that neuropathological changes often precede the clinical symptoms of neuropsychiatric disorders, it is hypothesized that variations in brain structure may underlie the causality between sleep habits and neuropsychiatric disorders, serving as either a cause or consequence of sleep patterns. Therefore, understanding the bidirectional causal relationships between sleep traits and brain imaging-derived phenotypes (IDPs) will help elucidate the mechanisms linking sleep behaviours with neuropsychiatric disorders.

At present, associations have been reported between global and regional differences in brain morphology and sleep quality in small case–control studies of insomnia, as well as moderate-sized observational studies conducted in general communities [10, 11, 15, 17,18,19]. However, the cross-sectional design of these neuroimaging studies limits their ability to infer the direction of the relationship between sleep patterns and brain structure. Furthermore, longitudinal studies addressing this issue have been unable to elucidate the temporal sequence between sleep habits and brain IDPs, as sleep conditions or brain magnetic resonance imaging (MRI) were evaluated at only one time point [16, 20,21,22,23,24]. Recently, the normal dynamics of bidirectional acute interaction between sleep duration and cortical thickness were revealed through microlongitudinal time series analyses based on data from one healthy individual [25]. However, no population-based longitudinal study or randomized controlled trial (RCT) has been published that can determine the bidirectional long-term causal relationships between sleep habits and brain structure.

Mendelian randomization (MR) analysis is an epidemiological study design that utilizes single nucleotide polymorphisms (SNPs) as instrumental variables (IVs) to establish causal effects of exposures on outcomes. This approach can be regarded as a natural RCT as it relies on the random allocation of alleles at gametogenesis, making it less prone to confounders and reverse causality compared to conventional observational multivariable regression. Previous studies have utilized MR analysis to explore the causal relationship between sleep and brain structure. However, these studies only examined the causal relationship between daytime napping and overall cortical volume [26], as well as sleep duration and overall cortical thickness [27]. A more comprehensive exploration of causal relationships between a wide range of sleep behaviors and global and regional brain structures is still warrant. This is critical for understanding what affects sleep habits and how adverse sleep behaviours contribute to a higher vulnerability to neuropsychiatric diseases. By elucidating the bidirectional causal relationships between sleep traits and brain structure, we may potentially provide guidance for prevention and intervention in psychiatric disorders and sleep disturbances.

Therefore, the objective of this study was to apply bidirectional two-sample MR analysis to investigate the causal effects between eight self-reported sleep phenotypes (namely insomnia complaint, sleep duration, long sleep, short sleep, chronotype, morningness, napping frequency and sleepiness severity) and 92 brain IDPs (including global and regional surface area [SA] and thickness [TH], volume of subcortical structures, and longitudinal changes in 15 brain structures) in adults of European ancestry, without any prior hypothesis.

Results

Overview of the study and instrument variants selection

We conducted a bidirectional two-sample MR analysis to investigate the relationships between all sleep traits-brain IDPs pairs. The study flowchart is presented in Fig. 1. To ensure that the samples used in the GWAS study for exposures were independent from those for the outcomes, we manually checked the sample description of each GWAS study. For subcortical brain structures, none of the participants were from the UK Biobank, indicating that there was no sample overlap between subcortical IDPs–sleep traits pairs; For cortical structures, 10 083 participants were from the UK Biobank, resulting in a sample overlap proportion of 29.7% (10,083/33,992); For longitudinal changes in brain structures, 2536 participants were from the UK Biobank, giving a sample overlap proportion of 16.8% (2536/15,100). After pruning for linkage disequilibrium, we removed the IVs associated with outcome as well as potential confounders. Outlier IVs detected by MR Pleiotropy RESidual Sum and Outlier (MR-PRESSO) outliers test were also excluded from exposure-related SNPs for subsequent MR analysis. The full lists of IVs used for forward and reverse MR tests were provided in Additional file 1: Table S1–S2. The Z scores of all exposure-outcome pairs derived from the random-effect inverse-variance weighted (IVW) method in the bidirectional two-sample MR analysis are showed in Fig. 2.

Fig. 1
figure 1

Flowchart of this bidirectional two-sample Mendelian randomization analysis. GWAS, genome-wide association studies; SNP, single-nucleotide polymorphism; BMI, body mass index; DD, depressive disorder; SCZ, schizophrenia; BD, bipolar disorder; ANX, anxiety disorder; OCD, obsessive–compulsive disorder; ASD, autism spectrum disorders; ADHD, attention deficit hyperactivity disorder; IVW, inverse-variance weighted; MR-PRESSO, MR Pleiotropy RESidual Sum and Outlier

Full size image
Fig. 2
figure 2

Z scores of all exposure-outcome pairs in the bidirectional two-sample Mendelian randomization analysis. All Z scores (β/SE) were derived using inverse-variance weighted method. The asterisk indicates significance after Bonferroni correction

Full size image

Forward Mendelian randomization

In the forward MR analysis, we identified that genetically predicted insomnia complaint was positively associated with the SA of medial orbitofrontal cortex (mOFC) (β, 0.26; 95% CI, 0.15–0.37; P = 5.27 × 10–6) (Fig. 3 and Additional file 1: Table S3). The other three MR methods corroborated this association (Fig. 3 and Additional file 1: Table S3). No heterogeneity was identified among the IVs for insomnia complaints based on the Cochran’s Q test (P = 0.899) (Additional file 1: Table S3). The P-values of MR-Egger intercept test and MR-PRESSO global test were 0.935 and 0.924, respectively, indicating the absence of horizontal pleiotropy (Additional file 1: Table S3). The scatter plot is shown in Additional file 1: Fig. S1. The estimate was not biased by any single SNP in the leave-one-out analysis (Additional file 1: Fig. S1). Furthermore, this association remained even when using non-overlapping IVs only in the MR analysis (β, 0.28; 95%CI 0.16–0.40; P = 3.65 × 10–6) (Additional file 1: Table S4). According to the study by Burgess et al. [28], the bias estimated from a 30% sample overlap for this exposure-outcome pair was less than 0.1%.

Fig. 3
figure 3

The significant causalities in the bidirectional two-sample Mendelian randomization analysis. These results all met the following criteria: the P-value derived from IVW method was < 3.40 × 10–5; all estimates from four Mendelian randomization methods showed consistent directions of association (positive or negative); no heterogeneity was identified by the Cochran’s Q test after removing the underlying outliers by MRPRESSO; P-value of MR-Egger intercept and MR-PRESSO global tests were both > 0.05; the estimate was not biased by single single-nucleotide polymorphism in leave-one-out analysis. SA, surface area; IVW, inverse-variance weighted; MR-PRESSO, MR Pleiotropy RESidual Sum and Outlier; IVs, instrumental variables

Full size image

Reverse Mendelian randomization

In the reverse MR analysis, we observed a negative correlation between the SA of global cortex and insomnia complaints (OR, 0.89; 95%CI, 0.85–0.94; P = 1.51 × 10–5), as well as short sleep (OR, 0.98; 95%CI, 0.97–0.99; P = 1.51 × 10–5) (Fig. 3 and Additional file 1: Table S5). We also found a negative association between the SA of posterior cingulate cortex (PCC) and sleep duration (β, − 0.09; 95%CI, − 0.13 to − 0.05; P = 1.21 × 10–5). The other three MR methods supported these associations (Fig. 3 and Additional file 1: Table S5). There was no heterogeneity or horizontal pleiotropy identified among these exposure-outcome pairs by the Cochran’s Q test, MR-Egger intercept test and MR-PRESSO global test (Additional file 1: Table S5). The scatter plots are shown in Additional file 1: Fig. S1, and the estimates were not biased by any single SNP in the leave-one-out analysis (Additional file 1: Fig. S1). There were no substantial alterations in the above associations when using non-overlapping SNPs only in the reverse MR analysis (Additional file 1: Table S6). According to the study by Burgess et al. [28], the bias estimated from a 30% sample overlap for these exposure-outcome pairs was all less than 0.1%.

Discussion

In this study, we provided reliable evidence for the bidirectional causal relationships between a wide range of sleep habits and brain MRI morphological measures using summary-level data from large-scale GWAS. Forward MR analysis suggests that insomnia complaints have a causal effect on the SA of mOFC. Reverse MR analysis showed that individuals with reduced global cortical SA may be predisposed to experiencing insomnia complaints and short sleep (≤ 6 h/day), while reduced cortical SA in PCC could contribute to a vulnerability for longer sleep duration.

There is considerable interest in identifying modifiable risk factors that exert causal effects on brain structure because changes in brain structure can precede the onset of cognitive decline or dementia by several years. The finding that insomnia complaint was associated with the morphology of mOFC appears to be robust, given that previous sleep-related MRI literature have most frequently detected morphological alterations in the OFC or PFC not only among insomnia patients [10,11,12,13], but also in healthy individuals with higher insomnia severity [14], early-morning awakenings [15], and poor sleep quality [16, 18]. However, these studies were mostly cross-sectional, with the exception of one derived from a longitudinal cohort study in which sleep quality was assessed at only one time point [16]. Therefore, these studies could not determine whether alterations in grey matter (GM) morphology precede or follow the development of sleep disturbances.

Using genetic instruments identified in large-scale GWAS, this MR study identified robust evidence supporting a causal effect of insomnia complaints on the SA of mOFC. Considering that insufficient sleep has adverse effects on a variety of neurobiological processes that can potentially affect GM morphology (e.g., metabolite clearance and synaptic homeostasis), it is plausible that insomnia can directly impact brain structure. Some prior studies that identified a correlation between insomnia and morphological changes in the mOFC tended to support the view that sleep is a consequence rather than a contributor to mOFC structural abnormalities, as the GM reduction was not specifically related to insomnia duration [10, 11], contradicting the findings of this study. However, it is possible that insomnia duration may be nonlinearly correlated with OFC structural changes, with compensatory GM enlargement occurring during early stages followed by subsequent atrophy. If this hypothesis holds true, it would make sense that some studies observed GM hypertrophy in primary insomnia [12, 13, 17, 29], while others reported a negative correlation between insomnia and GM volume [10,11,12, 14]. Evidence supporting this speculation is that the mean insomnia duration of patients in the studies where cortical atrophy was observed were 7.6 and 17.7 years [10, 11], compared to 4.9 years in study where cortical hypertrophy was observed [13]. Future longitudinal studies are warranted to determine whether the neuropathology of insomnia is initial hypertrophy and later atrophy. Additionally, it cannot be denied that heterogeneity in sample size, demographic characteristics, neuroimaging analysis approach (e.g., surface-based morphometry and voxel-based morphometry) and brain imaging indicators (e.g., TH, volume and density) may also contribute to these inconsistencies.

The present study found a causal effect of insomnia on the SA in mOFC, specifically, insomnia complaint may lead to an increase rather than a decrease in it, which is consistent with a recent study reporting that primary insomnia patients showed cortical thickening and increased cortical volume in the left OFC [13]. Moreover, insomnia-related GM hypertrophy in other brain regions has also been reported [12, 17, 29]. These hypertrophic cortices can be explained by activity-related changes, for instance, skilled musicians showed increases in auditory cortical representation [30]; extensive learning tasks resulted in an increased volume of right hippocampal [31]. Therefore, the increased SA of mOFC maybe a compensation of persistent deleterious effect of hyperarousal, as the mOFC is involved in decision-making, behavioral flexibility, and social behavior. The cellular mechanism underlying the increased cortical SA is still unknown, potential process may include hypertrophy of neurons or glial cells, changes in the size or density of the capillaries, and remodeling of dendritic spines and synaptic connections [17]. Besides, shorter sleep duration and poorer sleep quality have been proved to be associated with greater Aβ burden [32]. Therefore, the increment of the sleep-related cortical SA may be explained by the space-occupying effects of amyloid plaques and other metabolic waste.

In addition, it is also possible that the relationship between insomnia complaints and SA of mOFC could be mediated by potential confounders. In the current study, the IVs used for final MR analysis excluded the SNPs associated with education, BMI/obesity, hypertension, and seven common psychiatric disease (DD, SCZ, BD, anxiety disorder [ANX], obsessive–compulsive disorder [OCD], autism spectrum disorders [ASD] and attention deficit hyperactivity disorder [ADHD]). Furthermore, age, sex, and a maximum of 10 principal components were included as confounding regressors in the primary GWAS studies.

Emerging evidence suggests complex causalities between sleep problems and neuropsychiatric disorders [3]. However, the mechanisms remain unclear. The causal relationship between insomnia complaints and SA alteration in mOFC, identified in this study, may provide insights into the underlying mechanism. The mOFC has functional connectivity with a number of brain regions involved in higher cognitive functions (such as hippocampus, amygdala, prefrontal lobe, dorsolateral thalamic nucleus, anterior cingulate gyrus, etc.) and plays an important role in the pathophysiological mechanisms of multiple psychiatric disorders (such as DD, SCZ, BD, OCD, ADHD, etc.) [33]. Sleep disturbances often precede the onset of psychiatric disorders, or develop in their early stages as the chief complaint on the first visit [34]. A study designed to describe residual symptoms in 943 patients with remission from major DD after treatment with citalopram reported that the most common residual symptom domains were sleep disturbance (71.7%), indicating that sleep disturbance (primarily insomnia) did not disappear with the remission of depression [35]. All these clinical phenomena suggest that insomnia may lie “upstream” in the causal chain of psychiatric disorders. Therefore, we propose the hypothesis that structural alterations in mOFC may partially mediate the link between insomnia and neuropsychiatric disorders, which needs to be tested in future studies. If the hypothesis holds, it is necessary to further explore whether sleep interventions can arrest, slow or reverse the progression of neuropsychiatric disorders. At the very least, our findings support that sleep patterns may be a cause of brain structural alteration, enhancing the importance of measuring and adjust for insomnia complaints in neuropsychiatric research aiming to delineate the morphological correlates or even antecedents of psychopathology.

Alternatively, we also found that sleep architecture can be a consequence of brain structural abnormalities. In the reverse MR analysis, we identified a negative correlation of the SA of global cortex with insomnia complaint and short sleep, which may partially account for the higher prevalence of sleep impairment in older adults [36], since global brain atrophy is commonly seen as we age. Sleep problems are also frequently reported in neurodegenerative diseases [37], such as Alzheimer disease and Parkinson’s disease, the finding that brain atrophy could increase vulnerability to insomnia provides new insights into the mechanisms of developing chronic insomnia in neurodegenerative diseases and supports efforts to promote healthy brain aging through cognitive-behavioral therapy and lifestyle improvements. Interestingly, unlike the global cortical SA, the decreased SA in PCC may result in longer sleep duration. In the past few decades, scientists have devoted significant effort to identifying the subcortical brain regions responsible for wakefulness and sleep, and the possibility that cortical neurons regulate vigilance states has been overlooked. Recently, a study from Oxford University observed that silencing of layer 5 pyramidal and archicortical dentate gyrus granule cells in male mice markedly increased wakefulness and reduced rebound of slow-wave activity after sleep deprivation, supporting a role for the cortex in sleep homeostasis [38]. Additionally, the default mode network (DMN), a network of brain regions that display increased activity during wakeful rest in the absence of cognitively demanding tasks, plays a central role in the modulation of consciousness. During the transition from wakefulness to slow-wave sleep, functional connectivity within the DMN, particularly between the frontal (medial prefrontal, anterior cingulate) and posterior regions (PCC and precuneus) of the DMN, displayed a disconnected status, which is associated with the reduction in consciousness [39]. Therefore, it is plausible that a reduction in the SA of PCC may predispose individuals to sleep. However, the effect size was small, indicating that variables other than brain morphology significantly contribute to explaining the variance in sleep architecture.

There are some limitations in our study. First, sleep traits were self-reported rather than objectively measured, such as by using polysomnography, but it is usually not feasible in a large cohort study. Previous studies have suggested a moderate correlation between self-reported and objectively measured actigraphy [40]. Second, we did not explore the potential relationships between sleep habits and white matter microstructure, given that the GWAS data on human white matter microstructure were obtained from a meta-analysis of 43,802 subjects, including 36,624 from UK Biobank, which will lead to an overlap of up to 83.6% between the exposure and outcome samples. Third, in the inverse MR analysis, the genome-wide significance threshold was set at P < 5 × 10−6 due to the limited sample size of the brain structure GWAS, resulting in a lack of significant SNPs available when threshold was set at P < 5 × 10−8. This method of relaxing the statistical threshold for IVs has been used in previous high quality MR research when few associated SNPs are available [41]. Considering that this method carries the risk of introducing weak IVs, we selected SNPs with F statistics > 10 and conducted a series of sensitivity analyses. Only the consistent estimates from the four MR analyses, without horizontal pleiotropy and not biased by single SNP, were considered reliable. Fourth, there was partial overlap between the exposure and outcome datasets (29.7% for cortical structure and 16.8% for longitudinal change of brain structure), which may potentially bias the results. However, due to the lack of individual-level GWAS data, we were unable to remove the overlapping samples. Nonetheless, according to the study by Burgess et al. [28], the bias estimated from a 30% sample overlap was < 0.1% for the four causally associated exposure-outcome pairs identified in this MR research. Therefore, we speculate that the impact of sample overlap between GWAS studies of exposures and outcomes on our MR estimates is minor. Fifth, as the population we studied was Europeans, the estimates cannot be generalized to other races.

In conclusion, this study provided reliable evidence that sleep traits have a causal effect on brain GM structure. Additionally, it raised the possibility that the structural correlates of sleep measures in cross-sectional studies may represent pre-existing morphological GM deficits, which could increase susceptibility to certain sleep patterns. Undoubtedly, understanding whether modifiable lifestyle habits have a causal effect on the brain is crucial for allowing appropriate intervention. These findings imply that addressing and managing insomnia complaints may potentially mitigate changes in brain structural dynamics, thereby contributing to the prevention of neurological and psychiatric disorders. Further support for these suggestions is needed through additional sleep intervention trials and long-term cohort studies. This study also warrants further laboratory experiments to uncover the mechanism underlying the bidirectional causal relationships.

Methods

The aim, design and setting of the study

We conducted a bidirectional two-sample MR analysis to explore causal relationships between sleep behaviours and brain IDPs using summary-level results of published GWAS studies from the UK Biobank and the Enhancing Neuroimaging Genetics through Meta-Analysis (ENIGMA) Consortium. The study flowchart is presented in Fig. 1.

The definition of sleep traits

  1. (1)

    Insomnia complaint: Participants were asked: “Do you have trouble falling asleep at night or do you wake up in the middle of the night?” with responses “Never/rarely,” “Sometimes,” and “Usually”. Participants who answered “usually” were considered as having insomnia complaints, while those who answered “never/rarely” or “sometimes” were defined as controls in the GWAS study [42].

  2. (2)

    Sleep duration, long and short sleep: Participants were asked: “About how many hours sleep do you get in every 24 h? (Please include naps)”, with responses in hour increments. Sleep duration was treated as a continuous variable in GWAS analysis. Binary variables for short sleep (≤ 6 h vs. 7–8 h) and long sleep (≥ 9 h vs. 7–8 h) were also derived. Extreme responses of less than 3 h or more than 18 h were excluded [43].

  3. (3)

    Chronotype and morningness: Participants were prompted to answer the question “Do you consider yourself to be?” with answers: “Definitely a ‘morning’ person”, “More a ‘morning’ than ‘evening’ person”, “More an ‘evening’ than a ‘morning’ person”, “Definitely an ‘evening’ person”, or “Do not know”, which were coded as 2, 1, − 1, − 2, and 0, respectively. Participants who answered “Definitely a ‘morning’ person” and “More a ‘morning’ than ‘evening’ person” were categorized as cases for morningness, while those who answered “Definitely an ‘evening’ person” and “More an ‘evening’ than a ‘morning’ person” were considered as controls [44].

  4. (4)

    Napping frequency: Participants were asked “Do you have a nap during the day?” with responses “Never/rarely”, “Sometimes”, “Usually”. The responses were treated as a continuous variable in the GWAS [45].

  5. (5)

    Daytime sleepiness: This phenotype was determined by asking the question “How likely are you to dose off or fall asleep during the daytime when you don’t mean to? (e.g., when working, reading or driving)” with the response options of “Never/rarely”, “sometimes”, “often”, and “all of the time”. The responses were coded continuously as one to four, corresponding to the severity of daytime sleepiness [46].

GWAS summary data for sleep traits

Genetic variants associated with sleep traits, including insomnia complaints (N = 386,533) [42], sleep duration (N = 446,118) [43], long sleep (N = 339,926) [43], short sleep (N = 411,934) [43], chronotype (N = 449,734) [44], morningness (N = 403,195) [44], napping frequency (N = 452,633) [45], sleepiness severity (N = 452,071) [46], were available from corresponding GWAS studies conducted among European-ancestry adults in the UK Biobank. All associations have been adjusted for age, sex, a maximum of 10 principal components.

GWAS summary data for brain imaging-derived phenotypes

The summary-level GWAS data correlated with the human cerebral cortex structure were obtained from a meta-analysis of the GWAS studies conducted among 33,992 European-ancestry participants, including 23,909 participants from 49 cohorts affiliated with the ENIGMA Consortium and 10,083 participants from the UK Biobank [47]. The cohorts that were enrolled are listed in Additional file 1: Table S7. The SA and TH of the whole cortex and 34 brain regions defined by the Desikan-Killiany atlas were extracted from structural brain MRI scans. The summary-level GWAS data correlated with the volumes of seven subcortical regions (nucleus accumbens, caudate, putamen, pallidum, amygdala, hippocampus and thalamus) were obtained from a meta-analysis of 13,171 individuals of European ancestry from 28 cohorts participating in the ENIGMA Consortium [48]. The cohorts that were enrolled are listed in Additional file 1: Table S8. The subcortical measures were extracted by the automatic subcortical segmentation software packages: FIRST, part of the FMRIB Software Library (FSL), and the FreeSurfer. The summary-level GWAS data correlated with the longitudinal changes in 15 brain structures across the human lifespan were obtained from a meta-analysis of 15,100 European-ancestry participants, including 12,564 participants from 35 cohorts affiliated with the ENIGMA Consortium and 2536 participants from the UK Biobank [49]. The cohorts that were enrolled are listed in Additional file 1: Table S9. The 15 brain structures consisted of seven subcortical structures and eight global brain measures (total brain including cerebellum and excluding brainstem, SA measured at the gray–white matter boundary, average cortical TH, total lateral ventricle volume and cortical and cerebellar gray and white matter volume). All of these measures were extracted using the FreeSurfer processing pipeline.

Selection of instrument variant

We selected instrument variants (IVs) from two different GWAS summary results to perform a two-sample MR analysis, which can increase the estimated power. The selection of IVs should satisfy three assumptions: (1) IVs are strongly associated with exposure; (2) IVs should not be associated with potential confounders; (3) IVs influence the outcome only through the exposure of interest. To fulfill the first assumption, we chose two sets of P-values for genetic variants associated with the exposure in the bidirectional MR analysis. We used a threshold of P < 5 × 10−8 as the criterion for genome-wide significance to select IVs for estimating the causal effects of sleep traits on brain IDPs. In the inverse MR analysis for causal estimation of brain IDPs on sleep traits, the threshold of genome-wide significance was set at P < 5 × 10−6. We relaxed the statistical threshold for selecting IVs given that few (usually less than 4) or even none SNPs were identified when we used P < 5 × 10−8 as the threshold. However, as reported in the original GWAS study of the human brain structure [47], common variants explained 34% of the variation in total SA and 26% in average TH. These heritability estimates suggest that SNPs beyond those identified at P < 5 × 10−8 may contribute to variation in brain structure. This method of relaxing the statistical threshold for IVs has been used in previous high quality MR research when few associated SNPs are available [41]. Considering that this method carries the risk of introducing weak IVs, we calculated the F-statistic of each SNP and only the SNPs with F-statistic > 10 were retained. We also used linkage disequilibrium clumping (r2 > 0.001, and < 1 MB) to obtain independent SNPs associated with the exposure, and excluded the SNPs with minor allele frequency of < 0.01. For the second assumption, we removed SNPs associated with confounders that interfere with the association between brain structures and sleep traits. We determined age, sex, education, body mass index (BMI)/obesity, hypertension, and common psychiatric disorders (DD, SCZ, BD, ANX, OCD, ASD and ADHD) as confounders after reviewing a large amount of literature on the correlation between sleep and brain structure, as well as their respective influencing factors. These factors have been reported by previous studies to influence both sleep traits and brain IDPs [50,51,52]. In addition, these factors are also the most frequently adjusted covariates in sleep-related MRI studies [16, 53,54,55,56]. While some studies do not include mental status or psychological test scores as covariates, participants with psychiatric diseases are directly exclude [11, 20, 57]. Therefore, it is necessary for us to remove the SNPs related to these factors from the selected IVs for sleep traits or brain structures, in order to avoid attributing the causal effects of these factors to sleep habits or brain structure. Considering that all associations in the original GWAS studies have been adjusted for age, sex, a maximum of 10 principal components, then we further identified and removed the SNPs that are associated with education, BMI/obesity, hypertension, and common psychiatric disorders in the PhenoScanner database (http://www.phenoscanner.medschl.cam.ac.uk/) and the NHGRI-EBI GWAS catalog database (https://www.ebi.ac.uk/gwas/docs/file-downloads/) via PhenoScanner (version 1.0) [58] and gwasrapidd (version 0.99.13) [59] R Package. For the third assumption, we removed the outcome-related SNPs (p < 5 × 10–8) as well as the SNPs whose high LD SNPs are associated with the outcome. Besides, palindromic SNPs were also excluded after harmonizing the exposure and outcome data.

Mendelian randomization analysis

Four different methods of MR (random-effect IVW [60], MR-Egger [61], weighted median [62], and weighted mode [63]) were conducted to obtain causal effects of the exposure on outcome, taking into account variants heterogeneity and pleiotropy effects. The random-effect IVW method was implemented as the primary statistical analysis because it provides estimates with the highest precision by combining the ratios of SNP-exposure to SNP-outcome in a random-effects meta-analysis to estimate the causal relationship between the exposure and outcome. However, this method relies on the assumption of no directional horizontal pleiotropy, as it constrains the intercept of the regression to zero. Therefore, MR-Egger, weighted median and weighted mode were performed to complement and enhance the robustness of the results. The MR-Egger approach models a pleiotropy parameter by fitting an intercept term, allowing it to detect and correct for directional pleiotropy, albeit with compromised power [61]. The weighted median method can yield valid causal effects if at least half of the weight in the analysis comes from valid IVs [62]. The weighted mode method provides consistent estimates when the relaxed IV assumption has less bias and a lower type I error rate [63]. When only one genetic instrument was available, we used the Wald ratio method for MR analysis.

Sensitivity analysis

To improve the reliability of the genetic instruments, we conducted Cochran’s Q test to identify heterogeneity and removed any outlying SNPs by applying MR-PRESSO outliers test [64] before conducting the final MR analysis. We also performed the MR-Egger intercept test and MR-PRESSO global test to assess horizontal pleiotropy. Additionally, we applied leave-one-out analysis to check whether the causal relationship was mainly driven by a single SNP. To eliminate potential interference from overlapping IVs between exposure phenotypes, we re-conducted the MR analysis after removing the overlapping SNPs.

Statistics

All analyses were performed using the TwoSampleMR package (version 0.5.6) in R (version 4.1.2). A Bonferroni-corrected P-value threshold was set at 3.40 × 10−5 (0.05/[92 × 8 × 2]; 92 represents the number of brain IDPs, eight represents the number of sleep traits, and two represents both forward and reverse MR analysis). The significant estimates should meet the following criteria: The P-value derived from IVW method was < 3.40 × 10–5; all estimates from the four MR methods showed consistent directions of association (either positive or negative); no heterogeneity was identified by the Cochran’s Q test after removing the outlying SNPs by MRPRESSO; the P-values of MR-Egger intercept and MR-PRESSO global tests were both > 0.05, indicating the absence of horizontal pleiotropy; the estimate was not biased by a single SNP, as indicated by the leave-one-out plot. In the forward MR, the effect estimates [β and 95% CI] for cortical structures were divided by standard deviation (SD) to visually measure the effect size. However, for subcortical structures and longitudinal changes of brain structures, the original results were presented given that the GWAS study did not provide SD information. In the reverse MR, the effect estimates for cortical structures were calculated as the change in sleep traits per SD change in brain IDPs, while the effect estimates for subcortical structures and longitudinal changes of brain structures were presented as original results. Moreover, for binary outcome, including short sleep, long sleep, and morningness, the effect estimates were transformed from [β and 95% CI] to [OR and 95% CI].

Availability of data and materials

The summary-level datasets of the human brain structure analyzed during the current study are available from the ENIGMA Consortium website, http://enigma.ini.usc.edu/research/download-enigma-gwas-results/. The datasets of the sleep traits analyzed during the current study are available from https://ctg.cncr.nl/ or http://sleepdisordergenetics.org/.

Abbreviations

IDPs:

Imaging-derived phenotypes

MR:

Mendelian randomization

The ENIGMA Consortium:

The Enhancing Neuroimaging Genetics through Meta-Analysis Consortium

SA:

Surface area

TH:

Thickness

MRI:

Magnetic resonance imaging

RCT:

Randomized controlled trial

SNPs:

Single nucleotide polymorphisms

IVs:

Instrumental variables

GWAS:

Genome-wide association studies

FSL:

FMRIB Software Library

BMI:

Body mass index

DD:

Depressive disorder

SCZ:

Schizophrenia

BD:

Bipolar disorder

ANX:

Anxiety disorder

OCD:

Obsessive–compulsive disorder

ASD:

Autism spectrum disorders

ADHD:

Attention deficit hyperactivity disorder

MR-PRESSO:

MR Pleiotropy RESidual Sum and Outlier

IVW:

Inverse-variance weighted

SD:

Standard deviation

CI:

Confidence interval

OR:

Odds ratio

mOFC:

Medial orbitofrontal cortex

PCC:

Posterior cingulate cortex

PFC:

Prefrontal cortex

DMN:

Default mode network

GM:

Grey matter

References

  1. Sun X, Liu B, Liu S, Wu DJH, Wang J, Qian Y, et al. Sleep disturbance and psychiatric disorders: a bidirectional Mendelian randomisation study. Epidemiol Psychiatric Sci. 2022;31: e26.

    Google Scholar 

  2. Wang Z, Chen M, Wei YZ, Zhuo CG, Xu HF, Li WD, et al. The causal relationship between sleep traits and the risk of schizophrenia: a two-sample bidirectional Mendelian randomization study. BMC Psychiatry. 2022;22(1):399.

    CAS  PubMed  PubMed Central  Google Scholar 

  3. Krystal AD. Sleep therapeutics and neuropsychiatric illness. Neuropsychopharmacology. 2020;45(1):166–75.

    PubMed  Google Scholar 

  4. Sutton EL. Psychiatric disorders and sleep issues. Med Clin North Am. 2014;98(5):1123–43.

    PubMed  Google Scholar 

  5. Johnson EO, Roth T, Breslau N. The association of insomnia with anxiety disorders and depression: exploration of the direction of risk. J Psychiatr Res. 2006;40(8):700–8.

    PubMed  Google Scholar 

  6. Franzen PL, Buysse DJ. Sleep disturbances and depression: risk relationships for subsequent depression and therapeutic implications. Dialogues Clin Neurosci. 2008;10(4):473–81.

    PubMed  PubMed Central  Google Scholar 

  7. Klingaman EA, Palmer-Bacon J, Bennett ME, Rowland LM. Sleep disorders among people with schizophrenia: emerging research. Curr Psychiatry Rep. 2015;17(10):79.

    PubMed  Google Scholar 

  8. Wang J, Wu J, Liu J, Meng Y, Li J, Zhou P, et al. Prevalence of sleep disturbances and associated factors among Chinese residents: a web-based empirical survey of 2019. J Glob Health. 2023;13:04071.

    PubMed  PubMed Central  Google Scholar 

  9. Qi X, Wu C, Du Y, Cheng S, Wen Y, Ma M, et al. Comparing GWAS and brain structure-specific gene expression profiles identifies psychiatric disorder-related brain structures at different developmental stages. Neurosci Bull. 2020;36(9):1046–50.

    PubMed  PubMed Central  Google Scholar 

  10. Altena E, Vrenken H, Van Der Werf YD, van den Heuvel OA, Van Someren EJ. Reduced orbitofrontal and parietal gray matter in chronic insomnia: a voxel-based morphometric study. Biol Psychiat. 2010;67(2):182–5.

    PubMed  Google Scholar 

  11. Joo EY, Noh HJ, Kim JS, Koo DL, Kim D, Hwang KJ, et al. Brain gray matter deficits in patients with chronic primary insomnia. Sleep. 2013;36(7):999–1007.

    PubMed  PubMed Central  Google Scholar 

  12. Grau-Rivera O, Operto G, Falcón C, Sánchez-Benavides G, Cacciaglia R, Brugulat-Serrat A, et al. Association between insomnia and cognitive performance, gray matter volume, and white matter microstructure in cognitively unimpaired adults. Alzheimer’s Res Ther. 2020;12(1):4.

    Google Scholar 

  13. Yu S, Feng F, Zhang Q, Shen Z, Wang Z, Hu Y, et al. Gray matter hypertrophy in primary insomnia: a surface-based morphometric study. Brain Imaging Behav. 2020;14(5):1309–17.

    PubMed  Google Scholar 

  14. Falgàs N, Illán-Gala I, Allen IE, Mumford P, Essanaa YM, Le MM, et al. Specific cortical and subcortical grey matter regions are associated with insomnia severity. PLoS ONE. 2021;16(5): e0252076.

    PubMed  PubMed Central  Google Scholar 

  15. Stoffers D, Moens S, Benjamins J, van Tol MJ, Penninx BW, Veltman DJ, et al. Orbitofrontal gray matter relates to early morning awakening: a neural correlate of insomnia complaints? Front Neurol. 2012;3:105.

    PubMed  PubMed Central  Google Scholar 

  16. Sexton CE, Storsve AB, Walhovd KB, Johansen-Berg H, Fjell AM. Poor sleep quality is associated with increased cortical atrophy in community-dwelling adults. Neurology. 2014;83(11):967–73.

    PubMed  PubMed Central  Google Scholar 

  17. Winkelman JW, Plante DT, Schoerning L, Benson K, Buxton OM, O’Connor SP, et al. Increased rostral anterior cingulate cortex volume in chronic primary insomnia. Sleep. 2013;36(7):991–8.

    PubMed  PubMed Central  Google Scholar 

  18. Alperin N, Wiltshire J, Lee SH, Ramos AR, Hernandez-Cardenache R, Rundek T, et al. Effect of sleep quality on amnestic mild cognitive impairment vulnerable brain regions in cognitively normal elderly individuals. Sleep. 2019;42(3).

  19. Branger P, Arenaza-Urquijo EM, Tomadesso C, Mézenge F, André C, de Flores R, et al. Relationships between sleep quality and brain volume, metabolism, and amyloid deposition in late adulthood. Neurobiol Aging. 2016;41:107–14.

    CAS  PubMed  Google Scholar 

  20. Lo JC, Loh KK, Zheng H, Sim SK, Chee MW. Sleep duration and age-related changes in brain structure and cognitive performance. Sleep. 2014;37(7):1171–8.

    PubMed  PubMed Central  Google Scholar 

  21. Spira AP, Gonzalez CE, Venkatraman VK, Wu MN, Pacheco J, Simonsick EM, et al. Sleep duration and subsequent cortical thinning in cognitively normal older adults. Sleep. 2016;39(5):1121–8.

    PubMed  PubMed Central  Google Scholar 

  22. Zitser J, Anatürk M, Zsoldos E, Mahmood A, Filippini N, Suri S, et al. Sleep duration over 28 years, cognition, gray matter volume, and white matter microstructure: a prospective cohort study. Sleep. 2020;43(5).

  23. Fjell AM, Sørensen Ø, Amlien IK, Bartrés-Faz D, Bros DM, Buchmann N, et al. Self-reported sleep relates to hippocampal atrophy across the adult lifespan: results from the Lifebrain consortium. Sleep. 2020;43(5).

  24. Fjell AM, Sørensen Ø, Amlien IK, Bartrés-Faz D, Brandmaier AM, Buchmann N, et al. Poor self-reported sleep is related to regional cortical thinning in aging but not memory decline-results from the lifebrain consortium. Cereb Cortex. 2021;31(4):1953–69.

    PubMed  Google Scholar 

  25. Wall J, Xie H, Wang X. Interaction of sleep and cortical structural maintenance from an individual person microlongitudinal perspective and implications for precision medicine research. Front Neurosci. 2020;14:769.

    PubMed  PubMed Central  Google Scholar 

  26. Paz V, Dashti HS, Garfield V. Is there an association between daytime napping, cognitive function, and brain volume? A Mendelian randomization study in the UK Biobank. Sleep health. 2023.

  27. Andrews SJ, Fulton-Howard B, O’Reilly P, Marcora E, Goate AM. Causal associations between modifiable risk factors and the Alzheimer’s phenome. Ann Neurol. 2021;89(1):54–65.

    CAS  PubMed  Google Scholar 

  28. Burgess S, Davies NM, Thompson SG. Bias due to participant overlap in two-sample Mendelian randomization. Genet Epidemiol. 2016;40(7):597–608.

    PubMed  PubMed Central  Google Scholar 

  29. Baril AA, Gagnon K, Brayet P, Montplaisir J, De Beaumont L, Carrier J, et al. Gray matter hypertrophy and thickening with obstructive sleep apnea in middle-aged and older adults. Am J Respir Crit Care Med. 2017;195(11):1509–18.

    CAS  PubMed  Google Scholar 

  30. Pantev C, Oostenveld R, Engelien A, Ross B, Roberts LE, Hoke M. Increased auditory cortical representation in musicians. Nature. 1998;392(6678):811–4.

    CAS  PubMed  Google Scholar 

  31. Koch K, Reess TJ, Rus OG, Zimmer C. Extensive learning is associated with gray matter changes in the right hippocampus. Neuroimage. 2016;125:627–32.

    PubMed  Google Scholar 

  32. Spira AP, Gamaldo AA, An Y, Wu MN, Simonsick EM, Bilgel M, et al. Self-reported sleep and β-amyloid deposition in community-dwelling older adults. JAMA Neurol. 2013;70(12):1537–43.

    PubMed  PubMed Central  Google Scholar 

  33. Hiser J, Koenigs M. The multifaceted role of the ventromedial prefrontal cortex in emotion, decision making, social cognition, and psychopathology. Biol Psychiat. 2018;83(8):638–47.

    PubMed  Google Scholar 

  34. Zanini M, Castro J, Coelho FM, Bittencourt L, Bressan RA, Tufik S, et al. Do sleep abnormalities and misaligned sleep/circadian rhythm patterns represent early clinical characteristics for developing psychosis in high risk populations? Neurosci Biobehav Rev. 2013;37(10 Pt 2):2631–7.

    PubMed  Google Scholar 

  35. Nierenberg AA, Husain MM, Trivedi MH, Fava M, Warden D, Wisniewski SR, et al. Residual symptoms after remission of major depressive disorder with citalopram and risk of relapse: a STAR*D report. Psychol Med. 2010;40(1):41–50.

    CAS  PubMed  Google Scholar 

  36. Mander BA, Winer JR, Walker MP. Sleep and human aging. Neuron. 2017;94(1):19–36.

    CAS  PubMed  PubMed Central  Google Scholar 

  37. Wulff K, Gatti S, Wettstein JG, Foster RG. Sleep and circadian rhythm disruption in psychiatric and neurodegenerative disease. Nat Rev Neurosci. 2010;11(8):589–99.

    CAS  PubMed  Google Scholar 

  38. Krone LB, Yamagata T, Blanco-Duque C, Guillaumin MCC, Kahn MC, van der Vinne V, et al. A role for the cortex in sleep-wake regulation. Nat Neurosci. 2021;24(9):1210–5.

    CAS  PubMed  PubMed Central  Google Scholar 

  39. Horovitz SG, Braun AR, Carr WS, Picchioni D, Balkin TJ, Fukunaga M, et al. Decoupling of the brain’s default mode network during deep sleep. Proc Natl Acad Sci USA. 2009;106(27):11376–81.

    CAS  PubMed  PubMed Central  Google Scholar 

  40. Lauderdale DS, Knutson KL, Yan LL, Liu K, Rathouz PJ. Self-reported and measured sleep duration: how similar are they? Epidemiology. 2008;19(6):838–45.

    PubMed  PubMed Central  Google Scholar 

  41. Choi KW, Chen CY, Stein MB, Klimentidis YC, Wang MJ, Koenen KC, et al. Assessment of bidirectional relationships between physical activity and depression among adults: a 2-sample Mendelian randomization study. JAMA Psychiat. 2019;76(4):399–408.

    Google Scholar 

  42. Jansen PR, Watanabe K, Stringer S, Skene N, Bryois J, Hammerschlag AR, et al. Genome-wide analysis of insomnia in 1,331,010 individuals identifies new risk loci and functional pathways. Nat Genet. 2019;51(3):394–403.

    CAS  PubMed  Google Scholar 

  43. Dashti HS, Jones SE, Wood AR, Lane JM, van Hees VT, Wang H, et al. Genome-wide association study identifies genetic loci for self-reported habitual sleep duration supported by accelerometer-derived estimates. Nat Commun. 2019;10(1):1100.

    PubMed  PubMed Central  Google Scholar 

  44. Jones SE, Lane JM, Wood AR, van Hees VT, Tyrrell J, Beaumont RN, et al. Genome-wide association analyses of chronotype in 697,828 individuals provides insights into circadian rhythms. Nat Commun. 2019;10(1):343.

    CAS  PubMed  PubMed Central  Google Scholar 

  45. Dashti HS, Daghlas I, Lane JM, Huang Y, Udler MS, Wang H, et al. Genetic determinants of daytime napping and effects on cardiometabolic health. Nat Commun. 2021;12(1):900.

    CAS  PubMed  PubMed Central  Google Scholar 

  46. Wang H, Lane JM, Jones SE, Dashti HS, Ollila HM, Wood AR, et al. Genome-wide association analysis of self-reported daytime sleepiness identifies 42 loci that suggest biological subtypes. Nat Commun. 2019;10(1):3503.

    PubMed  PubMed Central  Google Scholar 

  47. Grasby KL, Jahanshad N, Painter JN, Colodro-Conde L, Bralten J, Hibar DP, et al. The genetic architecture of the human cerebral cortex. Science (New York, NY). 2020;367(6484).

  48. Hibar DP, Stein JL, Renteria ME, Arias-Vasquez A, Desrivières S, Jahanshad N, et al. Common genetic variants influence human subcortical brain structures. Nature. 2015;520(7546):224–9.

    CAS  PubMed  PubMed Central  Google Scholar 

  49. Brouwer RM, Klein M, Grasby KL, Schnack HG, Jahanshad N, Teeuw J, et al. Genetic variants associated with longitudinal changes in brain structure across the lifespan. Nat Neurosci. 2022;25(4):421–32.

    CAS  PubMed  PubMed Central  Google Scholar 

  50. Kurth F, Levitt JG, Phillips OR, Luders E, Woods RP, Mazziotta JC, et al. Relationships between gray matter, body mass index, and waist circumference in healthy adults. Hum Brain Mapp. 2013;34(7):1737–46.

    PubMed  Google Scholar 

  51. Besteher B, Gaser C, Langbein K, Dietzek M, Sauer H, Nenadić I. Effects of subclinical depression, anxiety and somatization on brain structure in healthy subjects. J Affect Disord. 2017;215:111–7.

    PubMed  Google Scholar 

  52. Lane CA, Barnes J, Nicholas JM, Sudre CH, Cash DM, Parker TD, et al. Associations between blood pressure across adulthood and late-life brain structure and pathology in the neuroscience substudy of the 1946 British birth cohort (Insight 46): an epidemiological study. The Lancet Neurology. 2019;18(10):942–52.

    CAS  PubMed  PubMed Central  Google Scholar 

  53. Lim AS, Fleischman DA, Dawe RJ, Yu L, Arfanakis K, Buchman AS, et al. Regional neocortical gray matter structure and sleep fragmentation in older adults. Sleep. 2016;39(1):227–35.

    PubMed  PubMed Central  Google Scholar 

  54. Yaffe K, Nasrallah I, Hoang TD, Lauderdale DS, Knutson KL, Carnethon MR, et al. Sleep duration and white matter quality in middle-aged adults. Sleep. 2016;39(9):1743–7.

    PubMed  PubMed Central  Google Scholar 

  55. Sexton CE, Zsoldos E, Filippini N, Griffanti L, Winkler A, Mahmood A, et al. Associations between self-reported sleep quality and white matter in community-dwelling older adults: a prospective cohort study. Hum Brain Mapp. 2017;38(11):5465–73.

    PubMed  PubMed Central  Google Scholar 

  56. Carvalho DZ, St Louis EK, Boeve BF, Mielke MM, Przybelski SA, Knopman DS, et al. Excessive daytime sleepiness and fatigue may indicate accelerated brain aging in cognitively normal late middle-aged and older adults. Sleep Med. 2017;32:236–43.

    PubMed  Google Scholar 

  57. Riemann D, Voderholzer U, Spiegelhalder K, Hornyak M, Buysse DJ, Nissen C, et al. Chronic insomnia and MRI-measured hippocampal volumes: a pilot study. Sleep. 2007;30(8):955–8.

    PubMed  PubMed Central  Google Scholar 

  58. Staley JR, Blackshaw J, Kamat MA, Ellis S, Surendran P, Sun BB, et al. PhenoScanner: a database of human genotype-phenotype associations. Bioinformatics (Oxford, England). 2016;32(20):3207–9.

    CAS  PubMed  Google Scholar 

  59. Magno R, Maia AT. gwasrapidd: an R package to query, download and wrangle GWAS catalog data. Bioinformatics (Oxford, England). 2020;36(2):649–50.

    CAS  PubMed  Google Scholar 

  60. Bowden J, Del Greco MF, Minelli C, Davey Smith G, Sheehan N, Thompson J. A framework for the investigation of pleiotropy in two-sample summary data Mendelian randomization. Stat Med. 2017;36(11):1783–802.

    PubMed  PubMed Central  Google Scholar 

  61. Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44(2):512–25.

    PubMed  PubMed Central  Google Scholar 

  62. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40(4):304–14.

    PubMed  PubMed Central  Google Scholar 

  63. Hartwig FP, Davey Smith G, Bowden J. Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. Int J Epidemiol. 2017;46(6):1985–98.

    PubMed  PubMed Central  Google Scholar 

  64. Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(5):693–8.

    CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We appreciated the UK Biobank and the Enhancing NeuroImaging Genetics through Meta-Analysis (ENIGMA) Consortium for generously sharing their GWAS summary statistics data.

Funding

This work was supported by the National Natural Science Foundation of China (Grant number 82071454), the Natural Science Foundation of Beijing Municipality (Grant number 7202062) and Beijing Municipal Science & Technology Commission (Grant number Z211100002921032) awarded to L.K.R. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Author information

Author notes

  1. Qiao Wang and Shimin Hu contributed equally to this work.

Authors and Affiliations

  1. Department of Neurology, Xuanwu Hospital, Capital Medical University, NO.45 Changchun Street, Xicheng District, Beijing, China

    Qiao Wang, Shimin Hu, Lei Qi, Xiaopeng Wang, Guangyuan Jin, Di Wu, Yuke Wang & Liankun Ren

  2. National Center for Neurological Disorders, Beijing, China

    Qiao Wang, Shimin Hu, Lei Qi, Xiaopeng Wang, Guangyuan Jin, Di Wu, Yuke Wang & Liankun Ren

  3. Beijing Key Laboratory of Neuromodulation, Beijing, China

    Shimin Hu

  4. Institute of Sleep and Consciousness Disorders, Center of Epilepsy, Beijing Institute for Brain Disorders, Capital Medical University, Beijing, China

    Shimin Hu

Authors
  1. Qiao Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Shimin Hu
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Lei Qi
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Xiaopeng Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Guangyuan Jin
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Di Wu
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Yuke Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Liankun Ren
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

QW developed the idea; QW, SMH and LKR contributed to the study design; WQ, SMH, LQ, XPW, GYJ, DW and YKW downloaded the GWAS data and carried out the statistical analysis; WQ, SMH, LQ, XPW, GYJ and LKR participated in the data interpretation; WQ, SMH, DW, and YKW prepared the first draft of the manuscript, and LKR revised the article All authors read and approved the final manuscript.

Corresponding author

Correspondence to Liankun Ren.

Ethics declarations

Ethics approval and consent to participate

This MR study used the de-identified summary-level data that have been made publicly available. The informed consent and ethical approval had been obtained in all original GWAS studies, thus this study did not require additional ethical approval.

Consent for publication

All authors consent for publication of the manuscript in its present form.

Competing interests

All authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Table S1.

Information of instrumental variables for all exposure-outcome pairs in forward Mendelian randomization analyses. Table S2. Information of instrumental variables for all exposure-outcome pairs in reverse Mendelian randomization analyses. Table S3. Forward Mendelian randomization analysis results. Table S4. Forward Mendelian randomization analysis results after removing the overlapping single nucleotide polymorphisms between exposure phenotypes. Table S5. Reverse Mendelian randomization analysis results. Table S6. Reverse Mendelian randomization analysis results after removing the overlapping single nucleotide polymorphisms between exposure phenotypes. Table S7. Descriptions of study cohorts participating in the genome-wide association studies meta-analysis to identify genetic variants associated with human cortical structure. Table S8. Descriptions of study cohorts participating in the genome-wide association studies meta-analysis to identify genetic variants associated with human subcortical brain structures. Table S9. Descriptions of study cohorts participating in the genome-wide association studies meta-analysis to identify genetic variants associated with longitudinal changes in brain structure across the lifespan.

Additional file 2: Figure S1.

Scatter plots and leave-one-out plots of significant estimates in both forward and reverse Mendelian randomization analyses. SA, surficial area.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wang, Q., Hu, S., Qi, L. et al. Causal associations between sleep traits and brain structure: a bidirectional Mendelian randomization study. Behav Brain Funct 19, 17 (2023). https://doi.org/10.1186/s12993-023-00220-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12993-023-00220-z

Keywords

Behavioral and Brain Functions

ISSN: 1744-9081

Contact us