Myopia and sleep in children—a systematic review

Abstract

Worldwide, approximately one in three people are myopic or short-sighted. Myopia in children is of particular concern as younger onset age implies a higher risk of progression, and consequently greater risk of developing vision-threatening complications. The importance of sleep in children’s health has long been acknowledged, but evidence for its role in childhood myopia is fairly new and mixed results were presented across studies. To facilitate better understanding of this relationship, a broad literature search, up to and including October 31, 2022, was performed using three databases (PubMed, Embase, and Scopus). Seventeen studies were included in the review, covering four main aspects of sleep, namely duration, quality, timing, and efficiency, and their associations with myopia in children. The present literature review discussed these studies, revealed potential limitations in their methodologies, and identified gaps that need to be addressed in the future. The review also acknowledges that current evidence is insufficient, and the role of sleep in childhood myopia is far from being fully understood. Future studies that primarily, objectively, and accurately assess sleep and myopia, taking other characteristics of sleep beyond duration into consideration, with a more diverse sample in terms of age, ethnicity, and cultural/environmental background, and control for confounders such as light exposure and education load are much needed. Although more research is required, myopia management should be a holistic approach and the inclusion of sleep hygiene in myopia education targeting children and parents ought to be encouraged.

Introduction

Myopia, also known as short-sightedness, is an ocular condition that affects almost one-third of the entire population of the world and that figure is rising [1]. Childhood myopia is of particular concern as the younger the age of onset, the higher the risk of progression, and consequently the greater the risk of developing high myopia [2]. High myopia is often accompanied by pathological changes or complications that can lead to vision impairment and blindness [3]. Several risk factors for myopia development have been well documented, amongst which, outdoor activities were found to be protective against myopia onset in children [4]. Although the underlying mechanism for this protective effect remains to be elucidated, it appears that bright, full-spectrum outdoor light might be one of the critical contributors [5, 6]. Additionally, there is a seasonal variation and a diurnal pattern of eye length and refractive error [7–10], potentially signaling a role of circadian rhythm in myopia [11].

Circadian rhythms, which regulate the sleep–wake cycle, play a crucial role in sleep. Evidence suggests that outdoor play is associated with better sleep outcomes in children as it can facilitate the regulation of melatonin secretion and circadian rhythm and promote regular sleep onset [12]. Sleep is vital for children’s health, with adequate sleep a prerequisite for optimal daily function, physical, and mental health. Poor sleep has been associated with many health problems such as childhood adiposity and psychosocial difficulties [13].

Does sleep affect eye health of children? Although the significance of sleep in the sphere of children’s health has long been acknowledged, discoveries of evidence for its role in childhood myopia are relatively recent and the findings are rather mixed. The inclusion of sleep hygiene in ocular health promotion messages and myopia management strategies has therefore been relatively scarce. Additionally, sleep is essentially a multidimensional concept with several aspects or dimensions that need to be taken into consideration when evaluating its impact on health outcomes [14]. The present systematic review will discuss recent findings regarding the relationship between sleep and childhood myopia, examine possible limitations in methodologies of these studies, and identify potential gaps for future research.

Method

A literature search following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (the PRISMA flowchart [15]) was conducted on peer-reviewed scientific journal articles using databases PubMed, Embase, and Scopus on October 31st, 2022. Search terms and filters are shown in Figure 1 (detailed information on paper identification process is presented in Supplementary Appendix 1). After combining search results from all three databases and removing duplicates, each article was evaluated to determine its inclusion in the review. Exclusion criteria applied during the selection process were: (1) book chapters, reviews, comments, letter, or conference abstracts, (2) inappropriate study population: mean age of study participants beyond 18 years old (except for study reported by Stafford-Bell et al. [16]., in which participants were followed up for 20 years from birth); orthokeratology lens wearers; population with other disorders such as idiopathic scoliosis and keratoconus, etc., (3) inappropriate study outcome measures: lack of clinical assessment on refractive error (i.e. a number of studies were excluded as participants’ refractive status were determined via questionnaires and/or only visual acuity was measured), and (4) other reasons: e.g. no full text available. Primary search and screening were conducted by XL and reviewed by PS and TN. A consensus was reached through discussions among all three authors. A total of 17 articles were deemed eligible and subsequently included in the present review.

Figure 1.

Literature search and publication selection process.

Results

A summary of these 17 studies is presented in Tables 1 and 2. Descriptive information on study type (population or clinic-based), sample, and myopia is included in Table 1, while sleep evaluation methods and related findings are shown in Table 2. Significant variations in methods and findings across these studies were revealed.

Table 1.

Summary of Descriptive Information on Study Type, Sample, and Myopia

Year	Author	Study type	Sample description				Myopia
			N	Ethnicity	Age (Years, mean ± SD)	Gender (Female, %)	Definition	Cycloplegic Refraction	Prevalence (n, %)
2015	Zhou et al. [17]	Population based	1902	Chinese	9.80 ± 0.44	879, 46.2%	SE ≤ −0.50D in both eyes	Yes	588, 30.9%
2015	Hua et al. [18]	Population based	317	Chinese	Intervention: 10.7 ± 2.4; Control: 10.5 ± 2.3	Intervention: 51.1%; Control: 53.2%	SE ≤ −0.50D in the right eye	Yes	91, 46%
2016	Ayaki et al. [28]	Clinic based	278	Japanese	14.2 ± 2.6	56%	SE ≤ −0.50D in the worse eye	No	177, 51.5%
2016	Jee et al. [29]	Population based	3625	Korean	15.5 (0.0) (standard error)	52.9%	SE ≤ −0.50D in the right eye	No	2895, 78.7%
2018	Sensaki et al. [19]	Population based	376	Chinese, Malay, and Indian	36 months at eye examination	**	SE ≤ −0.5D in the right eye	Yes	13, 3.5%
2019	Pan et al. [20]	Population based	2436	Chinese	13.8	**	SE < −0.5D	Yes	693, 29.5%
2020	Wei et al. [21]	Population based	2328	Chinese	7.09 ± 0.41	980, 42.10%	SE < −0.5D	Yes	**
2020	Ostrin et al. [25]	Clinic based	91	**	13.02 ± 1.37	53%	mean SE of both eyes ≤ –0.50D and one eye ≤ −0.75D	No	36, 39.6%
2020	Liu et al. [22]	Population based	6042	Chinese	7.36 ± 0.60	2835, 46.93%	SE ≤ −0.50D in the right eye	Yes	409, 6.77%
2021	Lu et al. [34]	Population based	556	Chinese	10.21 ± 0.89	**	SE < −0.50D in at least one eye	No	354, 63.7%
2021	Stafford-Bell et al. [16]	Population based	1194	Caucasian, East Asian, South Asian, Other/Mixed	20.0 ± 0.5 at 20-year follow up	581, 48.7%	SE ≤ −0.50D in either eye	Yes	303, 25.4%
2022	Cai et al. [30]	Clinic based	115	Chinese	9.60 ± 2.30	49, 42.6%	N/A	No	N/A
2022	Li et al. [23]	Population based	572	56.1% Chinese	Aged 9 years at analysis	289, 50.5%	SE ≤ −0.50D	Yes	**, 37.3%
2022	Wang et al. [26]	Population based	15765	Chinese	Range: 6–18 years	**	SE ≤ −0.50D	No	6758, 59.54%
2022	Saara et al. [31]	Population based	3850	Indian	15.08 ± 1.23	3231, 83.92%	SE < −0.50D	No	752, 19.53%
2022	Ma et al. [32]	Population based	913	Chinese	8.8 ± 2.9	456, 50%	SE <-0.50D and LogMAR VA <5.0D	No	152, 16.6%
2022	Peng et al. [33]	Population based	6154	Chinese	** (middle school students)	2868, 46.6%	VA <6/6 (Snellen) and SE < −0.50D	Yes but only when VA under 6/6	4035, 65.5%

** indicates no data available.

Table 2.

Summary of Sleep Evaluation Methods and Related Findings

Year	Author	Sleep evaluation methods		Findings	Overall bias*
		Instrument	Respondent (If applicable)
2015	Zhou et al. [17]	Children’s Sleep Habits Questionnaire (CSHQ)	Parents	Odds of myopia increased with worse CSHQ score (OR = 1.01 per point, 95% CI [1.001, 1.02], p = 0.014). The association between total sleep duration (night + midday) and myopia was not significant.	Low
2015	Hua et al. [18]	Interview with questionnaire	Participants and parents	Among non-myopic students, shorter sleeping hours associated with greater myopic SE change (p = 0.039) and greater axial elongation (p = 0.006).	Low
2016	Ayaki et al. [28]	Pittsburgh Sleep Quality Index (PSQI), Hospital Anxiety and Depression Scale (HADS), selected questions from the Morningness/Eveningness questionnaire	Participants	Myopic refractive error was correlated with poor PSQI score (p < 0.05), shorter sleep duration (p < 0.01) and later bedtime (timing) (p < 0.01).	Medium
2016	Jee et al. [29]	Health interview	Participants	Risks for myopia decreased by 10% per hour increase in sleep duration (p = 0.012), and refractive error increased by 0.1D per hour increase in sleep duration (p = 0.004).	Medium
2018	Sensaki et al. [19]	Brief Infant Sleep Questionnaire (BISQ)	Parents	Total sleep duration and number of night-wakings at 12 months were not associated with SE at 3 years. Infants with longer total sleep duration (at least 12.5 hours) were more likely to have a greater axial length (β: 0.20; 95% CI [0.06, 0.36]; p = 0.006)	Low
2019	Pan et al. [20]	Children’s Sleep Habits Questionnaire (CSHQ)	Parents	The odds ratio of myopia increased amongst those with disordered sleep before propensity score matching (OR: 1.43, 95% CI [1.05, 2.58], p = 0.01). After matching, the association was not significant.	Low
2020	Wei et al. [21]	Questionnaire	Parents	Sleep duration was not significantly associated with myopia progression or axial elongation for all children. No significant difference between bedtime (timing) levels in terms of myopia progression or axial elongation.	Low
2020	Ostrin et al. [25]	Wearable device (Actiwatch 2)	n/a	Neither refractive error change nor axial elongation were correlated with any of the sleep parameters tested include Mean Daily Bedtime, Wake Time (timing), Sleep Duration, Sleep Latency, and Sleep Efficiency. Myopic children exhibited significantly shorter sleep latency (p = 0.04) and more variability insleep duration by day of the week (p < 0.001) and season (p = 0.007).	Medium
2020	Liu et al. [22]	Questionnaire	Parents	Later bedtime (timing) was identified as a risk factor for myopia prevalence at baseline (odds ratio [OR] = 1.55, p = 0.04), 2-year myopia incidence (odds ratio [OR] = 1.44, p = 0.02) and progression over 24 months (p = 0.005). sleep duration did not differ between myopes and non-myopes at baseline (9.49 vs. 9.47 h, p = 0.6) or at 24-month visit (9.16 vs. 9.18 h, p = 0.6).	Low
2021	Lu et al. [34]	Questionnaire	Participants	Longer sleep duration (>8 h) was associated with higher prevalence of myopia (p < 0.05).	Medium
2021	Stafford-Bell et al. [16]	Child Behavior Checklist questionnaire (CBCL)	Parents	No significant association was found between 12-year sleep problem trajectories and changes in refractive error, axial length or corneal radius.	Low
2022	Cai et al. [30]	Questionnaire	Participants and parents	The monthly axial growth rate was negatively correlated with sleep duration (r = −0.197, p = 0.040).	Medium
2022	Li et al. [23]	Children’s Sleep Habits Questionnaire (CSHQ)	Parents	Sleep quality, duration, timing, and the consistency of specific sleep factors were not independently associated with myopia, SE, or AL.	Low
2022	Wang et al. [26]	Questionnaire	Parents	Shorter sleep duration (Less than 7 h per day) was a risk factor for high myopia (OR: 9.789, 95% CI [6.865 to 13.958]).	Medium
2022	Saara et al. [31]	Questionnaire	Participants	Shorter sleep duration (less than 7 h per day) were associated with myopia when compared to sleeping more than 7 h per day (OR: 2.77, 95% CI [2.03 to 3.79], p < 0.001).	Medium
2022	Ma et al. [32]	Questionnaire	Participants and parents	Longer sleep duration associated with less myopia progression.	Medium
2022	Peng et al [33].	Questionnaire	Participants and parents	Sleeping early (timing) was associated with lower prevalence of myopia, while duration was not associated.	Low

* Overall bias scoring: clinic based scores 1; sample size less than 100 scores 1; no cycloplegia scores 1; using subjective methods (questionnaire or interview) scores 1. Scores of 0/4 and 1/4 indicate low level of bias, 2/4 and 3/4 indicate medium level of bias, and 4/4 indicates high level of bias.

First, methods of defining myopia differed between studies. As presented in Table 1, when determining the amount of refractive error for each participant, cycloplegia was only used in nine (9/17) studies [16–23]. Lack of cycloplegia in children refractive studies can lead to misclassification of myopia by introducing bias towards greater myopia [24].

Secondly, instruments used to evaluate sleep also varied across studies. Table 2 showed that a wide range of questionnaires (i.e. Children’s Sleep Habits Questionnaire, CSHQ; Brief Infant Sleep Questionnaire, BISQ; Pittsburgh Sleep Quality Index, PSQI; Child Behavior Checklist questionnaire, CBCL; and other customized questionnaires) were employed in 16 of 17 studies to evaluate sleep, while only one study used an objective method (wearable devices) [25]. Moreover, among studies using questionnaires, parents were the main respondent in many studies (8/16) [16, 17, 19–23, 26].

Lastly, definition of study variables is inconsistent across studies. For example, of studies that examined sleep duration, only a few analyzed the total amount of sleep accumulated over 24 hours (night sleep plus daytime naps) [17, 19, 23]. Other studies, in contrast, either only used nighttime sleep hours or did not specify the classification of sleep duration.

The frequency of each sleep-related variable reported amongst the selected studies is demonstrated in Figure 2. Four main categories namely duration, quality, timing, and efficiency were adapted from the sleep dimensions proposed by Buysse [14]. Within the scope of the present review, duration refers to the amount of time spent on sleep (either overnight or over 24 hours); quality is subjectively rated “good” or “poor” sleep; timing refers to the clock time when participants go to sleep and/or wake up; efficiency is the ease with which participants initiate and maintain sleep, where sleep onset latency (time attempting to fall asleep) and night-wakings are both critical components in the measurement [27]. Other unspecified sleep problems that were treated as the outcome variable in the study is also listed separately [16]. In summation, myopia was found to be associated with duration, quality, and timing of sleep, although findings were sometimes conflicting between studies.

Figure 2.

Count of sleep variables reported in reviewed studies.

Duration

Sleep duration was the most frequently studied variable, where 15 of the 17 publications reported results on this variable’s relation with myopia (Figure 2). However, those results were mixed (Figure 3). Of these 15 studies, seven associated increased risk and/or severity of myopia with shorter sleep duration [18, 26, 28–32]. Two of these studies reported sleeping less than 7 hours per day increased the odds of myopia [26, 31], the other five studies identified greater odds of myopia, amount of myopic refractive error, and eye growth or axial elongation rate with shorter sleep duration [18, 28–30, 32].

Figure 3.

Association of shorter sleep duration with greater myopia.

In contrast, six studies did not find any relationship between sleep duration and myopia (Figure 3) [19, 21–23, 25, 33], although one infant study found a correlation between longer sleep hours and greater axial length that was not associated with refractive error [19].

Only two studies reported associations of longer sleep hours with higher odds and prevalence of myopia [17, 34], but such association became insignificant in one study when nighttime sleep duration was replaced by total sleep hours (night sleep plus daytime naps) [17].

Mean sleep hours were specified in six studies [17, 19, 21, 22, 25, 29], while other studies used categorical cutoffs. Interestingly, amongst the six studies that did not find associations between duration and myopia, participants in four studies (4/6, 66.7%) had sufficient sleep according to the recommendations of the American Academy of Sleep Medicine [19, 21–23]; whereas participants of 2/2 studies (data unavailable for five studies) that reported sleep duration and myopia associations had insufficient hours of sleep [29, 32, 35].

Quality

Subjectively rated sleep quality measured with CSHQ and PSQI was investigated in four studies [17, 20, 23, 28]. Three reported a relation between poor sleep quality and increased odds (odds ratio = 1.01 and 1.43, respectively) and amount of myopia [17, 20, 28], although the association was no longer significant after propensity score matching in one study [20]. Poor sleep quality is quite prevalent across study samples. In the two studies using CSHQ, 59.4% and 78.3% of participants had an average score of 41 or higher [17, 20], indicating problematic or disordered sleep [36]. Even for the study that did not find any significant association between poor sleep and myopia, their study sample scored an average of 45.8 (SD = 6.2) with CSHQ [23]. In the study that used PSQI, participants scored 3.51 (SD = 2.44) on average [28], where scores less than five indicate poor sleep [37].

Timing

Six studies investigated the timing of sleep, namely the clock time when sleep is initiated or bedtime, and the clock time when sleep ended or wake time [21–23, 25, 28, 33]. In three studies, late bedtime was associated with higher prevalence and greater degree, progression, and incidence of myopia [22, 28, 33]. Among the three studies, participants with a higher degree of myopia were found to have a later bedtime (74 minutes later than non-myopes) [28]; among children 6 to 9 years old, those who slept later than 09:30 pm (compared with sleeping before 09:00 pm) had higher odds of being myopic at baseline (odds ratio = 1.55), higher 2-year myopia incidence (odds ratio = 1.44), and greater progression over 24 months [22]; and sleeping after 11:00 pm (compared with before 10:00 pm) was associated with higher prevalence of myopia in both urban and rural middle schools’ students (odds ratio = 1.78 and 2.0, respectively) [33]. In contrast, the other three studies did not reveal any significant associations between myopia and the timing of sleep [21, 23, 25]. Nevertheless, these studies, as pointed out by the authors, had skewed distribution of bedtime where 88.7% of the children’s bedtime fell between 08:30 and 10:00 pm [21], small sample size (less than 100), and relatively narrow refractive error range (myopes’ mean SE –2.44 ± 1.52 D) [25].

Efficiency

Sleep efficiency was discussed in two studies, one study used questionnaire while the other employed objective measures (a wristworn actigraphy device with light sensor) [19, 25]. The former found that the number of night-wakings at the age of 1 year were not associated with SE at the age of three years, and the later reported no significant difference detected with the wearable device in sleep efficiency between myopes and non-myopes. However, the wearable device study identified significantly shorter sleep onset latency, a critical component in the measurement of sleep efficiency, by approximately 3.8 minutes amongst myopic children compared to non-myopes (p = 0.04) [25].

Discussion

A total of 17 studies were identified for inclusion, covering different features of sleep, namely duration, quality, timing, and efficiency. Although findings were mixed and a causal relationship between poor sleep and myopia cannot be concluded from current evidence, there are implications for an association of poor sleep with childhood myopia. Specifically, insufficient sleep hours (in 7/15 studies), poor sleep quality (in 3/4 studies), and late bedtime (in 3/6 studies) were reported associated with greater myopia. Meanwhile, in spite of null findings concerning sleep efficiency, significantly shorter sleep latency was observed among 10 to 15 years old myopes in a longitudinal study conducted in Australia [25], which may reflect greater underlying physiological sleepiness in children with myopia [38]. Additionally, this study reported significant variability in sleep–wake patterns between weekdays and weekends in myopes compared to non-myopic children [25]. Such variability is often recognized as an indicator for “social jetlag,” a term used to describe the discrepancy between social and biological time [39]. These findings allude to a relationship between disturbances to children’s body clock, the circadian rhythm, and myopia.

As a component of the sleep–wake cycle that is mediated by the circadian rhythm, sleep is, in fact, a multidimensional construct that is more than mere duration. Evidence supporting the involvement of sleep or circadian rhythm in myopia is gradually coming to light. A considerable body of laboratory animal studies have demonstrated that regular light–dark cycles are vital for normal ocular growth, and alterations to this cycle would result in refraction shifts [40–44]. With the advancement in the study of chronobiology through identification of specific genetic loci that control the circadian rhythm in flies and mammals [45], researchers further tested the impact of disruption to these clock genes on optical development of the eye with lab animal models. For example, a significant association between retinal clock disturbance and myopia was revealed in one study, where retinal-specific clock gene (Bmal1) knock-out mice developed myopia with an elongated vitreous chamber [46]. Furthermore, a meta-analysis of genome-wide association studies located more than 300 genetic loci for refractive error, within which, genes regulating circadian rhythms were also identified [47].

The role of circadian clock in myopia has also been explored in studies examining the association between melatonin and myopia. Melatonin is a hormone released in dim or the absence of light that promotes sleep and contributes to circadian entrainment to the solar day. Dim light melatonin onset (DLMO) has therefore been widely recognized as a circadian phase marker. Photic suppression of melatonin release is facilitated by a group of intrinsically photosensitive retinal ganglion cells (ipRGCs) in human retina. Exposure to bright light, particularly to light that has a peak wavelength close to the peak sensitivity of ipRGCs (~482 nm), would trigger melatonin suppression [48]. Significant differences in circadian timing (measured with DLMO) between myopic and non-myopic adults have been reported. Compared with non-myopic participants, myopes were found to exhibit a DLMO phase delay of more than 1 hour, and lower urinary melatonin output overnight [49].

It is also worth noting that there are, nonetheless, confounding issues concerning the relationship between myopia and sleep. For example, while myopic children were found to spend much less time in darkness (≤1 lux, most likely sleeping) compared to non-myopic children, they also spent significantly less time outdoors [50]. Time outdoors, the protective factor against myopia has also been positively associated with better sleep outcomes (i.e. earlier bedtime, increased sleep duration, and improved sleep efficiency) in children [5, 12, 25]. Could insufficient time outdoors simultaneously be the trigger and the link between poor sleep and myopia? To answer this question, future studies need to precisely measure, control, and account for different light conditions to tease out the confounding effect of light. Additionally, education load is also one of the major confounders that need to be taken into consideration. It is a known risk factor for myopia [5] and has been reported to correlate with bedtime in students [33].

Limitations and gaps

As detailed in Table 2, findings were sometimes conflicting between studies. Factors such as inappropriate or insensitive outcome measures, varied definitions of study variables, and less diverse sample selections, could have contributed to these inconsistencies.

To begin with, methods evaluating outcome variables (i.e. myopia and sleep) varied across studies. For myopia measurements, in order to ensure the accuracy when performing refraction on children, cycloplegia is required and considered the “gold standard.” Non-cycloplegic refractions often overestimate myopia in this young population with very active accommodation [51]. Of the 17 studies reviewed, cycloplegia was only performed in eight studies (8/17, 47%). Meanwhile, a variety of sleep evaluation methods were employed in these studies and different focus of each instrument could have contributed to the inconsistencies observed. Subjective methods, such as questionnaires or interviews, are inevitably prone to recall bias. Since participants of these studies were children and mostly young children, who might have difficulties answering questions accurately, the majority of the questions (either in questionnaires or interviews) were answered by parents (in eight studies) or by parents and children together (in four studies). Therefore, the accuracy of the information collected via this method may be compromised if the children were not properly consulted when filling in the questionnaires or if the children’s answers were influenced by parents or other factors. Actigraphy used in one study [25], on the other hand, is an objective method that can improve the sensitivity of assessments. But it has been reported to have poor specificity in detecting wake after sleep onset, which could lead to over or underestimates of sleep parameters (e.g. efficiency of sleep) [52].

Second, definition of study variables differed between studies. As mentioned in the results section, where sleep duration was concerned, some studies investigated only nighttime sleep while some used total amount of sleep accumulated over 24 hours. In addition, although only two studies examined sleep efficiency and both reported null findings in the present review, it has been suggested that inconsistencies in the formulas used to calculate sleep efficiency across sleep studies need to be addressed [27].

Additionally, whether study samples had sufficient sleep corresponding to their age could be another contributor to current mixed findings. A recent cohort study identified associations of refractive error with both too much and too little sleep, which revealed a U-shaped association between sleep duration at the age of 2 years and glasses wear (types of refractive error were unspecified) at 5 years old [53]. In the present review, where data were available, participants in studies that failed to identify any association between sleep duration and myopia had sufficient sleep, while participants in other studies reporting association between shorter sleep and more myopia did not have sufficient sleep.

Furthermore, the samples of studies reviewed in the present article varied considerably in terms of demographic features and behavioral or environmental factors. Sleep patterns change greatly with age and cultural contexts [54]. Conflicting findings have been revealed between studies with a large sample size, wide range of age and refractive error versus studies with relatively small sample or narrow distributions of age and refractive profiles [21–23, 25, 28, 33]. A study comparing sleep habits between schoolchildren (age range 7 to 13 years) in China and in the United States, found that Chinese children had shorter sleep duration and later bedtime compared to American children [55]. Meanwhile, prevalence of myopia is also differing vastly across age groups and regions/ethnicity [2], which again brings in confounding issues and adds to the complexity to illustrate a clear picture of the relationship between sleep habits and myopia.

Last but not the least, sleep measurements were in fact not the primary outcome in many of the reviewed studies (7/17 studies) [18, 26, 30–34], and they might not be precisely designed to tackle sleep health. Therefore, a causal relationship between poor sleep and myopia cannot be established based on current evidence.

Conclusion

In summation, the relationship between myopia and sleep have been explored in many studies but it is far from being concluded. While findings were sometimes conflicting, there is implication for an association of myopia with poor sleep (i.e. insufficient hours, poor quality, and irregular or late timing). Potential methodological limitations of studies that might have contributed to the inconsistencies in findings observed include inappropriate or insensitive outcome measures, varied definitions of study variables, and less diverse sample selections. Future studies that primarily, objectively, and accurately assess sleep and myopia, taking other characteristics of sleep beyond duration into consideration, with a more diverse sample in terms of age, ethnicity, and cultural/environmental background, and control for confounders such as light exposure and education load are much needed. An understanding of sleep as an essential component of the circadian rhythm is also necessary to better interpret the findings. Additionally, a universal definition of critical study variables, such as sleep duration and efficiency, could better facilitate the connection of findings between studies.

Nevertheless, myopia management should be a holistic approach and the inclusion of sleep hygiene, namely early bedtime, regular and sufficient sleep, in ocular health promotion messages and myopia management strategies targeting children should be encouraged.

Funding

This research has been supported by an Australian Government Research Training Program (RTP) Scholarship, the Dr David Wilson Memorial Scholarship, and the Brien Holden Vision Institute Limited.

Disclosure Statement

None declared.