1 Introduction
Appropriate sleep duration has been the focus of public health recommendations [1]. The National Sleep Foundation (NSF) and the American Academy of Sleep Medicine (AASM) recommend that all adults should sleep at least 7 hours per day [2,3]. However, the 2020 U.S. National Survey indicates that more than 25% of adults aged 18 and older sleep less than 7 hours [4]. Despite the adverse health effects of restorative sleep deprivation, sleep problems often go undiagnosed and untreated [5,6]. Although elevated challenges in sleeping may arise due to typical alterations in “sleep architecture,” persistent health conditions, or the side effects of medications, older individuals and their healthcare providers should not merely acknowledge sleep issues as an inherent aspect of aging without making efforts to address and alleviate them. A recent study found that adults with sleep problems, particularly insomnia, have significantly higher healthcare utilisation and healthcare spending. The total direct healthcare cost of sleep-related nationally exceeds $90 billion [7]. This may be related to the multiple downstream pathogenic events caused by too much or not enough sleep.
Recently, there has been increasing evidence that long sleep duration, short sleep duration, and insomnia are associated with cognitive function [8,9]. Cognition is a combination of brain processes, including the ability to learn, remember and make judgements. Impaired cognitive function may have a profound effect on an individual’s overall health and well-being. This is because cognitive decline ranges from mild cognitive impairment to dementia, which is a decline in ability severe enough to interfere with daily life [10]. Current treatments are not optimal for disorders associated with cognitive decline. Instead, sleep are manageable and preventable condition, so understanding sleep phenotypes has attracted increasing attention as a potential target for preventing cognitive decline.
However, current findings are not entirely consistent regarding the relationship between sleep duration and cognitive impairment. A cross-sectional study that included a community group of more than 2000 individuals aged 40 years or older and assessed cognitive performance using changes in scores on the Brief Mental Status Exercise (MMSE) scale showed that individuals with self-reported sleep duration ≤ 5h, 6h, 8h, and ≥9h had odds ratios (ORs) of 2.14, 1.13, 1.51, and 5.37 for cognitive impairment, respectively (all P < 0.01), which suggests a U-shaped association between sleep duration and cognitive impairment [11]. Keil et al. conducted a cross-sectional study of 826 older adults, which was a retrospective analysis of the Seattle Longitudinal Study [12]. The study used the MMSE and Mattis Dementia Rating Scale (DRS) to measure cognition, thereby exploring the association between self-reported sleep duration and cognitive performance in older adults. After moderating for multiple potential confounders, it was found that while both short sleep duration (<7 h/d) and long sleep duration (>7 h/d) showed an association with a high risk of cognitive impairment, short sleepers had a significantly increased risk of their cognitive impairment (HR = 3.67, P < 0.005), whereas the trend of the risk of cognitive impairment was not significant in the long sleepers (HR = 1.91, P > 0.05) [12]. A prospective longitudinal study included 119 healthy adults aged 55 years or older, subjectively assessed for sleep duration and quality, and underwent magnetic resonance imaging (MRI) testing and neuropsychological assessment every 2 years. After adjusting for multiple confounders, a linear regression relationship was found between short sleep duration and age-related brain atrophy including cerebral grey matter, superior frontal gyrus, and inferior frontal gyrus (all P < 0.05), in which each 1h decrease in sleep duration at baseline level (6.7 h) was associated with a 0.59% increase in ventricular dilatation (beta = –0.587, P < 0.01), which may suggest that short sleepers’ overall cognitive levels decreased [13]. Benito-leon et al. included 2751 community-dwelling older adults aged 65 years or older, including 298 short sleepers (<5h/d), 1086 long sleepers (≥9h/d), and 1331 regular sleepers (6–8h/d) to assess cognitive decline using changes in MMSE scores. After 3 years of follow-up, no statistically significant difference in cognitive function was found between short sleepers and regular sleepers (P = 0.142), whereas long sleepers showed a decline in cognitive function when compared with regular sleepers (P = 0.040), and the difference in cognitive function remained statistically significant after adjusting for relevant confounders (P < 0.05) [14].
It is worth noting that apart from the inconsistency of the findings, most of the previous studies were cross-sectional and cohort studies. However, there are some shortcomings in drawing causal inferences with the help of observational studies. First, in assessing the causal relationship between sleep characteristics and cognitive function, different observational studies used different adjustment models and did not fully incorporate possible confounders due to the limitations of the studies themselves. Second, when examining the role of sleep phenotypes on cognitive function, they are susceptible to reverse causality bias. Third, differences in definitions of cognitive function (scores) as well as long sleep duration, short sleep duration and insomnia allow for some heterogeneity between studies. The causal relationship between long sleep duration, short sleep duration, insomnia, and cognitive function is currently unclear. Mendelian randomization (MR) studies may be an effective way to explore this issue.
MR is a novel epidemiological method based on genome-wide association studies (GWAS) that uses genetic variants such as single nucleotide polymorphisms (SNPs) as instrumental variables (IVs) to reveal causal relationships [15]. These genetic variants are equally, randomly and independently distributed when assigned.The greatest advantage of MR is that it avoids the risk of bias associated with observational experiments such as potential confounding, reverse causation, etc., and is to some extent similar to randomised controlled studies [15].
This study investigated the causal relationships between long sleep duration, short sleep duration, insomnia and multiple cognitive outcomes through comprehensive two-sample MR analyses of seven cognitive function tests (including cognitive performance, fluid Intelligence score [FIS], memory performance, Trail Making (TM) test, Symbol Digit Substitution (SDS) test, Pair Matching (PM) test, reaction time (RT)) and four dementia types (including Alzheimer’s disease [AD], Lewy body dementia, vascular dementia and frontotemporal dementia) derived from the IEU Open GWAS project. The aim was to genetically elucidate the role of these sleep phenotypes in the development of cognitive decline, thereby contributing to the development of new preventive strategies.
2 Methods
2.1 Mendelian randomization study design
This study utilized a two-sample MR design employing summary statistics of SNPs as IVs to elucidate causal relationships between self-reported long sleep duration, short sleep duration, insomnia, and cognitive function (see Fig 1). The summary-level data utilized in this investigation were sourced from publicly accessible European pedigree GWAS. The data employed in this MR analysis were drawn from extensive GWAS endeavors, with the original cohorts having received appropriate ethical clearances and obtained informed consent from participants. Consequently, no separate ethical approval was deemed necessary for the present study.
[Figure omitted. See PDF.]
The validity of such analyses hinges upon the fulfillment of three pivotal assumptions. Firstly, the genetic variant is strongly associated with self-reported sleep phenotypes. Secondly, it is imperative that the genetic variant remains independent of potential confounding variables. Thirdly, the genetic variant does not have a direct relationship with cognitive function, but affects cognitive function only through pathways that involve these self-reported sleep phenotypes. Abbreviations. TM = Trail Making, SDS = Symbol Digit Substitution, PM = Pair Matching.
2.2 Data sources
2.2.1 Long sleep duration, short sleep duration, and insomnia.
Through a study analysed genome-wide associations in 446,118 adults of European ancestry, Dashti et al (2019) found 78 loci for sleep duration [16]. Participants in the study were drawn from the UK Biobank study, a prospective cohort study of over 500,000 UK participants. Notably, although the study sleep duration was self-reported, the results were supported by accelerometer-derived estimates. The 78 loci were generally consistent with accelerometer estimates of sleep duration. Participants measured this by wearing a wrist-worn accelerometer for up to 7 days [16]. Relative to a sleep duration of 7–8 h (n = 305,742 controls), separate GWAS for short sleep (<7 h; n = 106,192 cases) and long sleep (≥9 h; n = 34,184 cases) identified 27 and 8 loci, respectively [16]. Study participants (approximately 500,000) self-reported sleep duration at the baseline assessment. Participants were asked: Approximately how many hours of sleep do you get every 24 hours (please include naps)? Responses were given in hourly increments. Normal sleep duration is 7 or 8 hours. Sleeping 6 hours or less per 24 hours was defined as short sleep duration. Sleeping 9 hours or more per 24 hours was defined as long sleep duration. These thresholds were chosen based on previous epidemiological and clinical studies that have shown that both less than 6 hours and more than 9 hours of sleep are associated with adverse health outcomes, including increased risk of cardiovascular disease, metabolic dysfunction, and cognitive decline [3,16]. Extreme responses of less than 3 hours or more than 18 hours were excluded from the study and responses that were not known or unwilling to answer were set as missing [16].
On the other hand, a GWAS analysis of insomnia by Jansen et al (2019) identified novel risk loci and functional pathways, with original data from UK Biobank and 23andMe, both prospective cohort studies [17]. By combining data collected by UK Biobank version 2 (n = 386,533) and 23andMe (n = 944,477), the study achieved a sizeable sample size of 1,331,010 individuals. The study was meta-analysed for insomnia GWAS in the UK Biobank and 23andMe cohorts [17]. In the UK Biobank cohort, cases of insomnia were identified by self-report of the question: Do you have trouble falling asleep at night or waking up in the middle of the night with responses never/rarely, sometimes, usually, prefer not to answer. Insomnia was defined as the ‘usually’ response to this question. In the 23andMe cohort, the classification of individuals as insomnia cases or controls was assessed by scoring responses to seven conceptual sleep-related questionnaires [17]. (see Table 1 for basic information on SNPs)
[Figure omitted. See PDF.]
2.2.2 Cognitive function and four types of dementia.
The present study presents a comprehensive overview of the cognitive function tests utilized, as outlined in Table 2. These assessments encompass a spectrum of cognitive domains, including but not limited to the cognitive performance, FIS, memory performance, TM test, SDS test, PM test, and RT [18–28]. Each test comprises distinct subtests, scoring mechanisms, and participant demographics, all of which are meticulously delineated in Table 2 for clarity. This study also used four types of dementia from AD Neuroimaging Initiative dataset and Finnish GWAS database: AD (5,918 cases: 212,874 controls); Lewy body dementia (2,591 cases: 4,027 controls); vascular dementia (194 cases: 211,300 controls); and frontotemporal dementia (515 cases: 2,509 controls). For more detailed information refer to S1 Table.
[Figure omitted. See PDF.]
2.3 Instrumental variable selection
The flowchart of the study is shown in Fig 2. Briefly, sleep phenotype serves as the exposure and cognitive function serves as the outcome.
[Figure omitted. See PDF.]
Initially, SNPs were derived from aggregated GWAS datasets pertaining to sleep phenotypes. SNPs surpassing the genome-wide statistical significance threshold (p < 5 × 10−8) were identified as IVs and subjected to analysis, incorporating measures to mitigate potential biases associated with linkage disequilibrium. Specifically, SNPs were pruned based on a stringent threshold of r^2 < 0.001 and a genetic distance of 10,000 kb, ensuring their independence and reliability as IVs. Subsequently, data from both sleep phenotype and cognitive function GWAS databases were retrieved, harmonized, and merged to ensure coherence in the effect alleles pertaining to both phenotypes. Palindromic SNPs, characterized by allele pairs such as A/T or G/C, were systematically excluded from the analysis due to their propensity to introduce inconsistencies in strand orientation and allele coding, thereby potentially confounding the interpretation of effect alleles across datasets. To maintain consistency in strand orientation, only IVs derived from the same DNA strand orientation were retained for further analysis. Moreover, to assess potential associations of SNPs associated with sleep phenotypes with other genomic traits at the genome-wide level, a thorough search was conducted within the PhenoScanner database, with a particular emphasis on excluding potential confounders influencing cognitive outcomes. Following this rigorous selection process, the identified SNPs were utilized as IVs in subsequent two-sample MR analyses. Finally, to evaluate the likelihood of weak instrumental bias, F statistics were computed, with a value exceeding 10 indicative of a reduced probability of weak instrumental bias [29].
2.4 Statistical analyses
2.4.1. Primary analysis.
The primary analyses were conducted to investigate the causal relationships between sleep phenotypes and cognitive function through MR approaches. When dealing with phenotypes characterized by a IV, the Wald ratio test was employed to gauge the correlation between the identified IV and each cognitive function [30]. Conversely, for phenotypes incorporating multiple IVs, the Inverse-Variance Weighted (IVW) test was utilized [31]. The IVW method stands as the predominant approach for computing the weighted mean of all IV effect values [31]. This method operates under the assumption that all genetic variants utilized as IVs are valid and are solely linked to cognitive function through the intermediary of the sleep phenotypes, namely, long sleep duration, short sleep duration, and insomnia. Assuming the absence of horizontal pleiotropy, IVW is presumed to furnish unbiased estimates of causal effects. The crux of IVW lies in assigning greater weight to instruments with smaller standard errors, thereby prioritizing more precise estimates.
2.4.2 Sensitivity analysis.
Numerous sensitivity analyses were conducted to evaluate the robustness of the findings. Initially, two additional methods, namely MR-Egger regression and the weighted median estimator (WME) methods, were utilized in conjunction with the IVW method to refine the estimation outcomes and validate the consistency of MR results. MR-Egger regression operates under the assumption that the strength of the IVs (i.e., the association between genetic variants and the exposure) is independent of their association with the outcome [32]. Conversely, the weighted median estimator assumes that at least 50% of the analytical weight is derived from valid instruments unaffected by horizontal pleiotropy [33]. Incorporating MR-Egger regression and WME introduces flexibility by relaxing the assumption of absent horizontal pleiotropy in MR. Consequently, these approaches offer enhanced robustness against violations of the assumption that all genetic instruments remain unaffected by pathways other than the sleep phenotype, thereby bolstering confidence in the estimator when consistent results align with the IVW method. Subsequently, a funnel plot was constructed to preliminarily assess heterogeneity; deviation from a funnel-shaped distribution might indicate its presence. Cochran’s Q values were then computed to quantitatively evaluate SNP heterogeneity, with heterogeneity deemed present if the associated P-value was less than 0.05 [34,35]. The presence of heterogeneity does not necessarily contravene the fundamental assumptions of MR. In instances where heterogeneity was evident, the random effects IVW method was employed for the MR analysis. Furthermore, a horizontal pleiotropy test was conducted to validate the reliability of the MR analysis outcomes [36], often gauged through the intercept term of the MR-Egger test. An intercept term exceeding P > 0.05 suggests an absence of horizontal pleiotropy; conversely, the presence of horizontal pleiotropy undermines the reliability of the MR analysis results. Additionally, the MR Pleiotropy RESidual Sum and Outlier (MR-PRESSO) test was utilized to identify and adjust for outliers attributable to horizontal pleiotropy using the Outlier-corrected method within the model. Subsequent MR analyses were reiterated after excluding known pleiotropic SNPs. Lastly, leave-one-out analyses were conducted to ascertain whether a single SNP drove the causal effect, thereby further fortifying the robustness of the findings [37].
Statistical analyses were conducted using the TwoSampleMR package (version 0.5.7) within the R environment (version 4.2.2) [38,39]. In order to enhance the robustness of the findings, a Bonferroni correction was implemented to establish the threshold of significance (0.05/45 MR estimates = 0.001). Consequently, outcomes with a p-value <0.001 were deemed significant, while those falling within the range of 0.001 < p < 0.05 were considered suggestive; the latter category denotes results that did not attain the Bonferroni-corrected significance threshold but did achieve conventional significance at the 0.05 level.
3 Results
3.1 Instrumental variable selection
The F-statistics for the association of genetic instrumentation with self-reported short sleep duration, insomnia, and long sleep duration ranged from 26 to 67, suggesting that there was little evidence to suggest weak instrument bias in this study. Further screening was carried out according to the IV screening conditions and the final results are presented in S1 Table. According to the ME-Egger intercept test it was stated that the data were not horizontal pleiotropy (p > 0.05) and could be analysed for MR (see S1 Table).
3.2 Primary analysis
The findings suggest that short sleep duration may adversely affect multiple domains of cognitive function. MR analysis using an IVW method showed that short sleep duration showed a harmful impact on cognitive performance (beta = –0.15, 95% CI: –0.27 to –0.02, P = 0.02), which serve as a potential test for identifying people in the community with undiagnosed dementia, and their decline reflects a significant reduction in overall cognitive function. Similarly, there was a strong negative correlation between short sleep duration and FIS (beta = –0.38; 95% CI: –0.65 to –0.11; P = 0.006, IVW). The FIS is a physiologically based measure of cognitive ability related to arithmetic speed and reasoning. In terms of memory function, short sleep duration was linked to poorer memory performance (beta = –0.10, 95% CI: –0.20 to –0.0005; P = 0.04, IVW), further underscoring the broad cognitive impact of inadequate sleep (Fig 3).
[Figure omitted. See PDF.]
Beta with the 95% confidence intervals reveal the association estimates with the risk of cognitive decline: self-reported short sleep duration (<7 vs. 7–8 h/day). TM:Trail Making; SDS:Symbol Digit Substitution; PM:Pair Matching; 95%CI:95% Confidence Interval.
As shown in Fig 3, performance on the TM test, which assesses executive function and the ability to switch attention, was also negatively affected. Specifically, short sleep duration was associated with a longer interval between steps in the Trail 2 path (beta = 0.11, 95% CI: 0.01 to 0.21, P = 0.03, IVW) and a longer total duration to complete the Trail 2 path (beta = 0.11, 95% CI: 0.002 to 0.22, P = 0.04, IVW), suggesting impairments in attention and task-switching capabilities.
There was insufficient evidence to support a causal relationship between insomnia and performance on cognitive tests, including FIS (Fig 4). In addition, insomnia was causally associated with AD, with a significant odds ratio of 1.13 (95% CI: 1.02 to 1.24, P = 0.02, IVW) (Fig 6). However, there was insufficient evidence to support a causal relationship between long sleep duration and every cognitive-related outcome (P > 0.05 for all) (Fig 5 and Fig 6). Considering the limited number of SNPs, the causal relationship between long sleep duration and cognitive function still needs further investigation.
[Figure omitted. See PDF.]
Beta with the 95% confidence intervals reveal the association estimates with the risk of cognitive decline: self-reported insomnia. TM:Trail Making; SDS:Symbol Digit Substitution; PM:Pair Matching; 95%CI:95% Confidence Interval.
[Figure omitted. See PDF.]
Beta with the 95% confidence intervals reveal the association estimates with the risk of cognitive decline: self-reported long sleep duration (≥9 vs. 7–8 h/day). TM:Trail Making; SDS:Symbol Digit Substitution; PM:Pair Matching; 95%CI:95% Confidence Interval.
[Figure omitted. See PDF.]
OR with the 95% confidence intervals reveal the association estimates with the risk of cognitive decline, respectively: self-reported short sleep duration (<7 vs. 7–8 h/day); self-reported long sleep duration (≥9 vs. 7–8 h/day); self-reported insomnia. TM:Trail Making; SDS:Symbol Digit Substitution; PM:Pair Matching; OR:Odds ratios; 95%CI:95% Confidence Interval.
3.3 Sensitivity analyses
The MR-Egger and WME approaches complement each other, aligning with the primary IVW methods and thereby enhancing confidence in the findings (refer to S1 Fig). Indications of heterogeneity in cognitive function analyses emerge from funnel plots and Cochran’s Q tests results, as detailed in S1 Fig and S1 Table. According to the MR-Egger intercept test, there was no evidence of horizontal pleiotropy (all P > 0.05), indicating that the data were suitable for MR analysis (refer to S1 Table for details). As shown in S1 Table, the results of the MR analyses using the outlier-corrected method in MR-PRESSO were compared with those of the original MR analyses, revealing a slightly strengthened causal relationship between short sleep duration and cognitive function, as measured by cognitive performance and FIS. The leave-one-out method did not identify SNPs with a high impact on causal associations (refer to S1 Fig).
4 Discussion
Although there have been previous MR studies on sleep duration and cognitive function, the cognitive outcomes assessed in this study were much more comprehensive, encompassing not only four different types of dementia but also seven cognitive tests. This two-sample MR analysis provides causal evidence indicating that genetically determined self-reported short sleep duration is positively associated with an increased risk of cognitive decline. Furthermore, the findings support an association between genetically determined self-reported insomnia and a higher risk of AD. The evidence of causality between long sleep duration and cognitive function remains unclear and requires further investigation.
The results of this study are consistent with previous observational studies of healthy older adults. These studies have found an association between short sleep duration and insomnia and an increased risk of cognitive impairment. Using the English Longitudinal Study of Ageing (ELSA) cohort and the China Health and Retirement Longitudinal Study (CHARLS), Ma Y et al. found that, after adjusting for multiple covariates, overall cognitive scores declined more rapidly among elderly individuals who slept ≤4 hours per night compared with those who reported sleeping 7 hours per night, over 100,000 person-years of follow-up [40]. A meta-analysis that included nine cohorts with a total of 22,187 participants showed that short and long sleep durations increased the risk of cognitive impairment by 34% and 21%, respectively, and that the association between short sleep duration and cognitive impairment was stronger [41]. Yaffe K et al. reported in an 8-year follow-up study of 179,738 male veterans (aged 55 years and older) that individuals with insomnia had a 26% increased risk of developing AD compared with those without insomnia (RR = 1. 26), but did not have a significantly increased risk of vascular dementia or Lewy body dementia [42]. Similarly, Shi L et al. found that insomnia increased the risk of AD but not vascular dementia or all-cause dementia [43].
Reduced sleep duration and insomnia can affect the onset and progression of cognitive impairment and dementia through amyloid β-protein (Aβ) dynamics. Deposition of Aβ is considered a hallmark pathological feature of AD [44]. On the one hand, neuronal firing promotes the rate of Aβ production, which is reduced during slow-wave sleep, resulting in slower or suppressed Aβ production. Reduced sleep following waking leads to increased neuronal activity and frequent discharges that promote Aβ production and accumulation [44,45]. On the other hand, the brain removes metabolic wastes like Aβ more efficiently during slow-wave sleep. Chronic sleep restriction or deprivation can lead to reduced clearance of such harmful substances from the cerebrospinal fluid, thereby increasing Aβ deposition and the formation of amyloid plaques [45,46]. Once amyloid plaques form in the brain they can disrupt sleep-wake function and circadian rhythms. This disruption affects the positive feedback loop that results in poor sleep, which in turn accelerates amyloid deposition and affects sleep-promoting brain regions, creating a vicious cycle of sleep disruption. Animal studies suggest that the rate of Aβ clearance during sleep is twice as high as during wakefulness and that Aβ is significantly deposited during periods of sleep deprivation [47]. Beyond sleep duration, growing evidence suggests that sleep quality plays a critical role in brain health through its effects on neuroinflammation and glymphatic clearance. Disrupted or fragmented sleep has been shown to increase systemic and central nervous system levels of pro-inflammatory cytokines, such as interleukin-6 (IL-6) and tumor necrosis factor-alpha (TNF-α), which can exacerbate neuroinflammatory processes and contribute to neurodegeneration [48]. Inadequate or poor-quality sleep may also impair the function of the glymphatic system—a waste clearance pathway active during sleep—thereby reducing the removal efficiency of amyloid-β and other neurotoxic proteins from the brain [47,49]. These biological mechanisms highlight the importance of considering both the duration and quality of sleep when evaluating its causal role in cognitive aging and neurodegenerative disease.
Considering the limited number of SNPs (n = 7) available as instruments for long sleep duration, there was no strong evidence to support a causal effect of long sleep duration on cognitive outcomes. Long sleep duration is considered a marker of poor health in older adults. A cross-sectional study that included 4417 cognitively normal older adults aged 65–85 years found no significant difference in Aβ loads between long and normal sleep duration [50]. Longer sleep duration was associated with many other phenotypes associated with aging (e.g., poorer executive functioning, poorer subjective cognition, higher GDS scores, higher BMI), highlighting that the detrimental effects of long sleep duration are generally linked to the aging process itself, rather than to early Aβ pathology [8]. However, some observational studies have found different results. Some studies have suggested a U-shaped relationship between self-reported sleep duration and cognition [51–53]. The results of Li et al.‘s study showed that both insufficient and excessive sleep duration were positively associated with poorer performance on subsequent cognitive tasks in the middle-aged and older population [54]. There are two main differences between the Li et al. study and the present analysis. First, Li et al.’s research was based on a prospective cohort design and may be subject to residual confounding. Second, Li et al.’s study differed in its definition of long and short sleep duration. Li et al. dichotomised sleep duration into two categories (≤7 h/24 h and >7 h/24 h). In contrast, the present study defined short sleep duration as 6 hours or less per 24 hours and long sleep duration as 9 hours or more, using 7–8 hours as the reference [54]. A recent longitudinal cohort study enrolling 5,115 participants aged 18–30 years, with sleep duration measured subjectively and objectively using actigraphy, aimed to examine the relationship between sleep duration and quality in early midlife and performance in several cognitive domains in late midlife [55]. This study suggests that the association between sleep quality and cognition may become salient as early as midlife and suggests that sleep quality, rather than quantity, is particularly important for midlife cognitive health [55]. The difference between actigraphy-based and self-reported sleep measures was found to be approximately 0.8 hours [56]. Such methodological differences in sleep assessment may partially account for the inconsistencies across studies.
There have been some previous MR studies on sleep duration and cognitive function. Henry et al. examined the potential effects of sleep duration on cognitive function using pooled statistics from the UK Biobank cohort and the International Alzheimer’s Disease Genomics Project [57]. Results from linear MR suggest that a linear increase in sleep duration has a slight negative effect on reaction time and visual memory, but the true association may be nonlinear [57]. The MR study by Wang et al. explored the potential causal relationship between sleep characteristics including sleep duration, insomnia and cognitive impairment [58]. This study reported a negative correlation between self-reported sleep duration and reaction time, suggesting an adverse effect of short sleep duration on cognitive performance. This study did not find sufficient evidence to support a causal relationship between insomnia and cognitive function [58]. Qiu et al. combined data from the National Health and Nutrition Examination Survey (NHANES) and MR analyses to report an inverted U-shaped relationship between sleep duration and cognitive function in older adults in a cross-sectional study, but the MR analyses did not find a causal effect of sleep duration and insomnia on cognitive function [59]. There have also been some MR analyses of the relationship between sleep duration, insomnia, and dementia. MR analyses in the study by Yuan et al. did not find a statistically significant causal relationship between sleep duration and AD, although prospective observational studies in that study suggested that excessive sleep duration may increase the risk of AD [60]. Similarly, Huang et al. did not find any evidence to support a causal relationship between insomnia and AD risk [61].
Five previous MR studies have brought enlightening evidence [57–61]. Based on previous MR analyses, the present study further explored the causal relationship between sleep duration and cognitive functioning by further categorising sleep duration: long sleep duration and short sleep duration. In addition, the present study used insomnia GWAS data based on a larger sample size for exploration. A review of previous MR analyses on insomnia and cognitive functioning found that the study by Wang et al. used insomnia GWAS data from 386,533 individuals of European ancestry recruited from 22 assessment centres in the UK [58]; the study by Huang et al. included insomnia GWAS data from 237,627 individuals of European ancestry in the UK [61]. The present study used combined data collected by UK Biobank (n = 386,533 individuals) and 23andMe (n = 944,477 individuals) for a final sample size of up to 1,331,010 individuals of European ancestry. Insomnia studies with different sample sizes and study sites may yield different conclusions. On the other hand, Qiu et al.‘s MR analyses considered only one test of cognitive function [59]. Henry et al. judged cognitive function only by visual memory tests, reaction time and dementia or not [57]. However, the manifestations of cognitive decline are complex. It is difficult to fully reflect changes in cognitive functioning and the presence of dementia using only one or two tests. Compared with the study of Henry et al, Wang et al included more cognitive tests such as cognitive performance, fluid intelligence [58]. Based on previous studies, the present study added more cognitive function tests (seven cognitive tests including memory test, reaction time, cognitive performance and fluid intelligence). At the same time, previous MR studies did not provide more evidence about sleep characteristics and dementia subtypes. Qiu et al.’s MR analysis did not consider any dementia and its subtypes [59]; Henry et al., Yuan et al. and Huang et al. all focused only on AD, the most common form of dementia [57,60,61]; and Wang et al.’s focus fell on AD-related progression scores and changes in brain structure [58].
In the current study, the genetic association analyses yielded some interesting results. Short sleep duration seems to be more strongly associated with certain cognitive indicators, such as fluid intelligence, than insomnia. This may be due to differences in the definitions of short sleep duration and insomnia. In this study, short sleep duration was defined as 6 hours or less of sleep per 24 hours. Shorter sleep duration implies a reduction in slow wave sleep (SWS) and rapid eye movement (REM) sleep. These two sleep stages play different roles in cognitive functioning. SWS, characterised by high-amplitude, low-frequency brain waves, is essential for the hippocampal-neocortical dialogue during memory consolidation [62]. Diminished SWS impairs the reactivation of newly acquired information from the hippocampus and its transfer to long-term storage in the cortex [62]. Thus, SWS is critically associated with declarative memory consolidation and fact-based situational memory [62]. It has been shown that SWS disruption is associated with impaired situational memory and slower information processing, especially after accounting for demographic and lifestyle factors [62]. In contrast, REM sleep is known to promote synaptic plasticity and activate neural circuits involving the prefrontal cortex and striatum to support emotion regulation, attention, and executive function. REM sleep disruption has been shown to impair sustained attention, working memory, and cognitive flexibility, functions that are critically dependent on the integrity of frontal-striatal circuits [63]. The UK Biobank assessment reflects fluid intelligence – which encompasses the ability to process and apply new information, reason abstractly, and perform goal-directed behaviours effectively. Therefore, the observed association between short sleep duration and reduced fluid intelligence scores may reflect cumulative deficits caused by reduced SWS, which affects memory encoding and consolidation, and reduced REM sleep, which affects attention and executive functioning. Despite the small amount of effect observed, MR analyses support that short sleep duration performs equally poorly on other cognitive tests. Specifically, genetically predicted short sleep duration was causally associated with reduced memory performance (beta = –0.10, 95% CI: –0.20 to –0.0005, P = 0.04). This result further highlights the detrimental effects of reduced sleep duration on memory encoding and recall. In addition, impaired executive functioning due to short sleep duration was also reflected in poorer performance on the TM test. Individuals with genetically predicted short sleep duration exhibited longer step intervals (beta = 0.11, 95% CI: 0.01 to 0.21, P = 0.03) and longer total task duration (beta = 0.11, 95% CI: 0.002 to 0.22, P = 0.04), indicating impaired attentional switching and cognitive flexibility. Considering accumulation over time, even mild cognitive decline (e.g., 0.2–0.3 standard deviations) may predict increased risk of functional decline and dementia in later life [52,54,63].
Insomnia is defined as difficulty falling asleep at night or waking up in the middle of the night. Thus, insomniacs may still obtain a normal total sleep duration, albeit with poorer sleep quality. Alternatively, insomniacs may be more conscious of this compensation through planned daytime naps, but chronically sleep-deprived individuals tend to underestimate their cognitive deficits. During planned daytime naps, individuals may not only compensate for REM sleep, but also compensate for reduced SWS by prolonging REM sleep, thus demonstrating resilience to SWS deficits. From this perspective, some of the functions of SWS sleep may be redistributed during REM sleep stages or compensated to some extent by REM sleep-related processes [63]. Second, sleep duration in this study was supported by accelerometer estimates. Insomnia relies on subjective self-reports, and subjective symptoms are influenced by mood, anxiety, and personal distress, which complicates its connection to objective cognition. In contrast to disease diagnosis, fluid intelligence emphasises speed of information processing, ability to reason abstractly and executive function. Thus, fluid intelligence scores better capture the changes in cognitive functioning that occur as a result of reduced SWS and REM due to short sleep duration.
The present study suggests a close causal link between insomnia and AD, rather than cognitive test scores. There are several plausible explanations for this dissociation. Firstly, cognitive assessment methods provided by the UK Biobank may not readily detect cognitive decline in the preclinical stages of AD. These assessment methods, although suitable for large-scale population studies, may lack the sensitivity to capture subtle or domain-specific cognitive deficits, such as language deficits, disorientation disorders, which are often among the areas susceptible to AD [24,40]. Thus, individuals with early neurodegenerative conditions may still perform normally on these relatively crude tests. Secondly, insomnia may increase long-term risk for AD through pathways that do not lead to immediately measurable cognitive deficits. For example, insomnia is associated with impaired clearance of amyloid β-protein and tau protein by the lymphatic system, systemic inflammation, and neuroendocrine dysregulation, all of which are implicated in the pathogenesis of AD but may precede observable changes in cognitive performance by years or even decades [44,46,49]. Thirdly, individuals with chronic insomnia may also develop compensatory mechanisms or cognitive reserves that temporarily buffer measurable performance deficits, thereby delaying the emergence of impairments on cognitive tests [64]. Finally, in the GWAS included in this study, the diagnosis of AD captures a clinical endpoint, whereas the cognitive performance indicators in the UK Biobank reflect subclinical characteristics. The lack of association between insomnia and cognitive test scores does not necessarily imply a lack of cognitive impact, but should reflect the complexity and time lag between sleep disorders and cognitive decline.
In light of these considerations, it is also important to examine whether structural brain changes—such as brain atrophy—may mediate the relationship between sleep disturbances and cognitive decline. Higher rates of brain atrophy are considered a marker of declining brain health [65–68]. Brain atrophy rates increase with normal aging [65], cognitive decline [66], AD [67], and cardiovascular risk factors [68]. Fjell et al. analysed several large-scale longitudinal MRI datasets and found no evidence of an association between sleep duration and brain atrophy [64]. This result suggests that cognitive decline may not be mediated by sleep duration through increased brain atrophy. However, Fjell et al. noted that the analysis was based on cross-sectional data of brain morphometry and that many covariates potentially influencing the sleep–brain relationship—such as cardiovascular risk factors—were not controlled for [64]. In addition, the study assessed only brain morphometry. Other imaging measures may better capture the relationship between sleep duration and cognitive function, such as hippocampal volume [69] or Aβ accumulation [70]. Another cross-sectional study by Fjell et al. analysing data from 21,000 participants in the UK Biobank, found that sleep durations between 5 and 9 hours per 24-hour period were not associated with smaller hippocampal volumes, whereas shorter and longer sleep durations were [71]. More MR studies on sleep and brain structure, especially the rate of brain atrophy, may compensate for the shortcomings of cross-sectional analyses, thus digging deeper into the intrinsic mechanisms of the relationship between sleep and cognitive function.
This study has several strengths. First, by leveraging the random allocation of alleles at conception, MR analyses reduce bias from residual confounding and reverse causation. SNPs were used as IVs to estimate causal relationships between self-reported sleep phenotypes and cognitive outcomes, under the standard MR assumptions of relevance, independence, and exclusion restriction. Second, most previous studies focused solely on AD, this study included a broader range of dementia types and cognitive function outcomes, including CPT, FI, memory performance, TM test, SDS test, PM test, and RT. In addition, all participants in all GWAS datasets were of European origin, helping to reduce bias due to population stratification. Finally, the three MR methods help to ensure consistency and precision of results. At the same time, several limitations should be acknowledged. First, self-reported sleep duration or insomnia symptoms may introduce a potential misclassification bias due to memory and reporting errors, which can lead to biased MR analysis estimates towards zero values and underestimation of true causal effects. This bias mainly occurs in cases where there is no difference in measurement error. In this case, the noise introduced by self-reporting weakens the observed association between the genetic IV and the exposure, thereby reducing the strength of the IV. This dilution effect reduces the proportion of exposure variance explained by the IVs, leading to regression dilution bias. As a result, the estimated causal effect of sleep on cognitive or neurological outcomes is attenuated, making it appear weaker or even zero, despite the fact that there may be a genuine underlying association. This issue is particularly important for subjective sleep phenotypes such as insomnia, as individual perceptions can vary widely depending on factors such as mood, stress, and attentional biases. In addition, an increase in exposure misclassification reduces statistical efficacy and widens standard errors, thereby increasing the likelihood of Type II errors. The absence of a flawless method for measuring sleep duration without disrupting daily routines underscores the challenge in this domain [56]. Despite the often robust correlation between actigraphy and polysomnography findings [72], actigraphy tends to overestimate sleep duration, particularly beyond the confines of sleep laboratory environments [72–76]. Notably, genetic correlations with sleep duration appear to be consistent across self-reported and accelerometer-based measures [16], which provides some robustness to the use of self-reported data. In addition, international sleep guidelines continue to rely on self-reported sleep duration to formulate public health recommendations [2,77], which supports their relevance in both clinical and epidemiological contexts. Nonetheless, future research should prioritise the integration of objective sleep measures (e.g., somatokinetic recorders and polysomnography) into large-scale cohort studies to improve measurement accuracy and reduce potential bias in causal inference. Second, the limited number of SNPs used as IVs for long sleep duration likely constrained the statistical power of the MR analysis. Statistical power in MR is directly influenced by the strength and number of genetic instruments, as well as the proportion of variance in the exposure they explain [15]. Although the use of stringent selection criteria ensured strong associations between the selected SNPs and sleep duration—thereby reducing the risk of bias from weak instruments and facilitating the identification of independent variants—low SNP count can still result in limited explanatory variance. This, in turn, may lead to imprecise causal estimates [78]. Such limitations reduce the likelihood of detecting true causal effects and may produce wider confidence intervals and attenuation of effect sizes toward the null. Consequently, the null findings for long sleep duration should be interpreted with caution, as they may reflect limited instrument strength rather than a true absence of causal association. Third, the MR analyses were restricted to individuals of European ancestry, limiting the generalisability of the findings to other populations. This is a notable limitation given the significant differences in sleep patterns, prevalence of sleep disorders, and dementia risk profiles across ethnic groups. For example, research suggests that individuals of African, Asian, and Hispanic ancestry may have different circadian rhythms, sleep duration distributions, and susceptibility to sleep-related health outcomes compared with individuals of European ancestry [79]. Similarly, genetic architecture (including allele frequencies and patterns of linkage disequilibrium) may differ significantly across populations, which may alter the strength or direction of causal relationships [80]. If causal pathways vary by descent, clinical recommendations based on European data alone may not be used in all populations. Future studies should consider validation in cohorts of diverse ancestry. Fourth, horizontal pleiotropy cannot be completely ruled out. In order to increase the robustness of the findings as well as to reduce the potential impact of horizontal pleiotropy, this study supplemented the main IVW analysis with MR-Egger and WME methods. The WME method provides valid causal estimates even when up to 50% of the IVs are invalid. The convergence of the results derived from the three methods enhances confidence in the validity of the findings and reduces the likelihood that the observed associations are attributable to pleiotropic bias. In addition, MR-Egger regression can detect directional pleiotropy through its intercept term, which reflects the average pleiotropic effect across IVs. The results suggest that the MR-Egger intercept is not statistically significant, suggesting limited evidence of horizontal pleiotropy. In addition, the MR-PRESSO detects potential pleiotropy by identifying and correcting for anomalous variation that may overly affect the results. The absence of significant outliers or large distortions in the MR-PRESSO supports the reliability of the estimates. In addition, there may be residual effects and vertical pleiotropy in this study; in particular, insomnia may overlap phenotypically with self-reported short sleep duration. Finally, the MR study relied on GWAS data, which provide valuable but partial insights into the heritability of complex traits such as sleep duration. The polygenic structure of cognitive outcomes means that even large-scale GWAS can only capture part of the overall phenotypic variation [28,81]. Thus, the genetic tools used in MR studies explain only a limited portion of the complex causal pathways that influence cognition and dementia risk. An individual’s genetic risk may only manifest itself in a particular context. For example, individuals with the same risk genes perform differently in cognitive development, in part due to differences in education or nutrition. MR analyses typically assume homogenous genetic effects, which may underestimate the true complexity of gene-environment interactions. Nonetheless, despite these limitations, MR offers a powerful tool for assessing causality in observational research. By leveraging genetic instruments, MR enables causal inference even in the face of incomplete heritability information. Taken together, these findings provide important evidence for the causal role of sleep traits in cognitive decline and dementia, while also highlighting the need for cautious interpretation and further validation in diverse populations and methodological frameworks.
5 Conclusions
In this systematic MR study, self-reported short sleep duration was causally associated with cognitive decline and self-reported insomnia was associated with increased risk of AD. However, due to the limited number of SNPs available as instruments for long sleep duration, the evidence for a causal relationship between long sleep duration and cognitive function remains inconclusive and warrants further investigation.
Supporting information
S1 Table. Detailed information for cognitive function tests and dementia-related outcomes.
SNPs analysis of short sleep duration. SNPs analysis of insomnia. SNPs analysis of long sleep duration. MR-Egger test for MR analysis of sleep phenotypes and cognitive functions. Cochran’s Q statistics test for MR analysis of sleep phenotypes and cognitive functions. The results of MR analyses using the Outlier-corrected method in MR-PRESSO with the results of the original MR analyses. The no. of SNPs determined in this MR study.
https://doi.org/10.1371/journal.pone.0330782.s001
(DOCX)
S1 Fig. Scatterplot of Mendelian randomisation analysis.
Funnel plots for MR analysis of sleep phenotypes and cognitive functions. Leave-one-out analysis for MR analysis of sleep phenotypes and cognitive functions.
https://doi.org/10.1371/journal.pone.0330782.s002
(DOCX)
Acknowledgments
Thanks to the investigators who performed GWAS analyses and summarised the data available in this study.
References
1. 1. Morse AM, Kothare SV. Sleep disorders and concussion. Handb Clin Neurol. 2018;158:127–34. pmid:30482340
* View Article
* PubMed/NCBI
* Google Scholar
2. 2. Consensus Conference Panel, Watson NF, Badr MS, Belenky G, Bliwise DL, Buxton OM, et al. Joint Consensus Statement of the American Academy of Sleep Medicine and Sleep Research Society on the Recommended Amount of Sleep for a Healthy Adult: Methodology and Discussion. J Clin Sleep Med. 2015;11(8):931–52. pmid:26235159
* View Article
* PubMed/NCBI
* Google Scholar
3. 3. Hirshkowitz M, Whiton K, Albert SM, Alessi C, Bruni O, DonCarlos L, et al. National Sleep Foundation’s sleep time duration recommendations: methodology and results summary. Sleep Health. 2015;1(1):40–3. pmid:29073412
* View Article
* PubMed/NCBI
* Google Scholar
4. 4. Increase the proportion of adults who get enough sleep. Health.gov. 2022. https://odphp.health.gov/healthypeople/objectives-and-data/browse-objectives/sleep/increase-proportion-adults-who-get-enough-sleep-sh-03/data
5. 5. St-Onge M-P, Grandner MA, Brown D, Conroy MB, Jean-Louis G, Coons M, et al. Sleep Duration and Quality: Impact on Lifestyle Behaviors and Cardiometabolic Health: A Scientific Statement From the American Heart Association. Circulation. 2016;134(18):e367–86. pmid:27647451
* View Article
* PubMed/NCBI
* Google Scholar
6. 6. Kyle SD, Siriwardena AN, Espie CA, Yang Y, Petrou S, Ogburn E, et al. Clinical and cost-effectiveness of nurse-delivered sleep restriction therapy for insomnia in primary care (HABIT): a pragmatic, superiority, open-label, randomised controlled trial. Lancet. 2023;402(10406):975–87. pmid:37573859
* View Article
* PubMed/NCBI
* Google Scholar
7. 7. Huyett P, Bhattacharyya N. Incremental health care utilization and expenditures for sleep disorders in the United States. J Clin Sleep Med. 2021;17(10):1981–6. pmid:33949943
* View Article
* PubMed/NCBI
* Google Scholar
8. 8. Yaffe K, Falvey CM, Hoang T. Connections between sleep and cognition in older adults. Lancet Neurol. 2014;13(10):1017–28. pmid:25231524
* View Article
* PubMed/NCBI
* Google Scholar
9. 9. Sabia S, Fayosse A, Dumurgier J, van Hees VT, Paquet C, Sommerlad A, et al. Association of sleep duration in middle and old age with incidence of dementia. Nat Commun. 2021;12(1):2289. pmid:33879784
* View Article
* PubMed/NCBI
* Google Scholar
10. 10. Scheltens P, De Strooper B, Kivipelto M, Holstege H, Chételat G, Teunissen CE, et al. Alzheimer’s disease. Lancet. 2021;397(10284):1577–90. pmid:33667416
* View Article
* PubMed/NCBI
* Google Scholar
11. 11. Song QF, Liu XX, Hu WN, Han XC, Zhou WH, Lu AD, et al. Night Sleep Duration and Risk of Cognitive Impairment in a Chinese Population: A Cross-sectional Study. Biomed Environ Sci. 2017;30(10):749–57. pmid:29122095
* View Article
* PubMed/NCBI
* Google Scholar
12. 12. Keil SA, Schindler AG, Wang MX, Piantino J, Silbert LC, Elliott JE, et al. Longitudinal Sleep Patterns and Cognitive Impairment in Older Adults. JAMA Netw Open. 2023;6(12):e2346006. pmid:38048131
* View Article
* PubMed/NCBI
* Google Scholar
13. 13. Lo JC, Loh KK, Zheng H, Sim SKY, Chee MWL. Sleep duration and age-related changes in brain structure and cognitive performance. Sleep. 2014;37(7):1171–8. pmid:25061245
* View Article
* PubMed/NCBI
* Google Scholar
14. 14. Benito-León J, Louis ED, Bermejo-Pareja F. Cognitive decline in short and long sleepers: a prospective population-based study (NEDICES). J Psychiatr Res. 2013;47(12):1998–2003. pmid:24094933
* View Article
* PubMed/NCBI
* Google Scholar
15. 15. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013;37(7):658–65. pmid:24114802
* View Article
* PubMed/NCBI
* Google Scholar
16. 16. 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. pmid:30846698
* View Article
* PubMed/NCBI
* Google Scholar
17. 17. 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. pmid:30804565
* View Article
* PubMed/NCBI
* Google Scholar
18. 18. Burns T, Lawler K, Lawler D, McCarten JR, Kuskowski M. Predictive Value of the Cognitive Performance Test (CPT) for Staging Function and Fitness to Drive in People With Neurocognitive Disorders. Am J Occup Ther. 2018;72(4):7204205040p1–9. pmid:29953838
* View Article
* PubMed/NCBI
* Google Scholar
19. 19. Douglas A, Letts L, Eva K, Richardson J. Use of the cognitive performance test for identifying deficits in hospitalized older adults. Rehabil Res Pract. 2012;2012:638480. pmid:22720163
* View Article
* PubMed/NCBI
* Google Scholar
20. 20. Lee JJ, Wedow R, Okbay A, Kong E, Maghzian O, Zacher M, et al. Gene discovery and polygenic prediction from a genome-wide association study of educational attainment in 1.1 million individuals. Nat Genet. 2018;50(8):1112–21. pmid:30038396
* View Article
* PubMed/NCBI
* Google Scholar
21. 21. Davies G, Marioni RE, Liewald DC, Hill WD, Hagenaars SP, Harris SE, et al. Genome-wide association study of cognitive functions and educational attainment in UK Biobank (N=112 151). Mol Psychiatry. 2016;21(6):758–67. pmid:27046643
* View Article
* PubMed/NCBI
* Google Scholar
22. 22. Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife. 2018;7:e34408. pmid:29846171
* View Article
* PubMed/NCBI
* Google Scholar
23. 23. Picciotto G, Fabio RA. Does stress induction affect cognitive performance or avoidance of cognitive effort?. Stress Health. 2023;40(1):e3280. pmid:37306658
* View Article
* PubMed/NCBI
* Google Scholar
24. 24. Llinàs-Reglà J, Vilalta-Franch J, López-Pousa S, Calvó-Perxas L, Torrents Rodas D, Garre-Olmo J. The Trail Making Test. Assessment. 2016;24(2):183–96. pmid:26318386
* View Article
* PubMed/NCBI
* Google Scholar
25. 25. Jaeger J. Digit Symbol Substitution Test: The Case for Sensitivity Over Specificity in Neuropsychological Testing. J Clin Psychopharmacol. 2018;38(5):513–9. pmid:30124583
* View Article
* PubMed/NCBI
* Google Scholar
26. 26. Curiel RE, Crocco E, Rosado M, Duara R, Greig MT, Raffo A, et al. A Brief Computerized Paired Associate Test for the Detection of Mild Cognitive Impairment in Community-Dwelling Older Adults. J Alzheimers Dis. 2016;54(2):793–9. pmid:27567839
* View Article
* PubMed/NCBI
* Google Scholar
27. 27. Hagenaars SP, Harris SE, Davies G, Hill WD, Liewald DCM, Ritchie SJ, et al. Shared genetic aetiology between cognitive functions and physical and mental health in UK Biobank (N=112 151) and 24 GWAS consortia. Mol Psychiatry. 2016;21(11):1624–32. pmid:26809841
* View Article
* PubMed/NCBI
* Google Scholar
28. 28. Davies G, Lam M, Harris SE, Trampush JW, Luciano M, Hill WD, et al. Study of 300,486 individuals identifies 148 independent genetic loci influencing general cognitive function. Nat Commun. 2018;9(1):2098. pmid:29844566
* View Article
* PubMed/NCBI
* Google Scholar
29. 29. Burgess S, Thompson SG. Bias in causal estimates from Mendelian randomization studies with weak instruments. Stat Med. 2011;30(11):1312–23. pmid:21432888
* View Article
* PubMed/NCBI
* Google Scholar
30. 30. Burgess S, Small DS, Thompson SG. A review of instrumental variable estimators for Mendelian randomization. Stat Methods Med Res. 2015;26(5):2333–55. pmid:26282889
* View Article
* PubMed/NCBI
* Google Scholar
31. 31. Jansen H, Samani NJ, Schunkert H. Mendelian randomization studies in coronary artery disease. Eur Heart J. 2014;35(29):1917–24. pmid:24917639
* View Article
* PubMed/NCBI
* Google Scholar
32. 32. 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. pmid:26050253
* View Article
* PubMed/NCBI
* Google Scholar
33. 33. 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. pmid:27061298
* View Article
* PubMed/NCBI
* Google Scholar
34. 34. Greco M FD, Minelli C, Sheehan NA, Thompson JR. Detecting pleiotropy in Mendelian randomisation studies with summary data and a continuous outcome. Stat Med. 2015;34(21):2926–40. pmid:25950993
* View Article
* PubMed/NCBI
* Google Scholar
35. 35. Bowden J, Del Greco M F, Minelli C, Zhao Q, Lawlor DA, Sheehan NA, et al. Improving the accuracy of two-sample summary-data Mendelian randomization: moving beyond the NOME assumption. Int J Epidemiol. 2018;48(3):728–42. pmid:30561657
* View Article
* PubMed/NCBI
* Google Scholar
36. 36. Carter AR, Sanderson E, Hammerton G, Richmond RC, Davey Smith G, Heron J, et al. Mendelian randomisation for mediation analysis: current methods and challenges for implementation. Eur J Epidemiol. 2021;36(5):465–78.
* View Article
* Google Scholar
37. 37. Hemani G, Tilling K, Davey Smith G. Orienting the causal relationship between imprecisely measured traits using GWAS summary data. PLoS Genet. 2017;13(11):e1007081. pmid:29149188
* View Article
* PubMed/NCBI
* Google Scholar
38. 38. Cho Y, Haycock PC, Sanderson E, Gaunt TR, Zheng J, Morris AP, et al. Exploiting horizontal pleiotropy to search for causal pathways within a Mendelian randomization framework. Nat Commun. 2020;11(1):1010. pmid:32081875
* View Article
* PubMed/NCBI
* Google Scholar
39. 39. Verbanck M, Chen C-Y, 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. pmid:29686387
* View Article
* PubMed/NCBI
* Google Scholar
40. 40. Ma Y, Liang L, Zheng F, Shi L, Zhong B, Xie W. Association Between Sleep Duration and Cognitive Decline. JAMA Netw Open. 2020;3(9):e2013573.
* View Article
* Google Scholar
41. 41. Wu L, Sun D, Tan Y. A systematic review and dose-response meta-analysis of sleep duration and the occurrence of cognitive disorders. Sleep Breath. 2017;22(3):805–14. pmid:28589251
* View Article
* PubMed/NCBI
* Google Scholar
42. 42. Yaffe K, Nettiksimmons J, Yesavage J, Byers A. Sleep Quality and Risk of Dementia Among Older Male Veterans. Am J Geriatr Psychiatry. 2015;23(6):651–4. pmid:25794635
* View Article
* PubMed/NCBI
* Google Scholar
43. 43. Shi L, Chen S-J, Ma M-Y, Bao Y-P, Han Y, Wang Y-M, et al. Sleep disturbances increase the risk of dementia: A systematic review and meta-analysis. Sleep Med Rev. 2018;40:4–16. pmid:28890168
* View Article
* PubMed/NCBI
* Google Scholar
44. 44. Sehar U, Rawat P, Reddy AP, Kopel J, Reddy PH. Amyloid Beta in Aging and Alzheimer’s Disease. Int J Mol Sci. 2022;23(21):12924. pmid:36361714
* View Article
* PubMed/NCBI
* Google Scholar
45. 45. Duarte AC, Furtado A, Hrynchak MV, Costa AR, Talhada D, Gonçalves I, et al. Age, Sex Hormones, and Circadian Rhythm Regulate the Expression of Amyloid-Beta Scavengers at the Choroid Plexus. Int J Mol Sci. 2020;21(18):6813. pmid:32957439
* View Article
* PubMed/NCBI
* Google Scholar
46. 46. Nedergaard M, Goldman SA. Glymphatic failure as a final common pathway to dementia. Science. 2020;370(6512):50–6. pmid:33004510
* View Article
* PubMed/NCBI
* Google Scholar
47. 47. Xie L, Kang H, Xu Q, Chen MJ, Liao Y, Thiyagarajan M, et al. Sleep drives metabolite clearance from the adult brain. Science. 2013;342(6156):373–7. pmid:24136970
* View Article
* PubMed/NCBI
* Google Scholar
48. 48. Irwin MR, Opp MR. Sleep Health: Reciprocal Regulation of Sleep and Innate Immunity. Neuropsychopharmacol. 2017;42(1):129–55.
* View Article
* Google Scholar
49. 49. Heneka MT, Golenbock DT, Latz E. Innate immunity in Alzheimer’s disease. Nat Immunol. 2015;16(3):229–36.
* View Article
* Google Scholar
50. 50. Winer JR, Deters KD, Kennedy G, Jin M, Goldstein-Piekarski A, Poston KL, et al. Association of Short and Long Sleep Duration With Amyloid-β Burden and Cognition in Aging. JAMA Neurol. 2021;78(10):1187–96. pmid:34459862
* View Article
* PubMed/NCBI
* Google Scholar
51. 51. Virta JJ, Heikkilä K, Perola M, Koskenvuo M, Räihä I, Rinne JO, et al. Midlife sleep characteristics associated with late life cognitive function. Sleep. 2013;36(10):1533–41, 1541A. pmid:24082313
* View Article
* PubMed/NCBI
* Google Scholar
52. 52. Devore EE, Grodstein F, Duffy JF, Stampfer MJ, Czeisler CA, Schernhammer ES. Sleep duration in midlife and later life in relation to cognition. J Am Geriatr Soc. 2014;62(6):1073–81. pmid:24786726
* View Article
* PubMed/NCBI
* Google Scholar
53. 53. Leng Y, Yaffe K. Sleep Duration and Cognitive Aging-Beyond a U-Shaped Association. JAMA Netw Open. 2020;3(9):e2014008. pmid:32955568
* View Article
* PubMed/NCBI
* Google Scholar
54. 54. Li Y, Sahakian BJ, Kang J, Langley C, Zhang W, Xie C, et al. The brain structure and genetic mechanisms underlying the nonlinear association between sleep duration, cognition and mental health. Nat Aging. 2022;2(5):425–37. pmid:37118065
* View Article
* PubMed/NCBI
* Google Scholar
55. 55. Leng Y, Knutson K, Carnethon MR, Yaffe K. Association Between Sleep Quantity and Quality in Early Adulthood With Cognitive Function in Midlife. Neurology. 2024;102(2):e208056. pmid:38170947
* View Article
* PubMed/NCBI
* Google Scholar
56. 56. 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. pmid:18854708
* View Article
* PubMed/NCBI
* Google Scholar
57. 57. Henry A, Katsoulis M, Masi S, Fatemifar G, Denaxas S, Acosta D, et al. The relationship between sleep duration, cognition and dementia: a Mendelian randomization study. Int J Epidemiol. 2019;48(3):849–60. pmid:31062029
* View Article
* PubMed/NCBI
* Google Scholar
58. 58. Wang Q, Xu S, Liu F, Liu Y, Chen K, Huang L, et al. Causal relationship between sleep traits and cognitive impairment: A Mendelian randomization study. J Evid Based Med. 2023;16(4):485–94. pmid:38108111
* View Article
* PubMed/NCBI
* Google Scholar
59. 59. Qiu P, Dong C, Li A, Xie J, Wu J. Exploring the relationship of sleep duration on cognitive function among the elderly: a combined NHANES 2011–2014 and mendelian randomization analysis. BMC Geriatr. 2024;24(1).
* View Article
* Google Scholar
60. 60. Yuan S, Ma W, Yang R, Xu F, Han D, Huang T, et al. Sleep duration, genetic susceptibility, and Alzheimer’s disease: a longitudinal UK Biobank-based study. BMC Geriatr. 2022;22(1).
* View Article
* Google Scholar
61. 61. Huang J, Zuber V, Matthews PM, Elliott P, Tzoulaki J, Dehghan A. Sleep, major depressive disorder, and Alzheimer disease. Neurology. 2020;95(14).
* View Article
* Google Scholar
62. 62. Walker MP. The Role of Slow Wave Sleep in Memory Processing. Journal of Clinical Sleep Medicine. 2009;5(2 suppl).
* View Article
* Google Scholar
63. 63. Scullin MK, Gao C. Dynamic Contributions of Slow Wave Sleep and REM Sleep to Cognitive Longevity. Curr Sleep Med Rep. 2018;4(4):284–93. pmid:31737466
* View Article
* PubMed/NCBI
* Google Scholar
64. 64. Fjell AM, Sørensen Ø, Wang Y, Amlien IK, Baaré WFC, Bartrés-Faz D, et al. No phenotypic or genotypic evidence for a link between sleep duration and brain atrophy. Nat Hum Behav. 2023;7(11):2008–22. pmid:37798367
* View Article
* PubMed/NCBI
* Google Scholar
65. 65. Walhovd KB, Krogsrud SK, Amlien IK, Bartsch H, Bjørnerud A, Due-Tønnessen P, et al. Neurodevelopmental origins of lifespan changes in brain and cognition. Proc Natl Acad Sci U S A. 2016;113(33):9357–62. pmid:27432992
* View Article
* PubMed/NCBI
* Google Scholar
66. 66. Fjell AM, Walhovd KB. Structural brain changes in aging: courses, causes and cognitive consequences. Rev Neurosci. 2010;21(3):187–221. pmid:20879692
* View Article
* PubMed/NCBI
* Google Scholar
67. 67. Fjell AM, McEvoy L, Holland D, Dale AM, Walhovd KB, Alzheimer’s Disease Neuroimaging Initiative. What is normal in normal aging? Effects of aging, amyloid and Alzheimer’s disease on the cerebral cortex and the hippocampus. Prog Neurobiol. 2014;117:20–40. pmid:24548606
* View Article
* PubMed/NCBI
* Google Scholar
68. 68. Nyberg CK, Fjell AM, Walhovd KB. Level of body fat relates to memory decline and interacts with age in its association with hippocampal and subcortical atrophy. Neurobiol Aging. 2020;91:112–24. pmid:32224068
* View Article
* PubMed/NCBI
* Google Scholar
69. 69. Halliday G. Pathology and hippocampal atrophy in Alzheimer’s disease. Lancet Neurol. 2017;16(11):862–4. pmid:29029840
* View Article
* PubMed/NCBI
* Google Scholar
70. 70. Hampel H, Hardy J, Blennow K, Chen C, Perry G, Kim SH, et al. The Amyloid-β Pathway in Alzheimer’s Disease. Mol Psychiatry. 2021;26(10):5481–503.
* View Article
* Google Scholar
71. 71. 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. 2019;43(5).
* View Article
* Google Scholar
72. 72. Ancoli-Israel S, Cole R, Alessi C, Chambers M, Moorcroft W, Pollak CP. The role of actigraphy in the study of sleep and circadian rhythms. Sleep. 2003;26(3):342–92. pmid:12749557
* View Article
* PubMed/NCBI
* Google Scholar
73. 73. de Souza L, Benedito-Silva AA, Pires MLN, Poyares D, Tufik S, Calil HM. Further validation of actigraphy for sleep studies. Sleep. 2003;26(1):81–5. pmid:12627737
* View Article
* PubMed/NCBI
* Google Scholar
74. 74. Jackson CL, Patel SR, Jackson WB 2nd, Lutsey PL, Redline S. Agreement between self-reported and objectively measured sleep duration among white, black, Hispanic, and Chinese adults in the United States: Multi-Ethnic Study of Atherosclerosis. Sleep. 2018;41(6):zsy057. pmid:29701831
* View Article
* PubMed/NCBI
* Google Scholar
75. 75. Sivertsen B, Omvik S, Havik OE, Pallesen S, Bjorvatn B, Nielsen GH, et al. A comparison of actigraphy and polysomnography in older adults treated for chronic primary insomnia. Sleep. 2006;29(10):1353–8. pmid:17068990
* View Article
* PubMed/NCBI
* Google Scholar
76. 76. Hedner J, Pillar G, Pittman SD, Zou D, Grote L, White DP. A novel adaptive wrist actigraphy algorithm for sleep-wake assessment in sleep apnea patients. Sleep. 2004;27(8):1560–6. pmid:15683148
* View Article
* PubMed/NCBI
* Google Scholar
77. 77. Hirshkowitz M, Whiton K, Albert SM, Alessi C, Bruni O, DonCarlos L, et al. National Sleep Foundation’s updated sleep duration recommendations: final report. Sleep Health. 2015;1(4):233–43. pmid:29073398
* View Article
* PubMed/NCBI
* Google Scholar
78. 78. Davies NM, Holmes MV, Davey Smith G. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. BMJ. 2018;362:k601. pmid:30002074
* View Article
* PubMed/NCBI
* Google Scholar
79. 79. Chen X, Wang R, Zee P, Lutsey PL, Javaheri S, Alcántara C, et al. Racial/Ethnic Differences in Sleep Disturbances: The Multi-Ethnic Study of Atherosclerosis (MESA). Sleep. 2015;38(6):877–88. pmid:25409106
* View Article
* PubMed/NCBI
* Google Scholar
80. 80. Mester R, Hou K, Ding Y, Meeks G, Burch KS, Bhattacharya A, et al. Impact of cross-ancestry genetic architecture on GWASs in admixed populations. Am J Hum Genet. 2023;110(6):927–39. pmid:37224807
* View Article
* PubMed/NCBI
* Google Scholar
81. 81. Savage JE, Jansen PR, Stringer S, Watanabe K, Bryois J, de Leeuw CA, et al. Genome-wide association meta-analysis in 269,867 individuals identifies new genetic and functional links to intelligence. Nat Genet. 2018;50(7):912–9. pmid:29942086
* View Article
* PubMed/NCBI
* Google Scholar
Citation: Guo Y (2025) Mendelian randomization study of self-reported long sleep duration, short sleep duration, and insomnia and cognitive function. PLoS One 20(8): e0330782. https://doi.org/10.1371/journal.pone.0330782
About the Authors:
Yunyun Guo
Roles: Conceptualization, Methodology, Writing – original draft
E-mail: [email protected]
Affiliation: Ageing Epidemiology Research Unit, School of Public Health, Imperial College London, London, United Kingdom
ORICD: https://orcid.org/0009-0009-4572-1139
[/RAW_REF_TEXT]
1. Morse AM, Kothare SV. Sleep disorders and concussion. Handb Clin Neurol. 2018;158:127–34. pmid:30482340
2. Consensus Conference Panel, Watson NF, Badr MS, Belenky G, Bliwise DL, Buxton OM, et al. Joint Consensus Statement of the American Academy of Sleep Medicine and Sleep Research Society on the Recommended Amount of Sleep for a Healthy Adult: Methodology and Discussion. J Clin Sleep Med. 2015;11(8):931–52. pmid:26235159
3. Hirshkowitz M, Whiton K, Albert SM, Alessi C, Bruni O, DonCarlos L, et al. National Sleep Foundation’s sleep time duration recommendations: methodology and results summary. Sleep Health. 2015;1(1):40–3. pmid:29073412
4. Increase the proportion of adults who get enough sleep. Health.gov. 2022. https://odphp.health.gov/healthypeople/objectives-and-data/browse-objectives/sleep/increase-proportion-adults-who-get-enough-sleep-sh-03/data
5. St-Onge M-P, Grandner MA, Brown D, Conroy MB, Jean-Louis G, Coons M, et al. Sleep Duration and Quality: Impact on Lifestyle Behaviors and Cardiometabolic Health: A Scientific Statement From the American Heart Association. Circulation. 2016;134(18):e367–86. pmid:27647451
6. Kyle SD, Siriwardena AN, Espie CA, Yang Y, Petrou S, Ogburn E, et al. Clinical and cost-effectiveness of nurse-delivered sleep restriction therapy for insomnia in primary care (HABIT): a pragmatic, superiority, open-label, randomised controlled trial. Lancet. 2023;402(10406):975–87. pmid:37573859
7. Huyett P, Bhattacharyya N. Incremental health care utilization and expenditures for sleep disorders in the United States. J Clin Sleep Med. 2021;17(10):1981–6. pmid:33949943
8. Yaffe K, Falvey CM, Hoang T. Connections between sleep and cognition in older adults. Lancet Neurol. 2014;13(10):1017–28. pmid:25231524
9. Sabia S, Fayosse A, Dumurgier J, van Hees VT, Paquet C, Sommerlad A, et al. Association of sleep duration in middle and old age with incidence of dementia. Nat Commun. 2021;12(1):2289. pmid:33879784
10. Scheltens P, De Strooper B, Kivipelto M, Holstege H, Chételat G, Teunissen CE, et al. Alzheimer’s disease. Lancet. 2021;397(10284):1577–90. pmid:33667416
11. Song QF, Liu XX, Hu WN, Han XC, Zhou WH, Lu AD, et al. Night Sleep Duration and Risk of Cognitive Impairment in a Chinese Population: A Cross-sectional Study. Biomed Environ Sci. 2017;30(10):749–57. pmid:29122095
12. Keil SA, Schindler AG, Wang MX, Piantino J, Silbert LC, Elliott JE, et al. Longitudinal Sleep Patterns and Cognitive Impairment in Older Adults. JAMA Netw Open. 2023;6(12):e2346006. pmid:38048131
13. Lo JC, Loh KK, Zheng H, Sim SKY, Chee MWL. Sleep duration and age-related changes in brain structure and cognitive performance. Sleep. 2014;37(7):1171–8. pmid:25061245
14. Benito-León J, Louis ED, Bermejo-Pareja F. Cognitive decline in short and long sleepers: a prospective population-based study (NEDICES). J Psychiatr Res. 2013;47(12):1998–2003. pmid:24094933
15. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013;37(7):658–65. pmid:24114802
16. 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. pmid:30846698
17. 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. pmid:30804565
18. Burns T, Lawler K, Lawler D, McCarten JR, Kuskowski M. Predictive Value of the Cognitive Performance Test (CPT) for Staging Function and Fitness to Drive in People With Neurocognitive Disorders. Am J Occup Ther. 2018;72(4):7204205040p1–9. pmid:29953838
19. Douglas A, Letts L, Eva K, Richardson J. Use of the cognitive performance test for identifying deficits in hospitalized older adults. Rehabil Res Pract. 2012;2012:638480. pmid:22720163
20. Lee JJ, Wedow R, Okbay A, Kong E, Maghzian O, Zacher M, et al. Gene discovery and polygenic prediction from a genome-wide association study of educational attainment in 1.1 million individuals. Nat Genet. 2018;50(8):1112–21. pmid:30038396
21. Davies G, Marioni RE, Liewald DC, Hill WD, Hagenaars SP, Harris SE, et al. Genome-wide association study of cognitive functions and educational attainment in UK Biobank (N=112 151). Mol Psychiatry. 2016;21(6):758–67. pmid:27046643
22. Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife. 2018;7:e34408. pmid:29846171
23. Picciotto G, Fabio RA. Does stress induction affect cognitive performance or avoidance of cognitive effort?. Stress Health. 2023;40(1):e3280. pmid:37306658
24. Llinàs-Reglà J, Vilalta-Franch J, López-Pousa S, Calvó-Perxas L, Torrents Rodas D, Garre-Olmo J. The Trail Making Test. Assessment. 2016;24(2):183–96. pmid:26318386
25. Jaeger J. Digit Symbol Substitution Test: The Case for Sensitivity Over Specificity in Neuropsychological Testing. J Clin Psychopharmacol. 2018;38(5):513–9. pmid:30124583
26. Curiel RE, Crocco E, Rosado M, Duara R, Greig MT, Raffo A, et al. A Brief Computerized Paired Associate Test for the Detection of Mild Cognitive Impairment in Community-Dwelling Older Adults. J Alzheimers Dis. 2016;54(2):793–9. pmid:27567839
27. Hagenaars SP, Harris SE, Davies G, Hill WD, Liewald DCM, Ritchie SJ, et al. Shared genetic aetiology between cognitive functions and physical and mental health in UK Biobank (N=112 151) and 24 GWAS consortia. Mol Psychiatry. 2016;21(11):1624–32. pmid:26809841
28. Davies G, Lam M, Harris SE, Trampush JW, Luciano M, Hill WD, et al. Study of 300,486 individuals identifies 148 independent genetic loci influencing general cognitive function. Nat Commun. 2018;9(1):2098. pmid:29844566
29. Burgess S, Thompson SG. Bias in causal estimates from Mendelian randomization studies with weak instruments. Stat Med. 2011;30(11):1312–23. pmid:21432888
30. Burgess S, Small DS, Thompson SG. A review of instrumental variable estimators for Mendelian randomization. Stat Methods Med Res. 2015;26(5):2333–55. pmid:26282889
31. Jansen H, Samani NJ, Schunkert H. Mendelian randomization studies in coronary artery disease. Eur Heart J. 2014;35(29):1917–24. pmid:24917639
32. 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. pmid:26050253
33. 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. pmid:27061298
34. Greco M FD, Minelli C, Sheehan NA, Thompson JR. Detecting pleiotropy in Mendelian randomisation studies with summary data and a continuous outcome. Stat Med. 2015;34(21):2926–40. pmid:25950993
35. Bowden J, Del Greco M F, Minelli C, Zhao Q, Lawlor DA, Sheehan NA, et al. Improving the accuracy of two-sample summary-data Mendelian randomization: moving beyond the NOME assumption. Int J Epidemiol. 2018;48(3):728–42. pmid:30561657
36. Carter AR, Sanderson E, Hammerton G, Richmond RC, Davey Smith G, Heron J, et al. Mendelian randomisation for mediation analysis: current methods and challenges for implementation. Eur J Epidemiol. 2021;36(5):465–78.
37. Hemani G, Tilling K, Davey Smith G. Orienting the causal relationship between imprecisely measured traits using GWAS summary data. PLoS Genet. 2017;13(11):e1007081. pmid:29149188
38. Cho Y, Haycock PC, Sanderson E, Gaunt TR, Zheng J, Morris AP, et al. Exploiting horizontal pleiotropy to search for causal pathways within a Mendelian randomization framework. Nat Commun. 2020;11(1):1010. pmid:32081875
39. Verbanck M, Chen C-Y, 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. pmid:29686387
40. Ma Y, Liang L, Zheng F, Shi L, Zhong B, Xie W. Association Between Sleep Duration and Cognitive Decline. JAMA Netw Open. 2020;3(9):e2013573.
41. Wu L, Sun D, Tan Y. A systematic review and dose-response meta-analysis of sleep duration and the occurrence of cognitive disorders. Sleep Breath. 2017;22(3):805–14. pmid:28589251
42. Yaffe K, Nettiksimmons J, Yesavage J, Byers A. Sleep Quality and Risk of Dementia Among Older Male Veterans. Am J Geriatr Psychiatry. 2015;23(6):651–4. pmid:25794635
43. Shi L, Chen S-J, Ma M-Y, Bao Y-P, Han Y, Wang Y-M, et al. Sleep disturbances increase the risk of dementia: A systematic review and meta-analysis. Sleep Med Rev. 2018;40:4–16. pmid:28890168
44. Sehar U, Rawat P, Reddy AP, Kopel J, Reddy PH. Amyloid Beta in Aging and Alzheimer’s Disease. Int J Mol Sci. 2022;23(21):12924. pmid:36361714
45. Duarte AC, Furtado A, Hrynchak MV, Costa AR, Talhada D, Gonçalves I, et al. Age, Sex Hormones, and Circadian Rhythm Regulate the Expression of Amyloid-Beta Scavengers at the Choroid Plexus. Int J Mol Sci. 2020;21(18):6813. pmid:32957439
46. Nedergaard M, Goldman SA. Glymphatic failure as a final common pathway to dementia. Science. 2020;370(6512):50–6. pmid:33004510
47. Xie L, Kang H, Xu Q, Chen MJ, Liao Y, Thiyagarajan M, et al. Sleep drives metabolite clearance from the adult brain. Science. 2013;342(6156):373–7. pmid:24136970
48. Irwin MR, Opp MR. Sleep Health: Reciprocal Regulation of Sleep and Innate Immunity. Neuropsychopharmacol. 2017;42(1):129–55.
49. Heneka MT, Golenbock DT, Latz E. Innate immunity in Alzheimer’s disease. Nat Immunol. 2015;16(3):229–36.
50. Winer JR, Deters KD, Kennedy G, Jin M, Goldstein-Piekarski A, Poston KL, et al. Association of Short and Long Sleep Duration With Amyloid-β Burden and Cognition in Aging. JAMA Neurol. 2021;78(10):1187–96. pmid:34459862
51. Virta JJ, Heikkilä K, Perola M, Koskenvuo M, Räihä I, Rinne JO, et al. Midlife sleep characteristics associated with late life cognitive function. Sleep. 2013;36(10):1533–41, 1541A. pmid:24082313
52. Devore EE, Grodstein F, Duffy JF, Stampfer MJ, Czeisler CA, Schernhammer ES. Sleep duration in midlife and later life in relation to cognition. J Am Geriatr Soc. 2014;62(6):1073–81. pmid:24786726
53. Leng Y, Yaffe K. Sleep Duration and Cognitive Aging-Beyond a U-Shaped Association. JAMA Netw Open. 2020;3(9):e2014008. pmid:32955568
54. Li Y, Sahakian BJ, Kang J, Langley C, Zhang W, Xie C, et al. The brain structure and genetic mechanisms underlying the nonlinear association between sleep duration, cognition and mental health. Nat Aging. 2022;2(5):425–37. pmid:37118065
55. Leng Y, Knutson K, Carnethon MR, Yaffe K. Association Between Sleep Quantity and Quality in Early Adulthood With Cognitive Function in Midlife. Neurology. 2024;102(2):e208056. pmid:38170947
56. 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. pmid:18854708
57. Henry A, Katsoulis M, Masi S, Fatemifar G, Denaxas S, Acosta D, et al. The relationship between sleep duration, cognition and dementia: a Mendelian randomization study. Int J Epidemiol. 2019;48(3):849–60. pmid:31062029
58. Wang Q, Xu S, Liu F, Liu Y, Chen K, Huang L, et al. Causal relationship between sleep traits and cognitive impairment: A Mendelian randomization study. J Evid Based Med. 2023;16(4):485–94. pmid:38108111
59. Qiu P, Dong C, Li A, Xie J, Wu J. Exploring the relationship of sleep duration on cognitive function among the elderly: a combined NHANES 2011–2014 and mendelian randomization analysis. BMC Geriatr. 2024;24(1).
60. Yuan S, Ma W, Yang R, Xu F, Han D, Huang T, et al. Sleep duration, genetic susceptibility, and Alzheimer’s disease: a longitudinal UK Biobank-based study. BMC Geriatr. 2022;22(1).
61. Huang J, Zuber V, Matthews PM, Elliott P, Tzoulaki J, Dehghan A. Sleep, major depressive disorder, and Alzheimer disease. Neurology. 2020;95(14).
62. Walker MP. The Role of Slow Wave Sleep in Memory Processing. Journal of Clinical Sleep Medicine. 2009;5(2 suppl).
63. Scullin MK, Gao C. Dynamic Contributions of Slow Wave Sleep and REM Sleep to Cognitive Longevity. Curr Sleep Med Rep. 2018;4(4):284–93. pmid:31737466
64. Fjell AM, Sørensen Ø, Wang Y, Amlien IK, Baaré WFC, Bartrés-Faz D, et al. No phenotypic or genotypic evidence for a link between sleep duration and brain atrophy. Nat Hum Behav. 2023;7(11):2008–22. pmid:37798367
65. Walhovd KB, Krogsrud SK, Amlien IK, Bartsch H, Bjørnerud A, Due-Tønnessen P, et al. Neurodevelopmental origins of lifespan changes in brain and cognition. Proc Natl Acad Sci U S A. 2016;113(33):9357–62. pmid:27432992
66. Fjell AM, Walhovd KB. Structural brain changes in aging: courses, causes and cognitive consequences. Rev Neurosci. 2010;21(3):187–221. pmid:20879692
67. Fjell AM, McEvoy L, Holland D, Dale AM, Walhovd KB, Alzheimer’s Disease Neuroimaging Initiative. What is normal in normal aging? Effects of aging, amyloid and Alzheimer’s disease on the cerebral cortex and the hippocampus. Prog Neurobiol. 2014;117:20–40. pmid:24548606
68. Nyberg CK, Fjell AM, Walhovd KB. Level of body fat relates to memory decline and interacts with age in its association with hippocampal and subcortical atrophy. Neurobiol Aging. 2020;91:112–24. pmid:32224068
69. Halliday G. Pathology and hippocampal atrophy in Alzheimer’s disease. Lancet Neurol. 2017;16(11):862–4. pmid:29029840
70. Hampel H, Hardy J, Blennow K, Chen C, Perry G, Kim SH, et al. The Amyloid-β Pathway in Alzheimer’s Disease. Mol Psychiatry. 2021;26(10):5481–503.
71. 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. 2019;43(5).
72. Ancoli-Israel S, Cole R, Alessi C, Chambers M, Moorcroft W, Pollak CP. The role of actigraphy in the study of sleep and circadian rhythms. Sleep. 2003;26(3):342–92. pmid:12749557
73. de Souza L, Benedito-Silva AA, Pires MLN, Poyares D, Tufik S, Calil HM. Further validation of actigraphy for sleep studies. Sleep. 2003;26(1):81–5. pmid:12627737
74. Jackson CL, Patel SR, Jackson WB 2nd, Lutsey PL, Redline S. Agreement between self-reported and objectively measured sleep duration among white, black, Hispanic, and Chinese adults in the United States: Multi-Ethnic Study of Atherosclerosis. Sleep. 2018;41(6):zsy057. pmid:29701831
75. Sivertsen B, Omvik S, Havik OE, Pallesen S, Bjorvatn B, Nielsen GH, et al. A comparison of actigraphy and polysomnography in older adults treated for chronic primary insomnia. Sleep. 2006;29(10):1353–8. pmid:17068990
76. Hedner J, Pillar G, Pittman SD, Zou D, Grote L, White DP. A novel adaptive wrist actigraphy algorithm for sleep-wake assessment in sleep apnea patients. Sleep. 2004;27(8):1560–6. pmid:15683148
77. Hirshkowitz M, Whiton K, Albert SM, Alessi C, Bruni O, DonCarlos L, et al. National Sleep Foundation’s updated sleep duration recommendations: final report. Sleep Health. 2015;1(4):233–43. pmid:29073398
78. Davies NM, Holmes MV, Davey Smith G. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. BMJ. 2018;362:k601. pmid:30002074
79. Chen X, Wang R, Zee P, Lutsey PL, Javaheri S, Alcántara C, et al. Racial/Ethnic Differences in Sleep Disturbances: The Multi-Ethnic Study of Atherosclerosis (MESA). Sleep. 2015;38(6):877–88. pmid:25409106
80. Mester R, Hou K, Ding Y, Meeks G, Burch KS, Bhattacharya A, et al. Impact of cross-ancestry genetic architecture on GWASs in admixed populations. Am J Hum Genet. 2023;110(6):927–39. pmid:37224807
81. Savage JE, Jansen PR, Stringer S, Watanabe K, Bryois J, de Leeuw CA, et al. Genome-wide association meta-analysis in 269,867 individuals identifies new genetic and functional links to intelligence. Nat Genet. 2018;50(7):912–9. pmid:29942086
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
© 2025 Yunyun Guo. This is an open access article distributed under the terms of the Creative Commons Attribution License: http://creativecommons.org/licenses/by/4.0/ (the “License”), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Abstract
Background and aims
Causal relationship between sleep duration and cognitive function remains unclear. This study used a two-sample Mendelian randomization (MR) analysis to assess the causal relationship between self-reported short sleep duration, insomnia and long sleep duration and cognitive function.
Methods
A total of 26 single nucleotide polymorphisms (SNPs) associated with short sleep duration, 240 associated with insomnia, and 7 associated with long sleep duration were extracted from a genome-wide association study primarily based on European ancestry, to be used as instrumental variables. Summary-level statistics were obtained from the Dementia genome-wide association studies database. MR estimation was performed using the inverse variance weighted (IVW) method as the primary method, supplemented by MR-Egger regression and weighted median estimator methods. Finally, multiple sensitivity analyses were performed to obtain robust and valid estimates.
Results
Based on IVW methods, short sleep duration showed a harmful impact on cognitive performance score (beta = –0.15, 95% CI: –0.27 to –0.02, P = 0.02, IVW), fluid intelligence score (beta = –0.38; 95% CI: –0.65 to –0.11; P = 0.006, IVW), memory performance (beta = –0.10, 95% CI: –0.20 to –0.0005; P = 0.04, IVW) and Trail Making (TM) test (TM: interval in trail 2 path, beta = 0.11, 95% CI: 0.01 to 0.21; P = 0.03, IVW; TM: duration to complete trail 2 path, beta = 0.11, 95% CI: 0.002 to 0.22; P = 0.04, IVW). In addition, insomnia was causally associated with Alzheimer’s disease (OR = 1.13, 95% CI = 1.02–1.24, p = 0.02, IVW). However, due to the limited number of SNPs (n = 7) available as instruments for long sleep duration, there was no strong evidence to support a causal effect of long sleep duration on cognitive outcomes.
Conclusions
This study suggests self-reported short sleep duration was causally associated with cognitive decline and self-reported insomnia was causally associated with increased risk of Alzheimer’s disease in individuals of European ancestry. The evidence of causality between long sleep duration and cognitive function requires further investigation. These results may have implications for public health interventions aimed at reducing the risk of cognitive decline.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
Details
