Neurobiology of Aging 106 (2021) 320.e320–320.e12 Contents lists available at ScienceDirect Neurobiology of Aging journal homepage: www.elsevier.com/locate/neuaging.org Sleep/wake cycle alterations as a cause of neurodegenerative diseases: A Mendelian randomization study Natalia Cullell a, Jara Cárcel-Márquez b, Cristina Gallego-Fábrega c, Elena Muiño d, Laia Llucià-Carol e, Miquel Lledós f, Karol Enrique Uscamaita Amaut g, Jerzy Krupinski h, Israel Fernández-Cadenas i,∗ a Department of Neurology, Hospital Universitari MútuaTerrassa / Fundació Docència i Recerca MútuaTerrassa, Terrassa, Spain // Stroke Pharmacogenomics and Genetics, Biomedical Research Institute Sant Pau, Sant Pau Hospital, Barcelona, Spain // Faculty of Medicine, Universitat de Barcelona, Barcelona, Spain b Stroke Pharmacogenomics and Genetics, Biomedical Research Institute Sant Pau, Sant Pau Hospital, Barcelona, Spain c Department of Neurology, Hospital Universitari MútuaTerrassa / Fundació Docència i Recerca MútuaTerrassa, Terrassa, Spain // Stroke Pharmacogenomics and Genetics, Biomedical Research Institute Sant Pau, Sant Pau Hospital, Barcelona, Spain d Stroke Pharmacogenomics and Genetics, Biomedical Research Institute Sant Pau, Sant Pau Hospital, Barcelona, Spain e Stroke Pharmacogenomics and Genetics, Biomedical Research Institute Sant Pau, Sant Pau Hospital, Barcelona, Spainn f Stroke Pharmacogenomics and Genetics, Biomedical Research Institute Sant Pau, Sant Pau Hospital, Barcelona, Spain g AdSalutem Institute Sleep Medicine, Barcelona, Spain h Department of Neurology, Hospital Universitari MútuaTerrassa / Fundació Docència i Recerca MútuaTerrassa, Terrassa, Spain // Centre for bioscience, School of HealthCare Science, Manchester Metropolitan University, Manchester, UK i Stroke Pharmacogenomics and Genetics, Biomedical Research Institute Sant Pau, Sant Pau Hospital, Barcelona, Spain // Department of Neurology, Hospital Universitari MútuaTerrassa / Fundació Docència i Recerca MútuaTerrassa, Terrassa, Spain a r t i c l e i n f o a b s t r a c t Article history: Sleep and/or wake cycle alterations are common in neurodegenerative diseases (ND). Our aim was to de- Received 21 October 2020 termine whether there is a causal relationship between sleep and/or wake cycle patterns and ND (Parkin- Revised 30 March 2021 son’s disease (PD) age at onset (AAO), Alzheimer’s disease (AD), and amyotrophic lateral sclerosis (ALS)) Accepted 12 May 2021 using two-sample Mendelian Randomization (MR). We selected 12 sleep traits with available Genome- Available online 24 May 2021 Wide Association Study (GWAS) to evaluate their causal relationship with the ND risk through Inverse- Keywords: Variance Weighted regression as main analysis. We used as outcome the latest ND GWAS with available Mendelian randomization summary-statistics: PD-AAO (N = 17,996), AD (N = 21,235) and ALS (N = 40,136). MR results pointed to Parkinson’s disease a causal effect of subjective and objective-measured morning chronotype on later PD-AAO (95%CI:0.33- sleep and/or wake cycle 1.81, p = 8.47×10−09 and 95%CI:-7.28 to -4.44, p = 5.87×10−16 , respectively). Sleep efficiency was causally Alzheimer’s disease associated with a decreased AD risk (95%CI:-20.408 to -0.66, p = 0.04) and daytime sleepiness with an Neurodegenerative disease increased ALS risk (95%CI:0.15 to 1.61, p = 0.01). Our study suggests that sleep and/or wake patterns have causal relationship with ND. Given that sleep and/or wake patterns are modifiable risk factors, sleep in- terventions should be investigated as a potential treatment in PD-AAO, AD and ALS. © 2021 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) 1. Introduction Abbreviations: ND, neurodegenerative disease; MR, Mendelian Randomization; Neurodegenerative diseases (ND) constitute a heterogeneous AAO, Age at onset; GWAS, Genome-Wide Association Study; AD, Alzheimer disease; PD, Parkinson disease; IVW, Inverse variance weighted; ALS, Amyotrophic Lateral group of conditions for which the time of symptom onset varies Sclerosis; SNVs, single nucleotide variants; REM, rapid eye movement; RBD, REM between patients. Sleep and/or wake cycle alterations are very sleep behavior disorder; OSA, obstructive sleep apnea; LD, linkage disequilibrium; common in ND and are present in up to 60% of patients WM, Weighted-Median; MBE, mode-based estimate; M10, most-active 10 hours; MIDP, Sleep midpoint timing. ∗ Corresponding author at: Stroke Pharmacogenomics and Genetics. IIB-Sant Pau. c/ Sant Antoni Mª Claret,167,08025, Barcelona, Spain. Phone +34 935 56 56 17 E-mail address: israelcadenas@yahoo.es (I. Fernández-Cadenas). 0197-4580/$ – see front matter © 2021 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) https://doi.org/10.1016/j.neurobiolaging.2021.05.008 320.e2 N. Cullell, J. Cárcel-Márquez, C. Gallego-Fábrega et al. / Neurobiology of Aging 106 (2021) 320.e320–320.e12 (Malhotra, 2018). In many cases sleep and/or wake cycle alter- time (Henry et al., 2019), and between the longer sleep duration ations occur in the prodromal stage of the disease, before mani- and cortical thickness and lower AD risk (Andrews et al., 2021). A festation of the main symptoms and final diagnosis (Iranzo, 2016; potential causal association of daytime napping and AD risk was Malhotra, 2018). Due to the uncertainty regarding the starting identified by Anserson et al. The causal relationship between some point of ND, it is difficult to establish whether sleep and/or wake sleep-related traits and the risk of PD or ALS have been also eval- cycle alterations are an early manifestation of ND or whether uated (Noyce et al., 2019; Bandres-Ciga et al., 2019). However, PD these disorders are a causal risk factor for ND (Videnovic and Ab- age at onset (AAO) was not evaluated. Moreover, the association bott, 2016). between the risk of PD, AD and ALS and some objective measures Different studies support an idea that sleep and/or wake cycle of sleep quality such as sleep efficiency or obstructive sleep apnea alterations are an underlying symptom of ND. Degeneration of the (OSA) has not been studied. cerebral nuclei involved in sleep and wake cycle leads to sleep loss The objective of this study was to analyze the causality of dif- and disruption (Peter-Derex et al., 2015). For example, degenera- ferent sleep and/or wake patterns and PD-AAO, AD and ALS. If tion of the locus coeruleus and raphe nucleus occurs in the pre- some sleep and/or wake traits are identified as causing the ND, clinical Braak stages I and II of Parkinson’s disease (PD) and cor- their diagnosis would be important for early intervention in these relates temporally with the onset of sleep alterations, such as in- patients and could help prevent or delay the onset and progression somnia or parasomnias (Braak and Del Tredici, 2008). Degenera- of ND (Loddo et al., 2017). tion of specific neurons, such as motor neurons in Amyotrophic Lateral Sclerosis (ALS) is related with breathing-related sleep disor- ders (Boentert, 2020). Neuronal loss is commonly associated with a reduction in relevant neurotransmitters involved in the sleep and/or wake cycle (Ono et al., 2012). Melatonin in Alzheimer’s dis- ease (AD) (Shukla et al., 2017) and orexin in PD (Baumann et al., 2. Methods 2008; Wienecke et al., 2012) are reduced as part of the neurode- generative process and are linked to disruption and fragmentation 2.1. Selection of exposure and instruments of sleep. However, a causal role of specific sleep and/or wake cycle To select proper genetic instruments, we searched the GWAS patterns in the risk and progression of ND is also plausible Catalog and the Sleep Disorder Knowledge Portal (Sleep Disorder (Videnovic and Abbott, 2016). The presence of sleep abnormal- KP)(“Sleep Disorder Genomics,” n.d.; “The NHGRI-EBI GWAS Cata- ities, such as sleep disruption and fragmentation, has a nega- log of published genome-wide association studies, targeted arrays tive effect on cognition. The “glymphatic system” (Videnovic and and summary statistics 2019,” n.d.) for published Genome-Wide Abbott, 2016) has been described as a potential key factor in- Association Studies (GWAS) on sleep and/or wake patterns with volved in the causal effect of sleep and/or wake cycle alterations the aim of finding genetic tools for our analysis. We considered in ND. The glymphatic system consists of a perivascular network the most common sleep and/or wake cycle disorders in ND: insom- involved in the clearance of proteins from the interstitial fluid to nia, excessive daytime sleepiness, circadian rhythm changes, rapid the cerebrospinal fluid (Nedergaard, 2013). Abnormal accumulation eye movement (REM) sleep behavior disorder (RBD), periodic leg of amyloid-β and tau is a hallmark in AD. A similar process oc- movements in sleep, restless legs syndrome, OSA, and nocturnal curs with α -synuclein in PD and TDP-43, among others, in ALS. The stridor (Kutscher et al., 2014). Moreover, we considered chrono- glymphatic system flow is reduced in the above diseases, increas- type as a sleep and/or wake cycle pattern that could be a causal ing brain accumulation of these proteins (Boland et al., 2018). The risk factor for ND. These sleep and/or wake traits were selected efficiency of the clearance system increases during sleep (Xie et al., with the summary statistics available in the GWAS Catalog or in 2013). Thus, certain sleep and/or wake cycle patterns could re- the Sleep disorder KP. We identified the following traits: insomnia, sult in an accumulation of products, such as amyloid-β , tau, α - daytime sleepiness, subjective and objective (using accelerometer synuclein or TDP-43, which could trigger the onset of ND. device measures) chronotype, sleep duration and OSA (Table 1). Given the difficulty to establish a causal relationship between Summary statistics for insomnia (Lane et al., 2019), chronotype sleep and/or wake cycle alterations and ND and the challenge of (Jones et al., 2019a), accelerometer measures (Jones et al., 2019b) conducting clinical trials that alter the sleep and/or wake cycle, and sleep duration (Braak and Del Tredici, 2008) were down- Mendelian randomization (MR) analysis could be useful. MR is loaded from the GWAS Catalog (“The NHGRI-EBI GWAS Catalog an epidemiological method that uses the same principle of ran- of published genome-wide association studies, targeted arrays and domization as clinical trials (Allman et al., 2018). MR uses ge- summary statistics 2019,” n.d.). OSA (Chen et al., 2018) and day- netic factors that are randomized at conception as instruments to time sleepiness (Wang et al., 2019) data were downloaded from elucidate the causal relationship between a risk factor (exposure) the Sleep disorder KP. We considered the genome-wide significant and a disease risk (outcome). It calculates the proportion of the variants with p < 5×10−8 in the selected GWAS as genetic instru- change explained by genetics (usually polymorphisms or single nu- ments. For OSA, for which no SNV reached these p-values, we se- cleotide variants (SNV)) in the exposure and the change explained lected all the variants with p < 10−6 as instruments. by the same genetic factors on the outcome (Emdin et al., 2017; Further information regarding the exposure variables is set out Koellinger and de Vlaming, 2019). It is possible to calculate how in Supplementary Methods and Table 1. changes in the exposure influences the outcome. This approach Heritability was calculated using the LD SCore (LDSC) (Bulik- is useful to avoid reverse causality issues, such as the possibility Sullivan et al., 2015) for all the GWAS considered as exposure vari- that the outcome may not cause any changes in genetics. More- ables to ensure the plausible use of the traits in MR. We calcu- over, genetic instruments are independent of any confounding fac- lated the F-statistics to additionally test that the instruments se- tors (Grover et al., 2017). lected were strongly associated with the exposure. A F-statistic Previous MR studies analyzed the causal relationship between greater than 10 was used as a threshold for sufficient instrument sleep traits and cognitive outcomes, cognitive decline, demen- strength (Sanderson et al., 2019). The people in charge of the tia and the risk of AD. It was found that there is a significant above-mentioned GWAS studies were responsible for obtaining in- causal relationship between sleep duration and slower reaction formed consent. N. Cullell, J. Cárcel-Márquez, C. Gallego-Fábrega et al. / Neurobiology of Aging 106 (2021) 320.e320–320.e12 320.e3 Table 1 Description of sleep traits used as exposure Trait Variable definition Subjects (N) Cohort Ancestry Significant SNV SNV SNV SNV pruned AD pruned PD pruned ALS Self-reported Chronotype Continuous variable 449,734 UK Biobank European 15155 152 137 151 considering: morning person, more a morning than evening person, more an evening person than morning, evening person Insomnia Dicothomic variable 453,379 UK Biobank European 2500 38 38 40 considering: never/rarely (345,022 vs. sometimes/usually cases/108,357 controls) Sleep duration Continous variable: 446,118 UK Biobank European 433 9 9 9 number of hours Daytime Sleepiness Continuous variable 452,071 UK Biobank European 5659 36 36 37 considering: never, sometimes, often, all of the time Objective Measures OSA Continuous variable: AHI 19,733 ARIC, CFS, FHS, Multi-ethnic 20a 1 5 1 HCHS/SIL, MESA, MrOS, Starr Chronotype M10 Continous variable: most 85,670 UK Biobank European 129 1 1 1 measures active’s 10 hours L5 Continous variable: least 85,205 UK Biobank European 75 6 3 6 active 5 hours Sleep efficiency Continous variable: total 84,810 UK Biobank European 17 2 1 2 time of sleeping divided by the total SPW Sleep duration Continous variable: total 85,449 UK Biobank European 264 9 9 9 hours in all the sleep episodes during SPW Number of sleep Continous variable: 84,810 UK Biobank European 853 20 20 21 episodes number sleep episodes during SPW MIDP Continous variable: 84,810 UK Biobank European 1 1 1 1 number of hours from the previous midnight Diurnal inactivity Continous variable: otal 84,757 UK Biobank European 31 2 2 2 daily duration of estimated bouts of inactivity that fell outside of the SPW Main characteristics of the sleep traits analyzed by MR. ‘Significant SNV’ column show the number of SNV with p-value < 5×10−08 in the GWAS and ‘SNV pruned’ the number of SNV after pruning by LD. AHI, apnea-hypoapnea index; MIDP, midnight period; OSA, obstructive sleep apnea; SPW, sleep period window a SNV had p-val < 10−7 2.2. Outcome selection Considering that almost all the GWAS selected for the exposure traits were conducted in patients from this cohort, the selection of To select outcomes for the MR analysis we searched for an outcome GWAS including samples from the UK Biobank could GWAS in different ND: PD-AAO, AD, ALS, hereditary spinocerebel- have suppose bias due to overlapping in the two-sample MR anal- lar ataxia, multiple sclerosis, Lewy body dementia, multiple sys- ysis (Burgess et al., 2016). The Moreno-Grau et al. GWAS used tem atrophy, progressive supranuclear palsy, Huntington disease the analysis of endophenotypes plus the meta-analysis of previ- and frontotemporal dementia. We selected diseases with available ous GWAS for AD to identify new loci associated with AD. For ALS, GWAS summary statistics in the GWAS catalog that had not been we used data from the latest GWAS (Nicolas et al., 2018) including previously evaluated for causality between sleep and/or wake cy- sporadic ALS cases and controls. More information regarding the cle patterns and the risk of disease and that were the most re- GWAS used as outcome is shown in Table 2. cent in the field. Finally, the traits that met these criteria and that were therefore selected as outcomes were PD-AAO, AD, and ALS 2.3. Mendelian randomization analysis (Table 2). For PD, we selected a GWAS that had not been included in previous MR analyses and for which summary statistics were We performed two-sample MR analysis following the Two Sam- available. The selected GWAS analyzed PD-AAO as a PD phenotype pleMR package (Hemani et al., 2018) pipeline in R (version 3.6.2). (Blauwendraat et al., 2019). This GWAS included only PD patients A description of the workflow used for MR analysis is shown in with information regarding AAO of the disease. The available sum- Fig. 1. mary statistics in GWAS catalog from Blauwendraat et al. did not For each sleep trait, we selected genome-wide significant poly- include the 23andMe cohort. Among the latest available GWAS for morphisms (p < 5×10−08 ) as the instrument and extracted these AD risk, we selected the most recently published by Moreno-Grau instruments in each of the 3 selected GWAS as outcomes. We then et al., because they did not include samples from the UK Biobank. harmonized the exposure and outcome data. We clumped the SNV 320.e4 N. Cullell, J. Cárcel-Márquez, C. Gallego-Fábrega et al. / Neurobiology of Aging 106 (2021) 320.e320–320.e12 Table 2 Description of GWAS used as outcome Outcome Total Subjects Cases / Controls Ancestry Reference Parkinson’s age of onset 17,996 17,996 NR (Blauwendraat et al., 2019) Alzheimer 21,235 7,879 / 5,947 NR (Moreno- Disease 4,120 / 3,289 European Grau et al., Amiotrophic lateral sclerosis 80,610 20,806 / 59,804 European (2019 ) et al., 2018) Nicolas Description of the GWAS used as outcome. NR, not-reported Fig. 1. Methodological workflow. Scheme of the workflow followed for the MR analysis. IVW, inverse variance weighted, MBE, mode-based estimate by linkage disequilibrium (LD) with an R2 < 0.001 and kb = 10,0 0 0. variability in the effect estimated by each variant using Cochran’s The clumping step was run after harmonization of the data to en- Q test and the I2 index. Cochran’s Q p-value < 0.05 and I2 > 50% sure a minimum loss of SNV due to no overlap between exposure were considered statistically significant for heterogeneity. GWAS and outcome GWAS. Then, as additional sensitivity analyses, we tested the For the MR analysis, we selected Inverse Variance Weighted Weighted-Median (WM) and MR-Egger to confirm the results (IVW) regression as previously described (Marini et al., 2020) to obtained with IVW for traits with >2 valid instruments. These infer for a causal relationship between sleep traits and outcomes tests are valid in different situations: WM can provide valid esti- considering the 3 assumptions in MR: (1) genes are assigned ran- mates when 50% or more of the information in the analysis comes domly, (2) there are genes that influence the exposure studied and from SNV that are valid instrumental variables (Bowden et al., (3) these genes influence outcome directly due to the effect on ex- 2016). MR-Egger is more conservative when pleiotropic variants posure (no horizontal pleiotropy) (Bowden et al., 2017). For analy- are present (Burgess and Thompson, 2017). For results with a ses with only 1 SNV as an instrument, we used the Wald ratio test. p-value < 0.05 in all the previous MR tests, we also performed For traits with many weak instruments, which occurred when SNV an MR test with mode-based estimate (MBE), which is robust to had p-values > 5×10−8 , we conducted the MR RAPS test, a more horizontal pleiotropy, as a third sensitivity analysis (Hartwig et al., robust method in these situations (Zhao et al., 2019). 2017). For traits with heterogeneity (Cochran’s Q p-value < 0.05), we 2.4. Sensitivity analyses performed the MR-PRESSO test, which is valid for identifying hor- izontal pleiotropic outliers in analyses with multiple instruments For traits with an IVW p-value <0.05 and 2 or more SNV used as instruments, we performed heterogeneity tests to estimate the N. Cullell, J. Cárcel-Márquez, C. Gallego-Fábrega et al. / Neurobiology of Aging 106 (2021) 320.e320–320.e12 320.e5 (Verbanck et al., 2018). The specific strategy used to determine the measured with accelerometer (IVW p = 4.33×10−3 , 1.29×10−7 , effect of outliers is described in Supplementary Methods. 2.28×10−3 , 5.87×10−16 and 0.01, respectively) (Figs. 2,5). Insom- Moreover, we performed leave-one-out analysis for traits with nia and M10 remained causal risk factors for PD-AAO after mul- multiple instruments to ensure the causal effect was not driven tiple comparison correction with Bonferroni (p < 1.3×10−3 , con- by a single genetic variant. We assessed whether the assump- sidering all the 12 traits analyzed in 3 different outcomes). Morn- tion that exposure causes outcome was valid using the Steiger ing chronotype, OSA and insomnia were positively correlated with test with the function directionality_test from the TwoSampleMR later PD-AAO (b = 1.07 (se = 0.37); b = 1.24 (se = 0.40) and package. 4.7 (se = 0.89), respectively), while M10 and sleep duration were Taking into account all the different tests, we applied workflows negatively correlated with PD-AAO (b = -5.86 (se = 0.72) and - from a recently published MR study (Marini et al., 2020) to con- 2.01 (se = 0.84), respectively). Cochran’s Q and I2 indicated het- sider causal relationships. We accepted causality when IVW was erogeneity in all 5 significant traits for PD (Table 3). The MR- significant and only 1 of the following assumptions was met: (1) PRESSO test showed presence of horizontal pleiotropy in > 50% of No heterogeneity was detected and the WM, MR-Egger and MBE the instruments (Table 3). Thus, we further investigated the reli- tests were in the same direction; (2) Heterogeneity was detected ability of the results by performing additional MR tests that con- but was corrected by MR-PRESSO (when <50% of the instruments sidered other assumptions. The MR Egger and WM were signifi- are considered outliers) (Verbanck et al., 2018), or 3) Heterogene- cant and consistent in terms of effect direction with the IVW re- ity was detected and the MR-PRESSO test detected >50% of outliers sults for chronotype and sleep duration. Moreover, the MBE regres- but MR Egger and WM were significant and had the same effect di- sion was consistent in the effect direction in both traits. However, rection and MBE had the same direction of effect (Verbanck et al., for insomnia and OSA, the WM and MR Egger, respectively, were 2018). consistent in terms of effect direction but did not reach signifi- cance (p = 0.48 and 0.14, respectively). However, as weak instru- 2.5. Reverse causality ments were selected for OSA (SNV with p<10−7 instead of SNV with p < 10−8 ), we conducted MR RAPS test that showed signifi- We studied reverse causality to test whether ND were causally cant causal association (p = 6.4×10−4 , b = 1.29 (se = 0.38)). For associated with sleep and/or wake cycle pattern traits. We per- M10 only one SNV was used as an instrument, so sensitivity anal- formed MR analyses following the above-described steps to study yses were not implemented. The Steiger test for all the significant the causal relationship: (1) MR testing with IVW; (2) Heterogeneity traits associated with PD-AAO indicated correct causal direction tests; (3) Sensitivity analysis: additional MR tests; (4) MR-PRESSO (Table S3). testing for traits with detected heterogeneity; (5) Leave-one-out We performed leave-one-out analysis for the traits with a tests, and (6) Steiger analysis. causal relationship to PD-AAO that passed the sensitivity analysis. For chronotype, we did not observe any SNV leading the causality 2.6. Genetic correlation (Fig. S1A). The leave-one-out analysis for sleep duration (Fig. S1B) showed that the entire causal effect was driven by rs113851554. To explore genetic correlation between traits, we used the LD- Removal of that SNV resulted in the inversion of the beta values SCsoftware (Bulik-Sullivan et al., 2015). We assessed genetic corre- (p = 0.028; b = 4.31 (se = 1.96)). In that case, the WM was not lations between the 3 analyzed ND traits and all the different sleep significant (p = 0.84), thus not passing the sensitivity analysis. In and/or wake cycle traits. the OSA’s leave-one-out analysis, all the effect was guided by the 2.7. Data availability statement rs34188544 SNV (Fig. S1C). The removal of that variant resulted in the beta inversion in the MR RAPS test (b = -1.96 and p = 0.38). The data that support the findings of this study are available As only one SNV was selected as an instrument for M10, we did from the corresponding author upon reasonable request. not perform a leave-one-out analysis. 3. Results 3.2. MR for AD risk In total, 12 sleep and/or wake pattern traits were analyzed as From the 12 traits analyzed as possible causal risk factors for exposure factors to assess causality for the risk of 3 outcomes: PD- AD, sleep efficiency measured with an accelerometer was the only AAO, AD, and ALS. Tables 1 and 2 describe the exposure and out- trait that was significant (IVW p = 0.036) and showed a negative comes analyzed in detail. All but 1 GWAS included for the selection correlation (b = -10.53 (se = 5.03)) (Figs. 3,5). This result indicated of instruments for MR analysis were based on the UK Biobank co- that the risk of AD increased when the patient had poorer sleep hort (Table 1). A full description of the variables used in the expo- efficiency. As only 2 SNV were considered valid instruments in this sure GWAS and the significant variants before and after LD pruning analysis, the sensitivity test performed when multiple instruments is shown in Table 1. are used could not be conducted. The Steiger test confirmed the We assessed the heritability (h2 ) of the different GWAS used as causal effect of sleep efficiency on AD risk (Table S3). exposure variables to ensure the plausibility of the MR analysis. All traits but 1 (M10) had 4% or higher heritability (Table S1). Diurnal 3.3. MR for ALS risk inactivity measured with an accelerometer was the most heritable trait (44%). The M10 measure had negative h2 , indicating inaccu- We found daytime sleepiness causally associated with ALS risk racy in this heritability estimation. Moreover, we assessed the F- (IVW p = 0.018). The correlation was positive (b = 1.17 (se 1.70)) statistic to ensure the exposure strength. The F-statistic for all the indicating that increased daytime sleepiness was causally associ- exposures was higher than 10, the rule-of-thumb cut-off (Table S2). ated with increased risk of ALS (Figs. 4,5). The sensitivity analyses confirmed these results: neither heterogeneity (Table 3) nor single 3.1. MR for PD-AAO SNV driven the association (Fig. S1D) were identified. The different MR tests evaluated were consistent in effect direction despite the When we considered PD-AAO as an outcome, we identified MR Egger was not statistically significant (p = 0.49). The Steiger causality for chronotype, insomnia, OSA, M10 and sleep duration test confirmed the correct causal direction (Table S3). 320.e6 N. Cullell, J. Cárcel-Márquez, C. Gallego-Fábrega et al. / Neurobiology of Aging 106 (2021) 320.e320–320.e12 Fig. 2. Forest plot for Mendelian randomization analysis for PD-AAO outcome. The plots show the different methods used for the main (IVW) and sensitivity analyses. OSA, obstructive sleep apnea, ACC_M10, accelerometer-based measure for M10; ACC_L5, accelerometer-based measure for L5; ACC_SLEEP_EFF, accelerometer-based measure for sleep efficiency; ACC_SLEEP_DUR, accelerometer-based measure for sleep duration; ACC_N_SLEEP, accelerometer-based measure for number of sleep episodes; ACC_MIDP, accelerometer-based measure for sleep midpoint timing; ACC_DIURNAL_INACT, accelerometer-based measure for diurnal inactivity; IVW, inverse variance weighted; WM, Weighted-Median; MBE, mode-based estimate Table 3 Heterogeneity results Outcome Exposure Q_Cochran Q_pval I2 MR PRESSO % Outliers PD-AAO Chronotype 36528.26 <0.001 99.6% 4.8e-03 >50% PD-AAO Insomnia 217216.5 <0.001 98.8% 2.19e-06 >50% PD-AAO OSA 6600.37 <0.001 99.9% 0.38 - PD-AAO Acc. Sleep Duration 4233,64 <0.001 96.0% <2e-04 >50% AD Acc. Sleep Efficiency 0.09 0.76 0% - - ALS Acc. Daytime sleepiness 43.74 0.14 19.9% - - Results for heterogeneity test (Q-cochran, I2 and MR-PRESSO) performed to the significant results in IVW test. % Outliers indicated the proportion of instrumental variants considered outliers. PD-AAO, Parkinson’s disease age at onset; AD, Alzheimer’s disease; ALS, amyotrophic lateral sclerosis; OSA, obstructive sleep apnea; Acc, Accelerometer 3.4. Reverse MR analysis predicted risk of AD was causal for nine sleep and/or wake cycle traits: chronotype, insomnia, self-reported sleep duration, L5 mea- The reverse causality analysis was conducted to confirm the sure, sleep duration measured with accelerometer, number of sleep MR results and to assess potential new causal associations in the episodes, sleep midpoint timing (MIDP) and diurnal inactivity (Fig. reverse direction. In the case of PD-AAO, the reverse MR anal- S3). All these traits but chronotype were negatively correlated with ysis was based in only one weak instrument (Fig. S2) and has AD. Heterogeneity test for these traits only showed presence of been described to lead to inconclusive results (Burgess et al., Int heterogeneity for chronotype and number of sleep episodes. How- J Epi; 2011; Davies et al., BMJ; 2018). We found that genetically ever, the MR-PRESSO test did not detect outliers (Table S3). In all N. Cullell, J. Cárcel-Márquez, C. Gallego-Fábrega et al. / Neurobiology of Aging 106 (2021) 320.e320–320.e12 320.e7 Fig. 3. Forest plot for Mendelian randomization analysis for AD outcome. The plots show the different methods used for the main (IVW) and sensitivity analyses. OSA, obstructive sleep apnea; ACC_M10, accelerometer-based measure for M10; ACC_L5, accelerometer-based measure for L5; ACC_SLEEP_EFF, accelerometer-based measure for sleep efficiency; ACC_SLEEP_DUR, accelerometer-based measure for sleep duration; ACC_N_SLEEP, accelerometer-based measure for number of sleep episodes; ACC_MIDP, accelerometer-based measure for sleep midpoint timing; ACC_DIURNAL_INACT, accelerometer-based measure for diurnal inactivity; IVW, inverse variance weighted; WM, Weighted-Median; MBE, mode-based estimate cases, the effect direction among the different MR methods was 4. Discussion consistent and was statistically significant with the exception of MR Egger for chronotype and number of sleep episodes (Figs. S2- The results from our MR study support a causal effect of some S4). We did not identify any single SNV leading the causal effect sleep-wake patterns with PD-AAO, AD and ALS. Specifically, we (Fig. S5). The causal direction was predicted to be correct by the have found a potential causal effect of morning chronotype in Steiger test (Table S5). We did not find significant causal associa- delayed PD-AAO, which suggests protection against the onset of tion for sleep efficiency (p = 0.75), the unique trait identified to be the disease. These results were consistent in both analyses based a potential causal factor of AD. on self-reported and objective measures of chronotype. Our re- The 5 hours with less activity (L5) measured with accelerom- sults based on GWAS using objectively obtained variables (with ac- eter was found to be causal outcome of ALS (Fig. S4). The beta celerometer devices) showed a causal relationship between M10 for the association indicated a positive correlation (b = 0.018 time and PD-AAO. M10 time is an indicative measure of circa- (se = 0.003). The sensitivity analysis based in heterogeneity test, dian rhythms: the higher the M10 time value, the more of an additional MR tests, leave-one-out and Steiger test confirmed this evening person you are. These results indicated that the more of causal association (Table S4-S5; Figs. S4 and S5). We did not find an evening person you are, the earlier the PD-AAO. Looking at significant causal association for daytime sleepiness (p = 0.60), the both objective- and subjective-based results, these indicate that unique trait identified to be a potential causal factor of ALS. being a morning person could delay the onset of PD. Chrono- type is an indicator of the circadian rhythm, an important regu- lator of hippocampal function, that in turn is related to neurode- 3.5. Genetic correlation analysis generation and cognitive deficits (Cho, 2001). Circadian rhythms We studied whether apart from causal associations exists ge- are fundamental in the daily ordering of cellular metabolic cy- netic correlation between ND traits and sleep and/or wake cycle cles (Hastings and Goedert, 2013). For example, the metabolism traits. We did not find any significant genetic correlation (Fig. S6). of dopamine, a key neurotransmitter in PD, is influenced by the 320.e8 N. Cullell, J. Cárcel-Márquez, C. Gallego-Fábrega et al. / Neurobiology of Aging 106 (2021) 320.e320–320.e12 Fig. 4. Forest plot for mendelian randomization analysis for ALS outcome. The plots show the different methods used for the main (IVW) and sensitivity analyses. OSA, obstructive sleep apnea; ACC_M10, accelerometer-based measure for M10; ACC_L5, accelerometer-based measure for L5; ACC_SLEEP_EFF, accelerometer-based measure for sleep efficiency; ACC_SLEEP_DUR, accelerometer-based measure for sleep duration; ACC_N_SLEEP, accelerometer-based measure for number of sleep episodes; ACC_MIDP, accelerometer-based measure for sleep midpoint timing; ACC_DIURNAL_INACT, accelerometer-based measure for diurnal inactivity; IVW, inverse variance weighted; WM, Weighted-Median; MBE, mode-based estimate We have identified sleep efficiency to be causally associated with a decreased risk of AD. The results indicate that the greater the sleep efficiency, the lower the risk of developing AD. The Steiger test and the reverse causality analysis confirmed the plausi- bility of this causal direction. As previously seen, inadequate sleep could prime the brain for neurodegeneration by promoting pro- cesses such as inflammation and synaptic damage, which exert pathogenic effects across diseases(Musiek and Holtzman, 2016). This result could reflect the importance of sleep-in metabolic homeostasis through waste removal by the glymphatic system (Xie et al., 2013). We found daytime sleepiness to be potentially causally asso- Fig. 5. Summary of the MR analyses. Summary of the results obtained from the ciated with ALS risk. The causal direction was confirmed by the different MR analyses. Size of the bubbles the strength of the associations: the big- reverse causality analysis and the Steiger test. Different sleep dis- ger ones show significant traits in IVW results that were confirmed by sensitivity analyses. The smaller bubbles indicate significant traits in IVW results that were orders have been linked with ALS. Based on a recent published not confirmed by sensitivity analyses. Red color indicates positive correlation in the systematic review, subjective sleep quality and daytime sleepiness causal effect while blue indicate negative correlation in the causal effect (Color ver- were the sleep traits most commonly studied in relation with ALS sion of the figure is available online.) (Lucia et al., 2020). Liu et al., identified higher frequency of exces- sive daytime sleepiness in patients with ALS compared with con- trols. It was linked with more serious frontal behavior, cognitive circadian rhythm (Videnovic and Golombek, 2016). The effect of and physical impairment (Liu et al., 2018). In our reverse causal- chronotype on dopamine metabolism could be a possible explana- ity analysis, we found that genetically predicted increased ALS risk tion for the results of our study. N. Cullell, J. Cárcel-Márquez, C. Gallego-Fábrega et al. / Neurobiology of Aging 106 (2021) 320.e320–320.e12 320.e9 was causally associated with an increase in L5 time, an objective accordance with what we observed in our analysis of sleep dura- measure of evening chronotype. Circadian alterations have been tion and AD risk. We did not find significant association between linked with ALS in a mouse model of the disease (Huang et al., these 2 traits. However, our results showed a trend toward shorter 2018). sleep duration leading to higher risk of AD (p = 0.098, b = -0.47 Previously, some studies in animal models and observational (se = 0.28)). Anderson et al., studied MR using different sleep traits studies in humans indicated that sleep and/or wake cycle alter- as exposure. They did not show any significant causal association. ations could lead to the pathogenesis of ND (Benedict et al., 2014, Only daytime napping was potentially causally associated with AD Kang et al., 2009, Iliff et al., 2012). For example, some animal mod- risk (Anderson et al., 2020). None of the above articles studied the els of ALS and PD found that circadian disruption led to an increase causal association of sleep efficiency and AD risk. in neuroinflammation that could be related to exacerbation of the The MR study from Huang et al., studied the reverse causal as- neuropathology of the disease (Leng et al., 2019). Several human sociation between AD risk and sleep traits (Huang et al., 2020). observational studies in older subjects showed that sleep and/or They found that the genetically predicted increase risk of AD was wake cycle alterations were related to an increased risk of incident causally associated with increased risk of morning chronotype, ear- all-cause dementia, including AD, vascular dementia (Musiek and lier L5 timing and a decrease risk in insomnia, sleep duration Holtzman, 2016), and PD (Leng et al., 2019). However, the studies and number of sleep episodes. Our reverse causality study for AD analyzing the causative risk of sleep and/or wake cycle alterations showed similar results. We also found a significant causal associa- for developing ND are longitudinal or retrospective and could be tion of increased AD risk and morning chronotype, earlier L5 tim- biased due to causal reverse effect and confounding factors. MR ing, insomnia, sleep duration and number of sleep episodes. Ad- overcomes these issues and could be better used with the pur- ditionally, we also found some traits not studied by Huang et al., pose of finding causal relationships. Thus, the MR approach used that were causal outcomes of AD risk: diurnal inactivity and the in this study has allowed us to potentially identify causal relation- number of hours from the previous midnight (MIDP). ships between sleep and/or wake cycle patterns and the risk of The ALS MR study (Bandres-Ciga et al., 2019) analyzed the ND, which is important due to its potential therapeutic implica- causal association of >700 phenotypic traits from LD-hub and MR- tions (Musiek and Holtzman, 2016). base in ALS risk. Among these traits, they included some sleep Some previous MR studies have been published in the field of and/or wake cycle features. None of these traits reached statisti- sleep and/or wake related traits and ND: 1 in PD (Noyce et al., cally significance in their study. However, as far as we know, the 2019), 4 in AD (Anderson et al., 2020; Andrews et al., 2021; authors did not study some of the sleep traits that we did: sleep Henry et al., 2019; Huang et al., 2020) and 1 in ALS (Bandres-Ciga traits calculated through accelerometer devices (M10, L5, sleep ef- et al., 2019). ficiency, number of sleep episodes, MIDP, diurnal inactivity) and The PD MR study analyzed 5,839 GWAS as exposure traits, OSA. We have found daytime sleepiness to be causally associated including sleep and/or wake cycle-related phenotypes, for their with ALS risk. The Bandres-Ciga et al. included another summary causal relationship with PD risk (Noyce et al., 2019). They devel- statistic for this trait that only considered narcolepsy and thereby, oped a web-based portal to make all their results available. The it had a smaller sample size (N = 336,082 vs 452,071). available results for chronotype and risk of PD from the PD MR Our study has some limitations. Firstly, we did not have sum- Portal are consistent with our results. The results were significant mary statistics for all sleep and/or wake cycle-related phenotypes. for morning chronotype and risk of PD in MR-Egger (p = 0.029), It is, therefore, possible that other sleep or circadian traits could showing a negative correlation (b = -0.771). Compared with this be causally related to the risk of ND. Secondly, for OSA there MR study, we evaluated PD-AAO as phenotype instead of PD risk. were no valid instruments to assess causality. Despite using spe- PD-AAO has been found to be a genetically determined pheno- cific methodology to overcome this issue, the results obtained may type that shares some genetic factors with the risk of PD. How- not be totally reliable. Moreover, we were only able to consider PD, ever, GWAS of PD-AAO identified some additional genetic factors AD and ALS, as they were the only ND with complete summary exclusively associated with this phenotype and not with PD risk. statistics data. The genetic factors associated with this trait are also important To conclude, our study has allowed the identification of poten- for characterizing the pathways regulating PD (Blauwendraat et al., tial causal relationships between sleep and/or wake patterns and 2019). PD-AAO, AD and ALS. The causal relationships found in this study Three out of the 4 MR studies in AD that we identified have an- could lead to further studies investigating the therapeutic implica- alyzed the AD risk as outcome. The other one studied the reverse tions of sleep and/or wake cycle patterns to prevent or slow the causal association and considered AD risk as the exposure factor. progression of ND. Specifically, the effect of chronotype modifica- The summary statistics for the AD GWAS used as outcome in these tion on PD-AAO and even on the risk of the disease, the effect of studies differs from the one we used in our analysis. They were sleep efficiency on AD risk and the effect of daytime sleepiness in published before (Huang et al., 2017; Lambert et al., 2013) or they ALS should be evaluated in future studies. were based on the late-onset AD (Kunkle et al., 2019). In the MR study from Henry et al., 2019, the authors studied the causal as- Declaration of Competing Interest sociation of sleep duration with cognition and AD risk. They only identified a small negative causal effect of sleep duration on re- We do not have any conflict of interest. action time, a cognitive marker. However, they did not find any association between sleep duration and all-cause dementia or AD Credit author statement risk using MR (Henry et al., 2019). The GWAS from Andrews et al., studied different traits previously associated with AD risk in obser- Natalia Cullell: Conceptualization, Methodology, Formal analy- vational studies. Among them, 2 sleep traits were included: sleep sis, Investigation, Writing - Original Draft. Jara Cárcel-Márquez: duration and insomnia. They found that longer sleep duration was Validation, Writing - Review & Editing. Cristina Gallego-Fábrega: causality associated with cortical thickness and shorter sleep du- Writing - Review & Editing. Elena Muiño: Writing - Review & ration with an increased risk of AD (Andrews et al., 2021). They Editing, Data Curation. Laia Llucià: Writing - Review & Editing. did not reach conclusive results for insomnia because the sensitiv- Miquel Lledós: Writing - Review & Editing. Karol Enrique Usca- ity analyses did not confirm their observations. These results are in maita Amaut: Writing - Review & Editing, Data Curation. Jerzy 320.e10 N. Cullell, J. Cárcel-Márquez, C. Gallego-Fábrega et al. / Neurobiology of Aging 106 (2021) 320.e320–320.e12 Krupinski: Conceptualization, Supervision, Writing - Review & Edit- Grosset, D.G., Gasser, T., Heutink, P., Shulman, J.M., Wood, N., Hardy, J., Mor- ing, Project administration, Funding acquisition. Israel Fernández- ris, H.R., Hinds, D.A., Gratten, J., Visscher, P.M., Gan-Or, Z., Nalls, M.A., Single- ton, A.B.23andMe Research Team, 2019. International Parkinson’s Disease Ge- Cadenas: Conceptualization, Supervision, Writing - Review & Edit- nomics Consortium (IPDGC), 2019. Parkinson’s disease age at onset genome- ing, Project administration, Funding acquisition wide association study: Defining heritability, genetic loci, and α -synuclein mechanisms. Mov. Disord. Off. J. Mov. Disord. Soc 34, 866–875. doi:10.1002/mds. 27659. Funding sources Boentert, M., 2020. Sleep and Sleep Disruption in Amyotrophic Lateral Sclerosis. Curr. Neurol. Neurosci. Rep 20. doi:10.1007/s11910- 020- 01047- 1. Boland, B., Yu, W.H., Corti, O., Mollereau, B., Henriques, A., Bezard, E., Pas- This work was supported by the EPIGENESIS project (Instituto tores, G.M., Rubinsztein, D.C., Nixon, R.A., Duchen, M.R., Mallucci, G.R., Kroe- Carlos III [Carlos III Institute] (PI17/02089); FEDER funds, Marató mer, G., Levine, B., Eskelinen, E.-L., Mochel, F., Spedding, M., Louis, C., Mar- TV3 and Fundació MútuaTerrassa; the MAESTRO project (Carlos tin, O.R., Millan, M.J., 2018. Promoting the clearance of neurotoxic proteins in neurodegenerative disorders of aging. Nat. Rev. Drug Discov 17, 660–688. III Institute, PI18/01338); the SEDMAN Study (Boehringer Ingel- doi:10.1038/nrd.2018.109. heim); the APHAS Study (Bristol-Myers Squibb); and Ibiostroke Bowden, J., Davey Smith, G., Haycock, P.C., Burgess, S., 2016. Consistent estimation (Eranet Neuron). E. Muiño is supported by a Río Hortega Con- in mendelian randomization with some invalid instruments using a weighted tract (CM18/00198) from the Instituto de Salud Carlos III. J. median estimator. Genet. Epidemiol 40, 304–314. doi:10.1002/gepi.21965. Bowden, J., Del Greco, M, F., Minelli, C., Davey Smith, G., Sheehan, N., Thompson, J., Cárcel-Márquez is supported by an AGAUR Contract (agència 2017. A framework for the investigation of pleiotropy in two-sample summary de gestió d’ajuts universitaris i de recerca; FI_DGR 2020, grant data Mendelian randomization. Stat. Med 36, 1783–1802. doi:10.1002/sim.7221. number 2020FI_B1 00157) co-financed with Fons Social Europeu Braak, H., Del Tredici, K., 2008. Invited Article: Nervous system pathology in sporadic Parkinson disease. Neurology 70, 1916–1925. doi:10.1212/01.wnl. (FSE). C. Gallego-Fabrega is supported by a Sara Borrell Contract 0 0 0 0312279.49272.9f. (CD20/0 0 043) from Instituto de Salud Carlos III. M. Lledós is sup- Bulik-Sullivan, B.K., Loh, P.-R., Finucane, H., Ripke, S., Yang, J., Patterson, N., ported by a PFIS Contract (Contratos Predoctorales de Formación Daly, M.J., Price, A.L., Neale, B.M., 2015. LD score regression distinguishes con- founding from polygenicity in genome-wide association studies. Nat. Genet 47, en Investigación en Salud) from the Instituto de Salud Carlos III. 291–295. doi:10.1038/ng.3211. I. Fernández-Cadenas is supported by a research contract from Burgess, S., Davies, N.M., Thompson, S.G., 2016. Bias due to participant overlap in Miguel Servet Program (CM18/00198), from the Instituto de Salud two-sample Mendelian randomization. Genet. Epidemiol 40, 597–608. doi:10. 1002/gepi.21998. Carlos III. Burgess, S., Thompson, S.G., 2017. Interpreting findings from Mendelian randomiza- tion using the MR-Egger method. Eur. J. Epidemiol 32, 377–389. doi:10.1007/ s10654- 017- 0255- x. Acknowledgements Chen, H., Cade, B.E., Gleason, K.J., Bjonnes, A.C., Stilp, A.M., Sofer, T., Conomos, M.P., Ancoli-Israel, S., Arens, R., Azarbarzin, A., Bell, G.I., Below, J.E., Chun, S., We thank the International Stroke Genetics Consortium, the Evans, D.S., Ewert, R., Frazier-Wood, A.C., Gharib, S.A., Haba-Rubio, J., Ha- gen, E.W., Heinzer, R., Hillman, D.R., Johnson, W.C., Kutalik, Z., Lane, J.M., Spanish Stroke Genetics Consortium and the RETICS Network Larkin, E.K., Lee, S.K., Liang, J., Loredo, J.S., Mukherjee, S., Palmer, L.J., Papani- INVICTUS (RD16/0 019/0 0 02, RD16/0 019/0 010, RD16/0 019/0 011, colaou, G.J., Penzel, T., Peppard, P.E., Post, W.S., Ramos, A.R., Rice, K., Rotter, J.I., RD16/0 019/0 021). Sands, S.A., Shah, N.A., Shin, C., Stone, K.L., Stubbe, B., Sul, J.H., Tafti, M., Tay- lor, K.D., Teumer, A., Thornton, T.A., Tranah, G.J., Wang, C., Wang, H., Warby, S.C., Wellman, D.A., Zee, P.C., Hanis, C.L., Laurie, C.C., Gottlieb, D.J., Patel, S.R., Zhu, X., Supplementary materials Sunyaev, S.R., Saxena, R., Lin, X., Redline, S., 2018. Multiethnic meta-analysis identifies RAI1 as a possible obstructive sleep apnea–related quantitative trait locus in men. Am. J. Respir. Cell Mol. Biol 58, 391–401. doi:10.1165/rcmb. Supplementary material associated with this article can be 2017-0237OC. found, in the online version, at doi:10.1016/j.neurobiolaging.2021. Cho, K., 2001. Chronic “jet lag” produces temporal lobe atrophy and spatial cognitive 05.008. deficits. Nat. Neurosci 4, 567–568. doi:10.1038/88384. Emdin, C.A., Khera, A.V., Kathiresan, S., 2017. Mendelian Randomization. JAMA 318, 1925–1926. doi:10.1001/jama.2017.17219. Grover, S., Del Greco, M, F., Stein, C.M., Ziegler, A., 2017. Mendelian Randomization. References Methods Mol. Biol. Clifton NJ 1666, 581–628. doi:10.1007/978- 1- 4939- 7274- 6_ 29. Allman, P.H., Aban, I.B., Tiwari, H.K., Cutter, G.R., 2018. An introduction to Mendelian Hartwig, F.P., Davey Smith, G., Bowden, J, 2017. Robust inference in summary data randomization with applications in neurology. Mult. Scler. Relat. Disord 24, 72– Mendelian randomization via the zero modal pleiotropy assumption. Int. J. Epi- 78. doi:10.1016/j.msard.2018.06.017. demiol 46, 1985–1998. doi:10.1093/ije/dyx102. Anderson, E.L., Richmond, R.C., Jones, S.E., Hemani, G., Wade, K.H., Dashti, H.S., Hastings, M.H., Goedert, M., 2013. Circadian clocks and neurodegenerative diseases: Lane, J.M., Wang, H., Saxena, R., Brumpton, B., Korologou-Linden, R., Nielsen, J.B., time to aggregate? Curr. Opin. Neurobiol 23, 880–887. doi:10.1016/j.conb.2013. Åsvold, B.O., Abecasis, G., Coulthard, E., Kyle, S.D., Beaumont, R.N., Tyrrell, J., 05.004. Frayling, T.M., Munafò, M.R., Wood, A.R., Ben-Shlomo, Y., Howe, L.D., Hemani, G., Zheng, J., Elsworth, B., Wade, K.H., Haberland, V., Baird, D., Laurin, C., Lawlor, D.A., Weedon, M.N., Davey Smith, G., 2020. Is disrupted sleep a risk fac- Burgess, S., Bowden, J., Langdon, R., Tan, V.Y., Yarmolinsky, J., Shihab, H.A., Timp- tor for Alzheimer’s disease? Evidence from a two-sample Mendelian random- son, N.J., Evans, D.M., Relton, C., Martin, R.M., Smith, Davey, G., Gaunt, T.R., ization analysis. Int. J. Epidemiol doi:10.1093/ije/dyaa183. P.C., Haycock, 2018. The MR-Base platform supports systematic causal inference Andrews, S.J., Fulton-Howard, B., O’Reilly, P., Marcora, E., Goate, A.M., Farrer, L.A., across the human phenome. eLife 7. doi:10.7554/eLife.34408. Haines, J.L., Mayeux, R., Naj, A.C., Pericak-Vance, M.A., Schellenberg, G.D., Henry, A., Katsoulis, M., Masi, S., Fatemifar, G., Denaxas, S., Acosta, D., Garfield, V., Wang, L.collaborators of the Alzheimer’s Disease Genetics Consortium, 2021. Dale, C.E., 2019. The relationship between sleep duration, cognition and demen- Causal Associations Between Modifiable Risk Factors and the Alzheimer’s Phe- tia: a Mendelian randomization study. Int. J. Epidemiol 48, 849–860. doi:10. nome. Ann. Neurol 89, 54–65. doi:10.1002/ana.25918. 1093/ije/dyz071. Bandres-Ciga, S., Noyce, A.J., Hemani, G., Nicolas, A., Calvo, A., Mora, G., Tienari, P.J., Huang, J., Zuber, V., Matthews, P.M., Elliott, P., Tzoulaki, J., Dehghan, A., 2020. Sleep, Stone, D.J., Nalls, M.A., Singleton, A.B., Chiò, A., Traynor, B.J., 2019. Shared poly- major depressive disorder, and Alzheimer disease. Neurology 95, e1963–e1970. genic risk and causal inferences in amyotrophic lateral sclerosis. Ann. Neurol 85, doi:10.1212/WNL.0 0 0 0 0 0 0 0 0 0 010463. 470–481. doi:10.1002/ana.25431. International Genomics of Alzheimer’s Project, Alzheimer’s Disease Neuroimaging Baumann, C.R., Scammell, T.E., Bassetti, C.L., 2008. Parkinson’s disease, sleepiness Initiative Huang, K.-L., Marcora, E., Pimenova, A.A., Di Narzo, A.F., Kapoor, M., and hypocretin/orexin. Brain J. Neurol 131, e91. doi:10.1093/brain/awm220. Jin, S.C., Harari, O., Bertelsen, S., Fairfax, B.P., Czajkowski, J., Chouraki, V., Grenier- Benedict, C., Cedernaes, J., Giedraitis, V., Nilsson, E.K., Hogenkamp, P.S., Vågesjö, E., Boley, B., Bellenguez, C., Deming, Y., McKenzie, A., Raj, T., Renton, A.E., Budde, J., Massena, S., Pettersson, U., Christoffersson, G., Phillipson, M., Broman, J.-E., Lan- Smith, A., Fitzpatrick, A., Bis, J.C., DeStefano, A., Adams, H.H.H., Ikram, M.A., nfelt, L., Zetterberg, H., Schiöth, H.B., 2014. Acute sleep deprivation increases van der Lee, S., Del-Aguila, J.L., Fernandez, M.V., Ibañez, L., Sims, R., Escott- serum levels of neuron-specific enolase (NSE) and S100 calcium binding protein Price, V., Mayeux, R., Haines, J.L., Farrer, L.A., Pericak-Vance, M.A., Lambert, J.C., B (S-100B) in healthy young men. Sleep 37, 195–198. doi:10.5665/sleep.3336. van Duijn, C., Launer, L., Seshadri, S., Williams, J., Amouyel, P., Schellenberg, G.D., Blauwendraat, C., Heilbron, K., Vallerga, C.L., Bandres-Ciga, S., von Coelln, R., Zhang, B., Borecki, I., Kauwe, J.S.K., Cruchaga, C., Hao, K., Goate, A.M., 2017. A Pihlstrøm, L., Simón-Sánchez, J., Schulte, C., Sharma, M., Krohn, L., Siito- common haplotype lowers PU.1 expression in myeloid cells and delays onset of nen, A., Iwaki, H., Leonard, H., Noyce, A.J., Tan, M., Gibbs, J.R., Hernandez, D.G., Alzheimer’s disease. Nat. Neurosci 20, 1052–1061. doi:10.1038/nn.4587. Scholz, S.W., Jankovic, J., Shulman, L.M., Lesage, S., Corvol, J.-C., Brice, A., Huang, Z., Liu, Q., Peng, Y., Dai, J., Xie, Y., Chen, W., Long, S., Pei, Z., Su, H., Yao, X., van Hilten, J.J., Marinus, J., Eerola-Rautio, J., Tienari, P., Majamaa, K., Toft, M., 2018. Circadian Rhythm Dysfunction Accelerates Disease Progression in a Mouse N. Cullell, J. Cárcel-Márquez, C. Gallego-Fábrega et al. / Neurobiology of Aging 106 (2021) 320.e320–320.e12 320.e11 Model With Amyotrophic Lateral Sclerosis. Front. Neurol 9. doi:10.3389/fneur. McDavid, A.N, Gallacher, J., McKee, A.C, van den Bussche, H., Mesulam, M., 2018.00218. Brayne, C., Miller, B.L, Riedel-Heller, S., Miller, C.A, Miller, J.W, Al-Chalabi, A., Iliff, J.J., Wang, M., Liao, Y., Plogg, B.A., Peng, W., Gundersen, G.A., Benveniste, H., Morris, J.C, Shaw, C.E, Myers, A.J, Wiltfang, J., O’Bryant, S., Olichney, J.M, Vates, G.E., Deane, R., Goldman, S.A., Nagelhus, E.A., Nedergaard, M., 2012. A Alvarez, V., Parisi, J.E, Singleton, A.B, Paulson, H.L, Collinge, J., Perry, W.R, paravascular pathway facilitates CSF flow through the brain parenchyma and Mead, S., Peskind, E., Cribbs, D.H, Rossor, M., Pierce, A., Ryan, N.S, Poon, W.W, the clearance of interstitial solutes, including amyloid β . Sci. Transl. Med 4, Nacmias, B., Potter, H., Sorbi, S., Quinn, J.F, Sacchinelli, E., Raj, A., Spalletta, G., 147ra111. doi:10.1126/scitranslmed.3003748. Raskind, M., Caltagirone, C., Bossù, P., Donata Orfei, M., Reisberg, B., Clarke, R., Iranzo, A., 2016. Sleep in Neurodegenerative Diseases. Sleep Med. Clin 11, 1–18. Reitz, C., Smith, A.D, Ringman, J.M, Warden, D., Roberson, E.D, Wilcock, G., Ro- doi:10.1016/j.jsmc.2015.10.011. gaeva, E., Cecilia Bruni, A., Rosen, H.J, Gallo, M., Rosenberg, R.N, Ben-Shlomo, Y., Jones, S.E., Lane, J.M., Wood, A.R., van Hees, V.T., Tyrrell, J., Beaumont, R.N., Jef- Sager, M.A, Mecocci, P., Saykin, A.J, Pastor, P., Cuccaro, M.L, Vance, J.M, Schnei- fries, A.R., Dashti, H.S., Hillsdon, M., Ruth, K.S., Tuke, M.A., Yaghootkar, H., der, J.A, Schneider, L.S, Slifer, S., Seeley, W.W, Smith, A.G, Sonnen, J.A, Spina, S., Sharp, S.A., Jie, Y., Thompson, W.D., Harrison, J.W., Dawes, A., Byrne, E.M., Stern, R.A, Swerdlow, R.H, Tang, M., Tanzi, R.E, Trojanowski, J.Q, Troncoso, J.C, Tiemeier, H., Allebrandt, K.V., Bowden, J., Ray, D.W., Freathy, R.M., Mur- Van Deerlin, V.M, Van Eldik, L.J, Vinters, H.V, Paul Vonsattel, J., Weintraub, S., ray, A., Mazzotti, D.R., Gehrman, P.R., Lawlor, D.A., Frayling, T.M., Rutter, M.K., Welsh-Bohmer, K.A, Wilhelmsen, K.C, Williamson, J., Wingo, T.S, Woltjer, R.L, Hinds, D.A., Saxena, R., Weedon, M.N., 2019a. Genome-wide association analyses Wright, C.B, Yu, C.-E, Yu, L., Saba, Y., Pilotto, A., Bullido, M.J, Peters, O., Crane, P.K, of chronotype in 697,828 individuals provides insights into circadian rhythms. Bennett, D., Bosco, P., Coto, E., Boccardi, V., De Jager, P.L, Lleo, A., Warner, N., Nat. Commun 10, 343. doi:10.1038/s41467- 018- 08259- 7. Lopez, O.L, Ingelsson, M., Deloukas, P., Cruchaga, C., Graff, C., Gwilliam, R., For- Jones, S.E., van Hees, V.T., Mazzotti, D.R., Marques-Vidal, P., Sabia, S., van der nage, M., Goate, A.M, Sanchez-Juan, P., Kehoe, P.G, Amin, N., Ertekin-Taner, N., Spek, A., Dashti, H.S., Engmann, J., Kocevska, D., Tyrrell, J., Beaumont, R.N., Berr, C., Debette, S., Love, S., Launer, L.J, Younkin, S.G, Dartigues, J.-F, Corco- Hillsdon, M., Ruth, K.S., Tuke, M.A., Yaghootkar, H., Sharp, S.A., Ji, Y., Harri- ran, C., Arfan Ikram, M., Dickson, D.W, Nicolas, G., Campion, D., Tschanz, J., son, J.W., Freathy, R.M., Murray, A., Luik, A.I., Amin, N., Lane, J.M., Saxena, R., Schmidt, H., Hakonarson, H., Clarimon, J., Munger, R., Schmidt, R., Farrer, L.A, Rutter, M.K., Tiemeier, H., Kutalik, Z., Kumari, M., Frayling, T.M., Weedon, M.N., Van Broeckhoven, C., O’Donovan, M.C, Destefano, A.L, Jones, L., Haines, J.L, Gehrman, P.R., Wood, A.R., 2019b. Genetic studies of accelerometer-based sleep Deleuze, J.-F, Owen, M.J, Gudnason, V., Mayeux, R., Escott-Price, V., Psaty„ B.M, measures yield new insights into human sleep behaviour. Nat. Commun 10, Ramirez„ A., Wang, L.-S, Ruiz„ A., van Duijn, C.M, Holmans, P.A, Seshadri, S., 1585. doi:10.1038/s41467- 019- 09576- 1. Williams„ J., Amouyel, P., Schellenberg, G.D, Lambert, J.-C, Pericak-Vance, M.A, Kang, J.-E., Lim, M.M., Bateman, R.J., Lee, J.J., Smyth, L.P., Cirrito, J.R., Fujiki, N., 2019. Genetic meta-analysis of diagnosed Alzheimer’s disease identifies new Nishino, S., Holtzman, D.M., 2009. Amyloid-beta dynamics are regulated by risk loci and implicates Aβ , tau, immunity and lipid processing. Nat. Genet 51, orexin and the sleep-wake cycle. Science 326, 1005–1007. doi:10.1126/science. 414–430. doi:10.1038/s41588- 019- 0358- 2. 1180962. Kutscher, S.J., Farshidpanah, S., Claassen, D.O., 2014. Sleep Dysfunction and its Man- Koellinger, P.D., de Vlaming, R., 2019. Mendelian randomization: the challenge of agement in Parkinson’s Disease. Curr. Treat. Options Neurol 16, 304. doi:10. unobserved environmental confounds. Int. J. Epidemiol 48, 665–671. doi:10. 1007/s11940- 014- 0304- 7. 1093/ije/dyz138. Lambert, J.C., Ibrahim-Verbaas, C.A., Harold, D., Naj, A.C., Sims, R., Bellenguez, C., Kunkle, B.W., Grenier-Boley, B., Sims, R., Bis, J.C, Damotte, V., Naj, A.C, Boland, A., DeStafano, A.L., Bis, J.C., Beecham, G.W., Grenier-Boley, B., Russo, G., Thorton- Vronskaya, M., van der Lee, S.J, Amlie-Wolf, A., Bellenguez, C., Frizatti, A., Wells, T.A., Jones, N., Smith, A.V., Chouraki, V., Thomas, C., Ikram, M.A., Ze- Chouraki, V., Martin, E.R, Sleegers, K., Badarinarayan, N., Jakobsdottir, J., lenika, D., Vardarajan, B.N., Kamatani, Y., Lin, C.F., Gerrish, A., Schmidt, H., Kun- Hamilton-Nelson, K.L, Moreno-Grau, S., Olaso, R., Raybould, R., Chen, Y., kle, B., Dunstan, M.L., Ruiz, A., Bihoreau, M.T., Choi, S.H., Reitz, C., Pasquier, F., Kuzma, A.B, Hiltunen, M., Morgan, T., Ahmad, S., Vardarajan, B.N, Epelbaum, J., Cruchaga, C., Craig, D., Amin, N., Berr, C., Lopez, O.L., De Jager, P.L., Derame- Hoffmann, P., Boada, M., Beecham, G.W, Garnier, J.-G, Harold, D., Fitzpatrick, A.L, court, V., Johnston, J.A., Evans, D., Lovestone, S., Letenneur, L., Morón, F.J., Ru- Valladares, O., Moutet, M.-L, Gerrish, A., Smith, A.V, Qu, L., Bacq, D., Den- binsztein, D.C., Eiriksdottir, G., Sleegers, K., Goate, A.M., Fiévet, N., Huentel- ning, N., Jian, X., Zhao, Y., Del Zompo, M., Fox, N.C, Choi, S.-H, Mateo, I., man, M.W., Gill, M., Brown, K., Kamboh, M.I., Keller, L., Barberger-Gateau, P., Hughes, J.T, Adams, H.H, Malamon, J., Sanchez-Garcia, F., Patel, Y., Brody, J.A, McGuiness, B., Larson, E.B., Green, R., Myers, A.J., Dufouil, C., Todd, S., Wallon, D., Dombroski, B.A, Deniz Naranjo, M.C, Daniilidou, M., Eiriksdottir, G., Mukher- Love, S., Rogaeva, E., Gallacher, J., St George-Hyslop, P., Clarimon, J., Lleo, A., jee, S., Wallon, D., Uphill, J., Aspelund, T., Cantwell, L.B, Garzia, F., Galim- Bayer, A., Tsuang, D.W., Yu, L., Tsolaki, M., Bossù, P., Spalletta, G., Proitsi, P., berti, D., Hofer, E., Butkiewicz, M., Fin, B., Scarpini, E., Sarnowski, C., Bush, W.S, Collinge, J., Sorbi, S., Sanchez-Garcia, F., Fox, N.C., Hardy, J., Deniz Naranjo, M.C., Meslage, S., Kornhuber, J., White, C.C, Song, Y., Barber, R.C, Engelborghs, S., Bosco, P., Clarke, R., Brayne, C., Galimberti, D., Mancuso, M., Matthews, F., Moe- Sordon, S., Voijnovic, D., Adams, P.M, Vandenberghe, R., Mayhaus, M., Adri- bus, S., Mecocci, P., Del Zompo, M., Maier, W., Hampel, H., Pilotto, A., Bul- enne Cupples, L., Albert, M.S, De Deyn, P.P, Gu, W., Himali, J.J, Beekly, D., lido, M., Panza, F., Caffarra, P., Nacmias, B., Gilbert, J.R., Mayhaus, M., Lan- Squassina, A., Hartmann, A.M, Orellana, A., Blacker, D., Rodriguez-Rodriguez, E., nefelt, L., Hakonarson, H., Pichler, S., Carrasquillo, M.M., Ingelsson, M., Beekly, D., Lovestone, S., Garcia, M.E, Doody, R.S, Munoz-Fernadez, C., Sussams, R., Lin, H., Alvarez, V., Zou, F., Valladares, O., Younkin, S.G., Coto, E., Hamilton-Nelson, K.L., Fairchild, T.J, Benito, Y.A, Holmes, C., Karamujić-Čomić, H., Frosch, M.P, Thon- Gu, W., Razquin, C., Pastor, P., Mateo, I., Owen, M.J., Faber, K.M., Jons- berg, H., Maier, W., Roshchupkin, G., Ghetti, B., Giedraitis, V., Kawalia, A., son, P.V., Combarros, O., O’Donovan, M.C., Cantwell, L.B., Soininen, H., Blacker, D., Li, S., Huebinger, R.M, Kilander, L., Moebus, S., Hernández, I., Ilyas Kamboh, M., Mead, S., Mosley, T.H., Bennett, D.A., Harris, T.B., Fratiglioni, L., Holmes, C., Brundin, R., Turton, J., Yang, Q., Katz, M.J, Concari, L., Lord, J., Beiser, A.S, de Bruijn, R.F., Passmore, P., Montine, T.J., Bettens, K., Rotter, J.I., Brice, A., Dirk Keene, C., Helisalmi, S., Kloszewska, I., Kukull, W.A, Maria Koivisto, A., Morgan, K., Foroud, T.M., Kukull, W.A., Hannequin, D., Powell, J.F., Nalls, M.A., Lynch, A., Tarraga, L., Larson, E.B, Haapasalo, A., Lawlor, B., Mosley, T.H, Lip- Ritchie, K., Lunetta, K.L., Kauwe, J.S., Boerwinkle, E., Riemenschneider, M., ton, R.B, Solfrizzi, V., Gill, M., Longstreth, W.T, Montine, T.J, Frisardi, V., Diez- Boada, M., Hiltuenen, M., Martin, E.R., Schmidt, R., Rujescu, D., Wang, L.S., Fairen, M., Rivadeneira, F., Petersen, R.C, Deramecourt, V., Alvarez, I., Salani, F., Dartigues, J.F., Mayeux, R., Tzourio, C., Hofman, A., Nöthen, M.M., Graff, C., Ciaramella, A., Boerwinkle, E., Reiman, E.M, Fievet, N., Rotter, J.I, Reisch, J.S, Psaty, B.M., Jones, L., Haines, J.L., Holmans, P.A., Lathrop, M., Pericak-Vance, M.A., Hanon, O., Cupidi, C., Uitterlinden, A.G.A, Royall, D.R, Dufouil, C., Giovanni Launer, L.J., Farrer, L.A., van Duijn, C.M., Van Broeckhoven, C., Moskvina, V., Se- Maletta, R., de Rojas, I., Sano, M., Brice, A., Cecchetti, R., St George-Hyslop, P., shadri, S., Williams, J., Schellenberg, G.D., Amouyel, P.European Alzheimer’s Dis- Ritchie, K., Tsolaki, M., Tsuang, D.W, Dubois, B., Craig, D., Wu, C.-K, Soini- ease Initiative (EADI), Genetic and Environmental Risk in Alzheimer’s Disease, nen, H., Avramidou, D., Albin, R.L, Fratiglioni, L., Germanou, A., Apostolova, L.G, Alzheimer’s Disease Genetic Consortium, Cohorts for Heart and Aging Research Keller, L., Koutroumani, M., Arnold, S.E, Panza, F., Gkatzima, O., Asthana, S., in Genomic Epidemiology, 2013. Meta-analysis of 74,046 individuals identifies Hannequin, D., Whitehead, P., Atwood, C.S, Caffarra, P., Hampel, H., Quin- 11 new susceptibility loci for Alzheimer’s disease. Nat. Genet. 45, 1452–1458. tela, I., Carracedo, Á., Lannfelt, L., Rubinsztein, D.C, Barnes, L.L, Pasquier, F., doi:10.1038/ng.2802. Frölich, L., Barral, S., McGuinness, B., Beach, T.G, Johnston, J.A, Becker, J.T, Pass- Lane, J.M., Jones, S.E., Dashti, H.S., Wood, A.R., Aragam, K.G., van Hees, V.T., more, P., Bigio, E.H, Schott, J.M, Bird, T.D, Warren, J.D, Boeve, B.F, Lupton, M.K, Strand, L.B., Winsvold, B.S., Wang, H., Bowden, J., Song, Y., Patel, K., Ander- Bowen, J.D, Proitsi, P., Boxer, A., Powell, J.F, Burke, J.R, Kauwe, J.S.K, Burns, J.M, son, S.G., Beaumont, R.N., Bechtold, D.A., Cade, B.E., Haas, M., Kathiresan, S., Mancuso, M., Buxbaum, J.D, Bonuccelli, U., Cairns, N.J, McQuillin, A., Cao, C., Little, M.A., Luik, A.I., Loudon, A.S., Purcell, S., Richmond, R.C., Scheer, F.A.J.L., Livingston, G., Carlson, C.S, Bass, N.J, Carlsson, C.M, Hardy, J., Carney, R.M, Schormair, B., Tyrrell, J., Winkelman, J.W., Winkelmann, J., Hveem, K., Zhao, C., Bras, J., Carrasquillo, M.M, Guerreiro, R., Allen, M., Chui, H.C, Fisher, E., Ma- Nielsen, J.B., Willer, C.J., Redline, S., Spiegelhalder, K., Kyle, S.D., Ray, D.W., sullo, C., Crocco, E.A, DeCarli, C., Bisceglio, G., Dick, M., Ma, L., Duara, R., Zwart, J.-A., Brumpton, B., Frayling, T.M., Lawlor, D.A., Rutter, M.K., Wee- Graff-Radford, N.R, Evans, D.A, Hodges, A., Faber, K.M, Scherer, M., Fallon, K.B, don, M.N., Saxena, R., 2019. Biological and clinical insights from genetics of in- Riemenschneider, M., Fardo, D.W, Heun, R., Farlow, M.R, Kölsch, H., Ferris, S., somnia symptoms. Nat. Genet. 51, 387–393. doi:10.1038/s41588- 019- 0361- 7. Leber, M., Foroud, T.M, Heuser, I., Galasko, D.R, Giegling, I., Gearing, M., Hüll, M., Leng, Y., Musiek, E.S., Hu, K., Cappuccio, F.P., Yaffe, K., 2019. Association between Geschwind, D.H, Gilbert, J.R, Morris, J., Green, R.C, Mayo, K., Growdon, J.H, Feul- circadian rhythms and neurodegenerative diseases. Lancet Neurol 18, 307–318. ner, T., Hamilton, R.L, Harrell, L.E, Drichel, D., Honig, L.S, Cushion, T.D, Huentel- doi:10.1016/S1474- 4422(18)30461- 7. man, M.J, Hollingworth, P., Hulette, C.M, Hyman, B.T, Marshall, R., Jarvik, G.P, Liu, S., Huang, Y., Tai, H., Zhang, K., Wang, Z., Shen, D., Fu, H., Su, N., Shi, J., Ding, Q., Meggy, A., Abner, E., Menzies, G.E, Jin, L.-W, Leonenko, G., Real, L.M, Jun, G.R, Liu, M., Guan, Y., Gao, J., Cui, L., 2018. Excessive daytime sleepiness in Chi- Baldwin, C.T, Grozeva, D., Karydas, A., Russo, G., Kaye, J.A, Kim, R., Jessen, F., nese patients with sporadic amyotrophic lateral sclerosis and its association Kowall, N.W, Vellas, B., Kramer, J.H, Vardy, E., LaFerla, F.M, Jöckel, K.-H, Lah, J.J, with cognitive and behavioural impairments. J. Neurol. Neurosurg. Psychiatry Dichgans, M., Leverenz, J.B, Mann, D., Levey, A.I, Pickering-Brown, S., Lieber- 89, 1038–1043. doi:10.1136/jnnp- 2018- 318810. man, A.P, Klopp, N., Lunetta, K.L, Wichmann, H.-E, Lyketsos, C.G, Morgan, K., Loddo, G., Calandra-Buonaura, G., Sambati, L., Giannini, G., Cecere, A., Cortelli, P., Marson, D.C, Brown, K., Martiniuk, F., Medway, C., Mash, D.C, Nöthen, M.M, Provini, F., 2017. The Treatment of Sleep Disorders in Parkinson’s Disease: From Masliah, E., Hooper, N.M, McCormick, W.C, Daniele, A., McCurry, S.M, Bayer, A., Research to Clinical Practice. Front. Neurol. 8. doi:10.3389/fneur.2017.0 0 042. 320.e12 N. Cullell, J. Cárcel-Márquez, C. Gallego-Fábrega et al. / Neurobiology of Aging 106 (2021) 320.e320–320.e12 Lucia, D., McCombe, P.A., Henderson, R.D., Ngo, S.T., 2020. Disorders of sleep and Drory, J.S., Shaw, V.E., Turner, P.J., Talbot, M.R., Hardiman, K., Williams, O., Fi- wakefulness in amyotrophic lateral sclerosis (ALS): a systematic review. Amy- fita, K.L., Nicholson, J.A., Blair, G.A., Rouleau, I.P., Esteban-Pérez, G.A., García- otroph. Lateral Scler. Front. Degener 1–9. doi:10.1080/21678421.2020.1844755. Redondo, J., Al-Chalabi, A., Rogaeva, A., Zinman, E., Ostrow, L., Maragakis, L.W., Malhotra, R.K., 2018. Neurodegenerative Disorders and Sleep. Sleep Med. Clin 13, Rothstein, N.J., Simmons, J.D., Cooper-Knock, Z., Brice, J., Goutman, A., Feld- 63–70. doi:10.1016/j.jsmc.2017.09.006. man, S.A., Gibson, E.L., Taroni, S.B., Ratti, F., Gellera, A., Van Damme, C., Rob- Marini, S., Merino, J., Montgomery, B.E., Malik, R., Sudlow, C.L., Dichgans, M., Flo- berecht, P., Fratta, W., Sabatelli, P., Lunetta, M., Ludolph, C., Andersen, A.C., rez, J.C., Rosand, J., Gill, D., Anderson, C.D., 2020. International Stroke Genetics Weishaupt, P.M., Camu, J.H., Trojanowski, W., Van Deerlin, J.Q., Brown, V.M., Consortium, 2020. Mendelian Randomization Study of Obesity and Cerebrovas- van den Berg, R.H., Veldink, L.H., Harms, J.H., Glass, M.B., Stone, J.D., Tien- cular Disease. Ann. Neurol 87, 516–524. doi:10.1002/ana.25686. ari, D.J., Silani, P., Chiò, V., Shaw, A., Traynor, C.E., Landers, B.J., 2018. Genome- Moreno-Grau, S., de Rojas, I., Hernández, I., Quintela, I., Montrreal, L., Alegret, M., wide Analyses Identify KIF5A as a Novel ALS Gene. Neuron 97, 1268–1283.e6. Hernández-Olasagarre, B., Madrid, L., González-Perez, A., Maroñas, O., Rosende- doi:10.1016/j.neuron.2018.02.027. Roca, M., Mauleón, A., Vargas, L., Lafuente, A., Abdelnour, C., Rodríguez- Noyce, A.J., Bandres-Ciga, S., Kim, J., Heilbron, K., Kia, D., Hemani, G., Xue, A., Gómez, O., Gil, S., Santos-Santos, M.Á., Espinosa, A., Ortega, G., Sanabria, Á., Lawlor, D.A., Smith, G.D., Duran, R., Gan-Or, Z., Blauwendraat, C., Gibbs, J.R., Pérez-Cordón, A., Cañabate, P., Moreno, M., Preckler, S., Ruiz, S., Aguilera, N., Hinds, D.A., Yang, J., Visscher, P., Cuzick, J., Morris, H., Hardy, J., Wood, N.W., Pineda, J.A., Macías, J., Alarcón-Martín, E., Sotolongo-Grau, O., Marquié, M., Nalls, M.A., Singleton, A.B., 2019. The Parkinson’s Disease Mendelian Random- Monté-Rubio, G., Valero, S., Benaque, A., Clarimón, J., Bullido, M.J., García- ization Research Portal. Mov. Disord 34, 1864–1872. doi:10.1002/mds.27873. Ribas, G., Pástor, P., Sánchez-Juan, P., Álvarez, V., Piñol-Ripoll, G., García- Ono, K., Mochizuki, H., Ikeda, T., Nihira, T., Takasaki, J., Teplow, D.B., Yamada, M., Alberca, J.M., Royo, J.L., Franco, E., Mir, P., Calero, M., Medina, M., Rábano, A., 2012. Effect of melatonin on α -synuclein self-assembly and cytotoxicity. Neuro- Ávila, J., Antúnez, C., Real, L.M., Orellana, A., Carracedo, Á., Sáez, M.E., biol. Aging 33, 2172–2185. doi:10.1016/j.neurobiolaging.2011.10.015. Tárraga, L., Boada, M., Ruiz, A.GR@ACE consortium, DEGESCO consortium, Peter-Derex, L., Yammine, P., Bastuji, H., Croisile, B., 2015. Sleep and Alzheimer’s dis- Alzheimer’s Disease Neuroimaging Initiative, 2019. Genome-wide association ease. Sleep Med. Rev 19, 29–38. doi:10.1016/j.smrv.2014.03.007. analysis of dementia and its clinical endophenotypes reveal novel loci asso- Sanderson, E., Davey Smith, G., Windmeijer, F., Bowden, J., 2019. An examination of ciated with Alzheimer’s disease and three causality networks: The GR@ACE multivariable Mendelian randomization in the single-sample and two-sample project. Alzheimers Dement. J. Alzheimers Assoc 15, 1333–1347. doi:10.1016/j. summary data settings. Int. J. Epidemiol 48, 713–727. doi:10.1093/ije/dyy262. jalz.2019.06.4950. Shukla, M., Govitrapong, P., Boontem, P., Reiter, R.J., Satayavivad, J., 2017. Mecha- Musiek, E.S., Holtzman, D.M., 2016. Mechanisms Linking Circadian Clocks, Sleep, and nisms of Melatonin in Alleviating Alzheimer’s Disease. Curr. Neuropharmacol 15, Neurodegeneration. Science 354, 1004–1008. doi:10.1126/science.aah4968. 1010–1031. doi:10.2174/1570159x15666170313123454. Nedergaard, M., 2013. Garbage Truck of the Brain. Science 340, 1529–1530. doi:10. Sleep Disorder Genomics Accessed at: March 23,2020 http://sleepdisordergenetics. 1126/science.1240514. org/. n.d. Nicolas, A., Kenna, K.P., Renton, A.E., Ticozzi, N., Faghri, F., Chia, R., Domi- The NHGRI-EBI GWAS Catalog of published genome-wide association studies, tar- nov, J.A., Kenna, B.J., Nalls, M.A., Keagle, P., Rivera, A.M., van Rheenen, W., geted arrays and summary statistics 2019 Accessed at: March 23,2020 https: Murphy, N.A., van Vugt, J.J.F.A., Geiger, J.T., Van der Spek, R.A., Pliner, H.A., //www- ncbi- nlm- nih- gov.sire.ub.edu/pmc/articles/PMC6323933/. Shankaracharya, Smith, Marangi, B.N., Topp, G., Abramzon, S.D., Gkazi, Y., Verbanck, M., Chen, C.-Y., Neale, B., Do, R., 2018. Detection of widespread hori- Eicher, A.S., Kenna, J.D., Mora, A., Calvo, G., Mazzini, A., Riva, L., Mandrioli, N., zontal pleiotropy in causal relationships inferred from Mendelian randomiza- Caponnetto, J., Battistini, C., Volanti, S., La Bella, P., Conforti, V., Borghero, F.L., tion between complex traits and diseases. Nat. Genet 50, 693–698. doi:10.1038/ Messina, G., Simone, S., Trojsi, I.L., Salvi, F., Logullo, F., D’Alfonso, F.O., Cor- s41588- 018- 0099- 7. rado, S., Capasso, L., Ferrucci, M., de Araujo Martins Moreno, L., Kamalakaran, C., Videnovic, A., Abbott, S., 2016. Chronic sleep disturbance and neural injury: links to Goldstein, S., Gitler, D.B., Harris, A.D., Myers, T., Phatnani, R.M., Musunuri, H., neurodegenerative disease. Nat. Sci. Sleep 55. doi:10.2147/NSS.S78947. Evani, R.L., Abhyankar, U.S., Zody, A., Kaye, M.C., Finkbeiner, J., Wyman, S., Videnovic, A., Golombek, D., 2016. Circadian dysregulation in Parkinson’s disease. LeNail, S.K., Lima, A., Fraenkel, L., Svendsen, E., Thompson, C.N., Van Eyk, L.M., Neurobiol. Sleep Circadian Rhythms 2, 53–58. doi:10.1016/j.nbscr.2016.11.001. Berry, J.E., Miller, J.D., Kolb, T.M., Cudkowicz, S.J., Baxi, M., Benatar, E., Tay- Wang, H., Lane, J.M., Jones, S.E., Dashti, H.S., Ollila, H.M., Wood, A.R., van lor, M., Rampersaud, J.P., Wu, E., Wuu, G., Lauria, J., Verde, G., Fogh, F., Tiloca, I., Hees, V.T., Brumpton, B., Winsvold, B.S., Kantojärvi, K., Palviainen, T., Cade, B.E., Comi, C., Sorarù, G.P., Cereda, G., Corcia, C., Laaksovirta, P., Myllykangas, H., Jans- Sofer, T., Song, Y., Patel, K., Anderson, S.G., Bechtold, D.A., Bowden, J., Ems- son, L., Valori, L., Ealing, M., Hamdalla, J., Rollinson, H., Pickering-Brown, S., ley, R., Kyle, S.D., Little, M.A., Loudon, A.S., Scheer, F.A.J.L., Purcell, S.M., Rich- Orrell, S., Sidle, R.W., Malaspina, K.C., Hardy, A., Singleton, J., Johnson, A.B., mond, R.C., Spiegelhalder, K., Tyrrell, J., Zhu, X., Hublin, C., Kaprio, J.A., Kris- Arepalli, J.O., Sapp, S., McKenna-Yasek, P.C., Polak, D., Asress, M., Al-Sarraj, S., tiansson, K., Sulkava, S., Paunio, T., Hveem, K., Nielsen, J.B., Willer, C.J., Zwart, J.- King, S., Troakes, A., Vance, C., de Belleroche, C., Baas, J., ten Asbroek, F., Muñoz- A., Strand, L.B., Frayling, T.M., Ray, D., Lawlor, D.A., Rutter, M.K., Weedon, M.N., Blanco, A.L., Hernandez, J.L., Ding, D.G., Gibbs, J., Scholz, J.R., Floeter, S.W., Redline, S., Saxena, R., 2019. Genome-wide association analysis of self-reported Campbell, M.K., Landi, R.H., Bowser, F., Pulst, R., Ravits, S.M., MacGowan, J.M., daytime sleepiness identifies 42 loci that suggest biological subtypes. Nat. Com- Kirby, D.J.L., Pioro, J., Pamphlett, E.P., Broach, R., Gerhard, J., Dunckley, G., mun 10. doi:10.1038/s41467- 019- 11456- 7. Brady, T.L., Kowall, C.B., Troncoso, N.W., Le Ber, J.C., Mouzat, I., Lumbroso, K., Wienecke, M., Werth, E., Poryazova, R., Baumann-Vogel, H., Bassetti, C.L., Weller, M., Heiman-Patterson, S., Kamel, T.D., Van Den Bosch, F., Baloh, L., Strom, R.H., Waldvogel, D., Storch, A., Baumann, C.R., 2012. Progressive dopamine and Meitinger, T.M., Shatunov, T., Van Eijk, A., de Carvalho, K.R., Kooyman, M., Mid- hypocretin deficiencies in Parkinson’s disease: is there an impact on sleep and delkoop, M., Moisse, B., McLaughlin, M., Van Es, R.L., Weber, M.A., Boylan, M., wakefulness? J. Sleep Res 21, 710–717. doi:10.1111/j.1365-2869.2012.01027.x. Van Blitterswijk, K.B., Rademakers, M., Morrison, R., Basak, K.E., Mora, A.N., Xie, L., Kang, H., Xu, Q., Chen, M.J., Liao, Y., Thiyagarajan, M., O’Donnell, J., Chris- tensen, D.J., Nicholson, C., Iliff, J.J., Takano, T., Deane, R., Nedergaard, M., 2013. Sleep drives metabolite clearance from the adult brain. Science 342, 373–377. doi:10.1126/science.1241224. Zhao, Q., Wang, J., Hemani, G., Bowden, J., Small, D.S., 2019. Statistical inference in two-sample summary-data Mendelian randomization using robust adjusted profile score. Ann. Statist 48 (3), 1742–1769.