No perfect sleep! A systematic review of the link between multidimensional perfectionism and sleep disturbance

The view that perfectionists are prone to experiencing sleep disturbance is widely held. Yet, almost three decades of empirical research have yielded conflicting results. Whereas some researchers viewed perfectionism as a risk factor for sleep disturbance, others spoke of “adaptive” or “positive” forms of perfectionism in the context of sleep. The multidimensional conceptualisation of perfectionism may resolve this disagreement. Thus, this systematic review aimed to clarify the perfectionism‒sleep disturbance link using the widely accepted two‐dimensional perfectionism model, differentiating perfectionistic concerns (defined by worries over imperfections) and perfectionistic strivings (defined by excessively high personal standards). A systematic literature search returned 24 relevant empirical studies. Perfectionistic concerns were robustly linked to sleep disturbance. Perfectionistic strivings displayed comparatively small and inconsistent relations with sleep disturbance. Finally, cross‐sectional mediation studies suggested that psychological distress and dysfunctional cognitive processes might underlie the perfectionistic concerns‒sleep disturbance link. These findings show that considering perfectionistic concerns in explaining, predicting, and treating sleep disturbance may be a promising approach. In contrast, perfectionistic strivings appeared neither universally adaptive nor maladaptive. We identified several critical gaps in the empirical literature and point towards future research directions, highlighting the need for more longitudinal studies.

Additionally, perfectionism is thought to stimulate dysfunctional sleep-interpreting processes (e.g., regarding the consequences of prolonged sleep onset; Lundh & Broman, 2000), which interfere with sleep (Espie, 2002;Harvey, 2002). Perfectionistic persons may also put too much effort into attempting to fall asleep (van der Laar et al., 2010) and try to control a bodily status that cannot be controlled deliberately, which, in turn, exacerbates sleep difficulties (cf. the attention-intention-effort [AIE] pathway model; Espie et al., 2006).
Despite these strong theoretical links and almost three decades of research, a coherent picture of the perfectionism-sleep disturbance link is still missing. Empirical research has identified positive (e.g., Araújo et al., 2017), non-significant (e.g., Akram et al., 2015), and, seldom, negative relations (e.g., Molnar et al., 2020) between perfectionism measures and sleep disturbance indicators. The multidimensional nature of perfectionism may explain this seemingly puzzling heterogeneity in findings. Specifically, the two-dimensional perfectionism model (e.g.,  differentiates two broader dimensions, namely perfectionistic concerns (also labelled evaluative concerns perfectionism) and perfectionistic strivings (also labelled personal standards perfectionism). This perfectionism model is commonly accepted and frequently used for research syntheses (e.g., Burcaş & Creţu, 2020;Dahlenburg et al., 2019;Grugan et al., 2021;Ocampo et al., 2020;Osenk et al., 2020;Robinson & Wade, 2021;Smith et al., 2019Smith et al., , 2021Stricker et al., 2019;Vacca et al., 2021). Previous work showed that, compared to perfectionistic strivings, perfectionistic concerns are more consistently and strongly related to mental health difficulties (e.g., Burcaş & Creţu, 2020;Vacca et al., 2021). Thus, this systematic review used the two-dimensional perfectionism model to integrate the previously inconclusive empirical literature on perfectionism and sleep disturbance, reveal potential underlying mechanisms, and highlight future research directions.

| Multidimensional perfectionism
Perfectionism is a multidimensional personality trait defined as the tendency to hold excessively high standards for oneself paired with overly critical self-evaluations (Flett & Hewitt, 2002). Various measures exist that capture different components of perfectionism, including the Frost Multidimensional Perfectionism Scale (FMPS;Frost et al., 1990), the Hewitt and Flett (1991) Multidimensional Perfectionism Scale (HFMPS), the Almost Perfect Scale-Revised (APS-R; Slaney et al., 2001), the Perfectionism Inventory (PI; Hill et al., 2004), and the Multidimensional Inventory of Perfectionism in Sport (MIPS;; for a review, see Flett & Hewitt, 2015).
A wealth of factor-analytical and conceptual work indicates that two broader dimensions underlie the different perfectionism measures: perfectionistic concerns (characterised by concerns about imperfections) and perfectionistic strivings (characterised by the tendency to set unrealistically high standards for oneself; e.g., Bieling et al., 2004;Dunkley et al., 2006;Frost et al., 1993;Stoeber, 2017). Perfectionistic concerns encompass the impression that others expect perfection of oneself (Hewitt & Flett, 1991), concern over the consequences of making mistakes and doubts about one's competencies (Frost et al., 1990;Hill et al., 2004), a perceived discrepancy between one's standards and actual performance (Slaney et al., 2001), and negative reactions to imperfection . Perfectionistic strivings encompass the belief that oneself needs to be perfect and the tendency to hold overly high personal standards (Frost et al., 1990;Hewitt & Flett, 1991;Hill et al., 2004;Slaney et al., 2001;. The two perfectionism dimensions relate differentially to relevant outcomes. Perfectionistic concerns are robustly associated with maladaptive outcomes (e.g., Smith et al., 2021). In contrast, relations of perfectionistic strivings with maladaptive outcomes are smaller and sometimes inconsistent (e.g., Burcaş & Creţu, 2020). For that reason, perfectionistic strivings have sometimes been labelled "adaptive" or "positive" perfectionism (e.g., Lombardo et al., 2013;Xie et al., 2020). However, these labels are often rejected (e.g., Stoeber, 2017) because perfectionistic strivings may also be maladaptive as they contribute to various forms of psychopathology (Limburg et al., 2017).

| Sleep disturbance
Sleep disturbance can take on various facets, including dissatisfaction with the overall quality of one's sleep, difficulty falling asleep, early morning awakening, and problems in sleep continuity (e.g., difficulty returning to sleep), often resulting in low sleep efficiency (i.e., the ratio of time spent asleep to time in bed) and deficits in the total amount of sleep obtained (sleep duration; e.g., Buysse, 2014;Buysse et al., 2006).
In this systematic review, we focus on these core indicators of sleep disturbance (mostly pertaining to insomnia) rather than parasomnias (e.g., restless legs syndrome). This was because perfectionism is often viewed as a risk factor for difficulties initiating and maintaining restorative sleep, whereas relations with other sleep-related problems (e.g., parasomnias) are not well understood or researched.
Measures of sleep disturbance include retrospective multi-item self-report instruments (Bastien et al., 2001;Buysse et al., 1989), sleep diaries, structured clinical interviews, objective sleep markers (e.g., polysomnography), and single-item scales assessing singular aspects of disturbed sleep (e.g., Leguizamo et al., 2021). Other measures capture daytime correlates of sleep disturbance, such as fatigue, sleepiness, or impaired mood. These daytime correlates are important for measuring the distress and impairment caused by sleep disturbance. However, daytime phenomena are usually seen as potential effects rather than defining components of sleep disturbance (Buysse et al., 2006). Other consequences of chronically disturbed sleep include various health difficulties, including depression (e.g., Baglioni et al., 2011) or metabolic syndrome (Jennings et al., 2007), and neurocognitive deficits (e.g., Durmer & Dinges, 2005).
In the light of these severe consequences of sleep disturbance, the quest for its potential causes, such as perfectionistic tendencies, is a significant public health concern (Barnes & Drake, 2015).

| The link between multidimensional perfectionism and sleep disturbance
A growing body of popular (e.g., Ramlakhan, 2016) and scientific literature (e.g., Bos & Macedo, 2019) links perfectionism to sleep disturbance. In their seminal work on cognitive processes in insomnia, Lundh and Broman (2000) outlined two main pathways through which perfectionistic concerns and perfectionistic strivings (referred to as "perfectionistic standards" by Lundh & Broman, 2000) may contribute to sleep disturbance: perfectionistic strivings are thought to stimulate maladaptive sleep-interpreting processes, characterised by a low tolerance for suboptimal sleep (due to unrealistic standards). Perfectionistic concerns, in turn, are primarily associated with sleep-interfering processes, namely cognitive and emotional arousal stimulated by increased worry, rumination, and stronger negative reactions to adverse life events (Lundh & Broman, 2000). This theoretical assumption about perfectionistic concerns receives support from the substantial associations of perfectionistic concerns with worry, rumination (e.g., Piotrowski, 2019), physiological arousal (Besser et al., 2008), and negative reactions to failure (e.g., van der Kaap-Deeder et al., 2016). Lundh and Broman's (2000) hypotheses about the role of perfectionism in sleep disturbance also blend in well with later, classical cognitive models of insomnia, which highlight the role of arousal and dysfunctional interpretative processes for sleep disturbance (e.g., Espie, 2002;Harvey, 2002).
Further mechanisms may additionally contribute to the chronic sleep disturbance in persons with perfectionistic tendencies. Such mechanisms include a vicious cycle of maladaptive sleep-related cognitions, arousal, and poor sleep, as well as excessive (sleep impairing) efforts to fall asleep (cf. the AIE model; Espie et al., 2006;Akram et al., 2020;van der Laar et al., 2010). Alternatively, the influence of perfectionism dimensions on the development of psychopathology (e.g., depression and anxiety; Smith et al., , 2021 may indirectly stimulate and perpetuate sleep disturbance. Besides the usually assumed influence of perfectionism on sleep disturbance, other mechanisms that link the two constructs are possible. For example, poor sleep may influence personality development (e.g., Stephan et al., 2018), and sleep disturbance and personality may display bidirectional relations (Lau et al., 2021). Alternatively, third variables, such as broader personality dispositions (e.g., conscientiousness) may simultaneously influence the development of perfectionism (Stoeber et al., 2009) and sleep disturbance (e.g., Stephan et al., 2018).
Based on the theoretical links between perfectionism and sleep disturbance, various empirical studies have investigated the associations between the two constructs. This line of research has yielded strongly heterogeneous results, including positive (e.g., Schmidt et al., 2018;Xie et al., 2020), non-significant (e.g., Lin et al., 2019;Trudel-Fitzgerald et al., 2017), and, seldom, negative (e.g., Molnar et al., 2020) relations between perfectionism measures and sleep disturbance. The wealth of perfectionism subscales may have contributed to the current confusion about the perfectionism-sleep link. Hence, in this systematic review, we used the widely accepted two-dimensional perfectionism model. From a scientific and public health perspective, there are several reasons why clarification of the multidimensional perfectionismsleep disturbance link is needed. First, sleep disturbance is increasingly recognised as a cause of poor health (e.g., Anothaisintawee et al., 2016). Thus, identifying risk factors for sleep disturbance is an emerging priority (Barnes & Drake, 2015;Bartel et al., 2015). Second, mean levels of perfectionism in young populations have linearly increased over recent decades (for meta-analyses see Curran & Hill, 2019;Smith et al., 2019). Hence, understanding its detrimental health consequences has become an increasingly relevant concern (e.g., Limburg et al., 2017;Smith et al., 2021). Third, interventions exist that reduce perfectionism (for meta-analyses see Galloway et al., 2021;Lloyd et al., 2015). Thus, perfectionism may be a modifiable risk factor for sleep disturbance. Therefore, it has been suggested to include perfectionism interventions in psychotherapeutic programmes targeting disturbed sleep (e.g., Jansson-Fröjmark & Linton, 2007;Johann et al., 2018;Küskens et al., 2021;Richardson & Gradisar, 2020). Yet, a better understanding of the perfectionism-sleep disturbance link is needed to implement effective interventions.

| The present study
Due to the strong heterogeneity in empirical findings, the link be- We applied the following inclusion criteria: (i) the study includes at least one self-report measure of perfectionistic concerns or perfectionistic strivings. Indicators of perfectionistic concerns and perfectionistic strivings were selected based on widely accepted recommendations Stoeber, 2017).
Specifically, indicators of perfectionistic concerns were the FMPS concern over mistakes and doubts about actions scales, the HFMPS socially prescribed perfectionism scale, the APS-R discrepancy scale, the MIPS negative reactions to imperfection scale, and the PI concern over mistakes scale. Indicators of perfectionistic strivings were the FMPS personal standards scale, the HFMPS self-oriented perfectionism scale, the APS-R standards scale, the MIPS striving for perfection scale, and the PI striving for excellence scale. We also included short forms, adaptations, and translated versions of these scales (e.g., the Child-Adolescent Perfectionism Scale [CAPS]-14; Flett et al., 2016). (ii) The study includes at least one indicator of sleep disturbance (e.g., pertaining to sleep quality, sleep duration, sleep initiation, early morning awakening, sleep efficiency, or sleep continuity). Included sleep disturbance indicators were the Insomnia Severity Index (ISI; Bastien et al., 2001), the Pittsburgh Sleep Quality Index (PSQI; Buysse et al., 1989), the Karolinska Sleep Diary (KSD; Åkerstedt et al., 1994), the Uppsala Sleep Inventory (USI; Liljenberg et al., 1988), the Basic Nordic Sleep Questionnaire (BNSQ; Gislason et al., 1988), self-developed sleep disturbance scales/items, structured clinical interviews, and objective sleep markers. (iii) The study results are reported in English, German, French, or Spanish. (iv) The study reports original quantitative data on the relation (e.g., correlations or group comparisons) between perfectionistic concerns and/ or perfectionistic strivings with sleep disturbance (i.e., no re-analysis of data published more completely elsewhere). When the relevant relations were not reported, we contacted the corresponding author of the respective article via email. Of the six contacted authors, four replied, and two provided the requested information, allowing the inclusion of two additional studies in this systematic review (Table 1).

| Study selection and data extraction
We determined study eligibility in two steps. In Step 1, the first author and the second author independently examined the titles and F I G U R E 1 Flow diagram depicting the study search and inclusion process abstracts of all studies obtained in the literature search (n = 100).
Based on this screening, 46 articles were selected for further evaluation. In Step 2, the first author and the second author inspected the full texts of the remaining articles and independently decided whether to include (n = 24) or exclude (n = 22) an article based on the inclusion criteria. Of the 24 included articles, 23 articles were identified in the systematic database search, and one article was identified in the additional exploratory search (Faber & Schlarb, 2018). Table S1 displays an overview of the studies excluded in Step 2 with reasons for exclusion. In all, 13 studies were excluded because they did not contain a measure of sleep disturbance or perfectionistic concerns/striving (Bartczak & Ogińska-Bulik, 2012;Berglund, 1986;Ellis & Fox, 2004;Frost & Henderson, 1991;Győrffy & Girasek, 2014;Hyman et al., 2002;Kowal & Pritchard, 1990;Palaia et al., 2011;Regen et al., 2015;Sivertsen et al., 2014Sivertsen et al., , 2015Toy, 2009;Wojtowicz & Banez, 2015). Four studies were excluded because the relevant relations were neither reported in the manuscript nor provided via email (Archer et al., 2007;Huang et al., 2020;Lundh & Broman, 2006;Vincent et al., 2004). Two studies were excluded because the data had been reported more completely elsewhere (Azevedo et al., 2009(Azevedo et al., , 2010. Two additional studies were excluded because they were only available in languages not familiar to the study authors (i.e., Chinese and Serbian; Lin et al., 2012;Totić-Poznanović et al., 2012). Finally, one study was excluded because it did not report empirical findings (Brenner et al., 2019).
After determining inclusion/exclusion, the first author coded the country of data collection, sample characteristics, measures, main findings for each perfectionism dimension, and main findings of mediation analyses. The second author independently checked these data. The inter-rater agreement was 92.93% in Step 1 and 97.83% in Step 2. Any discrepancies were resolved by consulting the original articles.

| Study quality assessment
We assessed the methodological quality of all included primary studies. Based on a previously used quality assessment tool (Grugan et al., 2021) and general recommendations for assessing primary study quality (Petticrew & Roberts, 2008), we developed an instrument for quality assessment tailored specifically for the present review. We included the following quality dimensions: (i) measurement of multidimensional perfectionism, (ii) measurement of sleep disturbance, (iii) research design, (iv) sample size, (v) sample design, and (vi) peer-review status. Table S2 displays details of the study quality assessment criteria. We computed an overall methodological quality score (MQS; range: 0-8) from the dimension-specific quality evaluations. The first and second authors independently assessed the methodological primary study quality (95.83% inter-rater agreement). We resolved any disagreements by consulting the original studies. Table 1 displays an overview of the samples, measures, and main findings of the 24 included studies. The median (range) publication

| The link between perfectionistic concerns and sleep disturbance
Perfectionistic concerns were robustly related to sleep disturbance (see seventh column of Table 1). In all, 21 of the included studies assessed with the PSQI subscales) varied more strongly in magnitude and statistical significance than relations with composite scores of sleep disturbance measures (e.g., Brand et al., 2015;Lin et al., 2019;Molnar et al., 2020;Xie et al., 2020).
Only three studies (13%) reported no significant relation between perfectionistic concerns and sleep disturbance (Palo & Das, 2021;Raft, 2012;Vincent & Walker, 2000). In one such study, HFMPS socially prescribed perfectionism was unrelated to a single-item assessment of sleep duration in 419 Indian professionals (Palo & Das, 2021). In a different study, HFMPS socially prescribed perfectionism was negatively related to difficulty staying asleep (after controlling for two other perfectionism scales) in 35 community adults with DSM-IV-text revision (DSM-IV-TR) insomnia (Raft, 2012). Finally, concern over mistakes and doubts were unrelated to sleep quality assessed with a daily diary in 148 government employees (Flaxman et al., 2018). The null findings in the Palo and Das (2021)  This assumption is further supported by the Vincent & Walker (2000) study, which found no significant correlation of concern over mis-

| The link between perfectionistic strivings and sleep disturbance
The empirical findings on relations between perfectionistic strivings and sleep disturbance were highly heterogeneous (see eighth column of Table 1 Table 2 displays an overview of the included mediation studies (n = 8).

| Mediators of the perfectionism-sleep disturbance link
For perfectionistic concerns, dysfunctional cognitive processes, and psychological distress mediated relations with sleep disturbance.
Dysfunctional sleep-related cognitions (e.g., "I am worried that I may lose control over my abilities to sleep.") mediated the effect of FMPS doubts about actions on insomnia severity (Akram et al., 2020), worry and rumination mediated the effect of FMPS concern over mistakes and doubts on poor sleep quality (Flaxman et al., 2018), and worry and rumination mediated the effect of APS-R discrepancy on poor sleep quality (Lin et al., 2019). Regarding psychological distress, anxiety mediated the effect of FMPS doubts about actions on insomnia severity (Akram et al., 2020), perceived stress mediated the effects of APS-R discrepancy and HFMPS socially prescribed perfectionism on sleep disturbance (Molnar et al., 2020) and anxiety, but not depression, mediated the effects of FMPS concern over mistakes and doubts about actions on the probability to experience insomnia (Akram et al., 2017).

Moreover, anxiety mediated the effect of insomnia on doubts about
actions 1 year later (without controlling for baseline levels of doubts about actions; Akram et al., 2015). One additional study (Schmidt et al., 2018) found that a measure capturing a combination of counterfactual thoughts and emotions at bedtime mediated the effects of FMPS concern over mistakes and doubts about actions on insomnia severity.
For perfectionistic strivings, perceived stress mediated the indirect negative effect of APS-R standards and FMPS personal standards on sleep disturbance (Molnar et al., 2020). Additionally, vulnerability to stress mediated the relation of self-oriented perfectionism (critical and striving) with insomnia in female participants, and anxiety mediated the relation of self-oriented perfectionism (critical) on insomnia in male participants (Akram et al., 2015). Importantly, none of the included mediation studies investigated longitudinal processes controlling for baseline levels of perfectionism or sleep disturbance.

| DISCUSS ION
The assumption that perfectionism is closely linked to sleep disturbance is broadly accepted. Yet, previous empirical work on this question has produced conflicting results. In this first systematic review addressing multidimensional perfectionism and sleep disturbance, we clarified the link between the two constructs using the two-dimensional perfectionism model. For perfectionistic concerns (i.e., worrying about the consequences of imperfections), a robust relation with sleep disturbance emerged. For perfectionistic strivings (i.e., setting exceedingly high standards for oneself), comparatively small and inconsistent relations with sleep disturbance emerged. Finally, cross-sectional mediation studies indicated that dysfunctional cognitive processes and psychological distress might underlie the perfectionistic concerns-sleep disturbance link.

| Theoretical and practical implications
The differential relations of perfectionistic concerns and perfec- Persons with high perfectionistic concerns appeared to be prone to experiencing problems falling asleep, problems in sleep continuity, problems regarding sleep duration, and dissatisfaction with their sleep quality. These robust relations with sleep disturbance indicators mirror previous findings that relate perfectionistic concerns to various forms of poor mental (e.g., Limburg et al., 2017) and physical health (e.g., Molnar et al., 2006). Regarding mechanisms underlying the perfectionism-sleep disturbance link, the available cross-sectional mediation studies suggest that dysfunctional cognitive processes and psychological distress are potential mediators.
Notably, not all forms of psychological distress (i.e., anxiety, but not depression; Akram et al., 2017) mediated the perfectionism-sleep disturbance link. These differential patterns may have emerged because anxiety often precedes sleep disturbance, whereas depression is frequently viewed as a consequence of sleep disturbance (Akram et al., 2017;Ford & Kamerow, 1989).
Perfectionistic strivings were only weakly and, mostly, nonsignificantly related to sleep disturbance. Notably, almost all studies that identified significant relations found perfectionistic strivings to be related to more pronounced sleep disturbance (e.g., Lundh et al., 1994). Only one study (Molnar et al., 2020) found negative relations of perfectionistic strivings and sleep disturbance. This relation was only present when perfectionistic concerns and other covariates were controlled for. Thus, as in other contexts (e.g., Limburg et al., 2017), there was little evidence that perfectionistic strivings are "positive" or "adaptive" in the context of sleep.
The studies reporting null-relations of perfectionistic strivings and sleep disturbance used large samples and great width of measures. Hence, the statistically non-significant relations are unlikely to be a methodological artefact. It appears more plausible that perfectionistic strivings have positive and negative effects on sleep disturbance that might neutralise: on the one hand, perfectionistic strivings contain elements of conscientiousness and extraversion (Stricker et al., 2019) that contribute to better sleep (e.g., Stephan et al., 2018) and on the other hand, perfectionistic strivings also contain an element of neuroticism (Stricker et al., 2019) associated with increased sleep disturbance (e.g., Stephan et al., 2018). Previous research shows that contextual factors determine whether adaptive or maladaptive consequences of perfectionistic strivings prevail (e.g., Chang et al., 2008). Also for sleep disturbance, this systematic review identified some hints pointing to required conditions for effects of perfectionistic strivings to emerge. In one study with sleep laboratory patients (Johann et al., 2017), FMPS personal standards were related to sleep disturbance in the first night (i.e., presumably under increased stress due to the unfamiliar sleep laboratory setting), but not in the second night (i.e., presumably under reduced stress). Similarly, in a study with working adults (Ogus, 2006) Anxiety, but not depression, partially mediated the positive effects of concern over mistakes and doubts about actions on the probability to experience insomnia Akram et al.  Hewitt and Flett's (1991) Multidimensional Perfectionism Scale; ISI, Insomnia Severity Index (Bastien et al., 2001); KSD, Karolinska Sleep Diary (Åkerstedt et al., 1994); MIPS, Multidimensional Inventory of Perfectionism in Sports ; PSQI, Pittsburgh Sleep Quality Index (Buysse et al., 1989); USI, Uppsala Sleep Inventory (Liljenberg et al., 1988).
perfectionistic strivings might be particularly vulnerable to sleep disturbance when they experience increased stress. However, in a different study (Molnar et al., 2020), self-reported stress over the last month did not moderate the perfectionistic strivings-sleep disturbance link in university students and community adults.
Regarding differential relations of perfectionistic strivings measures with sleep disturbance, it appears noteworthy that HFMSP self-oriented perfectionism and FMPS personal standards, but not APS-R standards, were sometimes positively related to sleep disturbance. Thus, as in other fields (Blasberg et al., 2016), APS-R standards may capture less maladaptive perfectionism elements in the context of sleep disturbance. More broadly, this finding shows that different subscales used as indicators of the broader perfectionistic concerns/strivings dimensions possess unique correlates and characteristics (also see Smith et al., 2019;Stricker et al., 2019).
The differential relations of perfectionistic concerns and perfectionistic strivings with sleep disturbance also bear some broader implications for cognitive models of sleep disturbance. The cognitions associated with perfectionistic concerns may play a more pronounced role in sleep disturbance than the cognitions associated with perfectionistic strivings. Specifically, the robust and substantial perfectionistic concerns-sleep disturbance link supports the assumed role of worry and rumination in sleep disturbance (Espie, 2002;Harvey, 2002;Lundh & Broman, 2000). However, given the weak perfectionistic strivings-sleep disturbance link, there was less evidence for the assumed importance of exceedingly high personal standards in maladaptive sleep interpreting cognitions (Lundh & Broman, 2000). Having said that, perfectionism scales assess the general tendency to hold exceedingly high standards. Hence, as perfectionism levels may differ between life domains (Haase et al., 2013), persons with sleep difficulties may hold high standards specifically for sleep.
Besides theoretical advances, the present study also bears some practical implications. Yet, we urge to interpret any practical impli- To date, only one study (Maia et al., 2011) has studied longitudinal relations of perfectionism and sleep disturbance with controlling for baseline. However, this study simultaneously entered concurrent and prior sleep disturbance as predictors of multidimensional perfectionism, complicating the interpretation of longitudinal effects.
Additionally, no study has investigated the effects of perfectionism on sleep disturbance after controlling for baseline sleep disturbance.
Longitudinal effects of perfectionistic concerns on mental health difficulties (e.g., Smith et al., 2021) and broader personality traits on sleep disturbance (e.g., Križan et al., 2019) are well established.
Hence, longitudinal effects of perfectionistic concerns on sleep disturbance appear plausible. Yet, sleep disturbance may also shape personality development (Stephan et al., 2018), and third variables, such as environment effects (e.g., through adverse childhood experiences), may influence perfectionism and sleep simultaneously (Krizan et al., 2021 Regarding processes underlying the perfectionism-sleep disturbance relation, the available mediation studies provide only limited insights. To date, no mediation exists that controls for baseline levels. This is problematic because cross-sectional tests of assumed longitudinal processes introduce methodological biases (e.g., Maxwell & Cole, 2007). Another central limitation of current mediation studies lies in the statistical techniques employed to infer mediation. Some mediation studies (e.g., Akram et al., 2015) used simple hierarchical regression models, which have significant shortcomings (i.e., they do not test the assumed indirect effects; Hayes, 2018). Only recently, more advanced, and rigorous statistical techniques have been used to test variables that mediate the perfectionism-sleep disturbance link (e.g., Akram et al., 2020;Molnar et al., 2020). Future longitudinal mediation studies may benefit from applying state-of-the-art techniques for assessing mediation (e.g., Hayes, 2018;Nguyen et al., 2021). This future work could explicitly test assumptions made in the theoretical literature on perfectionism and sleep. For example, mediation studies are needed to test whether the general tendency to hold excessively high personal standards (i.e., perfectionistic strivings) translates into more specific sleep-related maladaptive thought patterns (e.g., regarding the required amount of sleep), and whether these cognitions explain potential effects of perfectionistic strivings on sleep disturbance (Lundh & Broman, 2000). Similarly, future longitudinal studies could test whether the influence of perfectionistic concerns on cognitive and emotional arousal mediates the link with sleep disturbance (Lundh & Broman, 2000). On a more general level, these future mediation studies could also clarify the role of perfectionism within broader cognitive theories of insomnia, e.g., by testing to which degree perfectionism confers a risk factor for hindering sleep interpreting and sleep interfering processes (Espie, 2002;Harvey, 2002).

| Limitations
The present systematic review has some limitations. First, we predominantly relied on the peer-reviewed literature. Thus, publication bias might have distorted our conclusions. Yet, we included some unpublished results, and many of the included studies reported non-significant relations. Second, we did not pre-register the procedure for the systematic review. Third, the samples included in this systematic review were highly heterogeneous. On the one hand, this is a strength of this review. For example, we could show that the relation of perfectionistic concerns with sleep disturbance replicates across student, clinical, and community populations. On the other hand, differences in sample composition might partly explain the heterogeneity in findings, particularly for perfectionistic strivings. Further research using a standardised set of instruments across diverse samples is needed to isolate sample composition and measurement effects. A fourth limitation is our reliance on self-report measures of perfectionism. This approach was inevitable because no validated behavioural perfectionism indicators exist, and no informant-report studies were available.
However, one innovative study has demonstrated increased perfectionistic tendencies (e.g., making additional written comments in psychometric testing) in patients with insomnia compared to healthy controls (Regen et al., 2015). Hence, there is some preliminary evidence that the perfectionism-sleep disturbance link exists beyond self-reports. Yet, further validation of behavioural perfectionism measures is needed to draw robust conclusion. Fifth, we qualitatively rather than quantitatively aggregated findings on the perfectionism-sleep disturbance link. We chose this approach because the broad spectrum of sleep disturbance measures did not allow meaningful statistical aggregation.

| CON CLUS ION
This systematic review showed that two perfectionism dimensions are differentially related to sleep disturbance. Perfectionistic concerns were substantially associated with sleep disturbance. In contrast, perfectionistic strivings were only weakly and inconsistently associated with sleep disturbance. Dysfunctional cognitive processes and psychological distress may mediate the perfectionistic concerns-sleep disturbance link. For neither perfectionism dimension, we found convincing evidence for positive or adaptive effects on sleep. Future longitudinal and intervention research is needed to clarify whether addressing perfectionism may aid in treating sleep disturbance.

CO N FLI C T O F I NTE R E S T
No potential conflict of interest was reported by the authors.

DATA AVA I L A B I L I T Y S TAT E M E N T
All information required to replicate this systematic review are included in the manuscript and Supplementary Information.