Daily axial length and choroidal thickness variations in young adults: Associations with light exposure and longitudinal axial length and choroid changes

Journal Pre-proof Daily axial length and choroidal thickness variations in young adults: Associations with light exposure and longitudinal axial length and choroid changes Sekar Ulaganathan, Scott A. Read, Michael J. Collins, Stephen J. Vincent PII:

S0014-4835(19)30425-7

DOI:

https://doi.org/10.1016/j.exer.2019.107850

Reference:

YEXER 107850

To appear in:

Experimental Eye Research

Received Date: 11 June 2019 Revised Date:

11 October 2019

Accepted Date: 18 October 2019

Please cite this article as: Ulaganathan, S., Read, S.A., Collins, M.J., Vincent, S.J., Daily axial length and choroidal thickness variations in young adults: Associations with light exposure and longitudinal axial length and choroid changes, Experimental Eye Research (2019), doi: https://doi.org/10.1016/ j.exer.2019.107850. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Ltd.

1

Daily axial length and choroidal thickness variations in young adults: Associations with light exposure and longitudinal axial length and choroid changes Sekar Ulaganathan, Scott A. Read, Michael J. Collins, Stephen J. Vincent.

Affiliation: Contact Lens and Visual Optics Laboratory, School of Optometry and Vision Science, Queensland University of Technology, Brisbane, Australia.

Authors Email: Sekar Ulaganathan: [email protected] Scott A. Read: [email protected] Michael J. Collins: [email protected] Stephen J. Vincent: [email protected]

Corresponding author: Sekar Ulaganathan Contact Lens and Visual Optics Laboratory, School of Optometry, Queensland University of Technology, Room B559, O Block, Victoria Park Road, Kelvin Grove 4059 Brisbane, Queensland, Australia. Phone: 617 3138 5705, Fax: 617 3138 5880

Declarations of interest: none

Word count: 6851

2

Abstract Evidence from animal studies suggests that the eye’s natural diurnal rhythms can be disrupted by altering the light/dark cycle or during refractive error development. Although diurnal variations in axial length (AL) and choroidal thickness are well documented in human eyes, the relationship between ambient light exposure, refractive error progression and diurnal AL and choroidal thickness variations is not well understood. Therefore we examined the association between objective ambient light exposure and daily variations in AL and subfoveal choroidal thickness (SFCT), along with longer-term changes in AL and SFCT over 12 months. Thirty-four young adult emmetropes and myopes had their daily variations (measurements ~every 3 hours from 9 am to 9 pm) in AL and SFCT assessed on a weekday and weekend in winter and then six months later in summer. All participants returned six months later for a single measurement session to determine the longer-term change in AL and SFCT. Personal ambient light exposure was captured in winter and summer using wrist-worn light sensors (Actiwatch-2) worn for 14 days over the same period of time when the diurnal measurements were collected. Linear mixed model analyses revealed significant daily variations in AL and SFCT (each p<0.05). The mean daily peak to trough difference (amplitude) in AL was significantly greater in myopes (0.020 mm; 95% CI: 0.014 to 0.026 mm) compared to emmetropes (0.010 mm; 95% CI: 0.005 to 0.015 mm) (p<0.01), but the SFCT variations were not significantly different between the refractive groups (p=0.45). Daily variations in AL were negatively associated with the daily SFCT variations (r = -0.603, p<0.001). Correlation analyses indicated that the amplitude of daily AL variations was negatively associated with the daily time exposed to bright light (r = -0.511, p=0.002) and positively associated with the longitudinal AL changes over 12 months (r = 0.381, p=0.04). There was an inverse association between the longer-term changes in AL and SFCT (r = -0.352, p=0.002). The daily ocular diurnal variations were not significantly different between weekdays and weekends, or between summer and winter (each p>0.05). In summary, diurnal variations in AL were higher in amplitude in myopes compared to emmetropes and were also associated with longitudinal changes in AL. These findings suggest that diurnal variations may be associated with longer-term axial eye growth. Time spent in bright light also significantly influenced the amplitude of daily AL variations, with more time exposed to bright light associated with a smaller amplitude of diurnal AL change. Choroidal thickness exhibited an inverse association with the AL changes, implying a potential role for the choroid in eye growth. Keywords: Diurnal variations, Light exposure, Axial length, Choroid, Myopia progression.

3

1. Introduction: Diurnal variations are periodic rhythms that occur in many of the body’s structures and physiological processes over the course of the day. A range of ocular parameters undergo diurnal variations, including axial length (AL) and choroidal thickness (Wilson et al. 2006; Read et al. 2008; Usui et al. 2012). AL is the primary biometric determinant of refractive error and demonstrates a diurnal variation typically characterised by a peak in AL during the day and a trough at night with an amplitude of change of ~25-46 µm over 24 hours (Stone et al. 2004; Wilson et al. 2006; Read et al. 2008; Chakraborty et al. 2011; Burfield et al. 2018). These diurnal AL rhythms appear consistent when measured on two consecutive days (Chakraborty et al. 2011), however other studies have noted larger between-day differences in diurnal variations across weeks or months (Stone et al. 2004; Wilson et al. 2006), which could potentially be due to seasonal variations (Chakraborty et al. 2011). However, the effects of seasons upon these diurnal ocular variations have not been studied in humans. Recent evidence also emphasises the role of the choroid in eye growth, as animal studies demonstrate that a thinning of the choroid occurs during myopia development, which precedes changes in scleral growth (Wallman et al. 1995). Human studies have also documented an association between longitudinal choroidal thickness changes and axial eye growth, with less choroidal thickening or choroidal thinning observed in children undergoing faster axial eye growth (Read et al. 2015a; Fontaine et al. 2017). Choroidal thickness also varies diurnally, typically in antiphase to AL variations, being thicker at night and thinner during the day (Chakraborty et al. 2011; Tan et al. 2012; Usui et al. 2012; Seidel et al. 2015; Kinoshita et al. 2017; Burfield et al. 2018). Animal studies provide evidence suggesting that diurnal rhythms may play a role in the regulation of eye growth (Nickla 2013). Form-deprivation stimulates excessive eye growth resulting in myopia development and form-deprived chicks developing myopia also exhibit altered (magnitude increased) diurnal rhythms in AL and choroidal thickness (Weiss & Schaeffel 1993; Wallman et al. 1995; Nickla et al. 1998b; Papastergiou et al. 1998). On the other hand, the diurnal variations of AL are phase delayed and the choroidal rhythm is phase advanced (shifting the two rhythms into phase with each other) when chicks are exposed to myopic defocus (a stimulus to slow/retard eye growth) (Nickla et al. 1998b; Papastergiou et al. 1998; Nickla 2006). The influence of short-term (12-hours) monocular defocus upon the pattern of diurnal ocular variations in humans has also been investigated. Monocular hyperopic defocus resulted in an

4

increase in the amplitude of daily variations in young adults, with no phase shift (Chakraborty et al. 2013). On the other hand, imposing monocular myopic defocus reduced the amplitude and also delayed the phase in AL and advanced the phase in choroidal thickness (Chakraborty et al. 2012), consistent with animal studies. Although an effect of defocus upon daily ocular changes suggests the potential for daily ocular variations to be associated with changes in eye growth, the relationship between short-term diurnal changes in AL and choroidal thickness and longer-term changes in eye growth in human eyes has not been previously studied. Environmental light information (light/dark cycle) drives many of the body’s diurnal rhythms (LeGates et al. 2014). Altering the normal light/dark cycle also alters normal refractive development in animal models (Nickla 2013), which supports the notion that interruptions to ocular diurnal rhythms leads to changes in ocular growth and the development of refractive error in animals. Recent evidence from human and animal studies suggests that ambient light exposure also plays an important role in the regulation of eye growth and refractive error development. Animal studies have shown that ambient light exposure can influence the rate of eye growth and the development of experimental myopia (Ashby et al. 2009; Ashby & Schaeffel 2010; Cohen et al. 2011). In humans, studies of both children and adults show an association between greater objective daily ambient light exposure and slower eye growth which supports a role for light exposure in human eye growth (Read et al. 2015b; Ulaganathan et al. 2019). However, a recent cross-sectional study investigating personal light exposure and ocular diurnal variations in adults showed no significant association between ambient light exposure and ocular diurnal variations measured over one day (Burfield et al. 2019). Together, these findings suggest that ambient light exposure may be an important factor involved in regulating eye growth, however, the relationship between ocular diurnal variations, light exposure and refractive error and eye growth in the human eye has not been examined in detail. There is consistent evidence that ambient light exposure influences eye growth in both humans and animals. Evidence from animal models indicate that short-term daily changes in eye length and choroidal thickness may play a role in the regulation of longer-term eye growth, and that ambient light exposure can influence the ocular diurnal rhythms. In this study, we have examined the daily variations occurring in AL and subfoveal choroidal thickness (SFCT) of emmetropic and myopic young adults and explored the association between objectively measured light exposure and these diurnal variations. Since the daily patterns of light exposure differ between weekdays and weekends in children (Read et al. 2014), light exposure and ocular diurnal variations were assessed on both

5

weekdays and weekends. We also examined whether seasonal variations exist in diurnal ocular variations, and examined for the first time in the human eye, the association between short-term daily AL and choroidal variations and longer-term changes in AL and choroidal thickness. 2. Materials and methods: 2.1 Study participants: Thirty four young adults (21 females and 13 males) with an average age of 22.3 ± 4.1 years (range: 18 to 30 years) participated in this prospective longitudinal study examining diurnal variations and longer term (12-months) variations in AL and SFCT. Each participant had their diurnal variations in AL and SFCT assessed in winter (between May 2015 and September 2015) and then 6-months later in summer (between November 2015 and February 2016). An additional single measurement session (conducted in the morning between 9 am and 11 am) was carried out 6 months later (between May 2016 and August 2016) in order to determine the longer-term (annual) change in AL and SFCT. Subjective non-cycloplegic refraction was also measured at baseline and follow up visits, but the change in AL was used as the primary outcome of eye growth since changes in AL can be measured with higher precision than the changes in refractive error. The participants were recruited from the student population of the Queensland University of Technology. Before the commencement of the study, all subjects underwent an ophthalmic screening to rule out any history or evidence of ocular or systemic diseases, and/or ocular surgeries/injuries. Subjects with conditions that could disrupt their natural diurnal rhythm or habitual patterns of light exposure were excluded from the study (e.g. insomnia, night shift work). In this study, the subjective noncycloplegic spherical equivalent refraction (SER) was used to classify subjects as emmetropes (SER < 0.75 and > -0.75 D, mean SER +0.06 ± 0.34 D, range: +0.62 to -0.62 D, n = 18, mean age 21.9 ± 3.9 years) or myopes (SER ≤ -0.75 DS, mean SER -3.77 ± 2.18 D, range: -0.75 to -8.25 D, n = 16, mean age 22.8 ± 4.3 years). None of the subjects had anisometropia > 1.00 D or cylindrical refraction >1.25 D. All participants had visual acuity of 0.00 logMAR or better in each eye. Approval from the QUT human research ethics committee was obtained prior to data collection and written informed consent was obtained from all of the study participants. All participants were treated in accordance with the tenets of the Declaration of Helsinki. 2.2 Diurnal ocular variation measurements:

6

The daily ocular variations in AL and SFCT were assessed in winter and summer to study the potential influence of seasonal variations in light exposure on daily ocular variations. These measurements were acquired only during the academic semesters in an attempt to ensure subjects had similar near-workloads during the two seasons of measurement. Thirty-four and 29 subjects attended the winter and summer diurnal measurements respectively, and 31 subjects attended the final single measurement session at the end of 12-months (Table 1). Twenty-eight of the 34 subjects (16 emmetropes and 12 myopes) who attended the winter measurements had diurnal variations measured on both a weekday and a weekend (allowing an assessment of between day differences in diurnal variations). The remaining 6 subjects (4 myopes and 2 emmetropes) were unable to attend for measurements on the weekend and had their diurnal variations measured only on a weekday. For the summer season, 25 subjects (15 emmetropes and 10 subjects) attended both weekday and weekend measurement sessions, and the remaining 4 subjects (2 emmetropes and 2 myopes) had their measurements collected only on a weekday. One myopic subject was unable to attend one of their measurement sessions on the weekday in winter. Table 1: Overview of the mean ± SD measurement times and the number of emmetropic and myopic participants who attended each of the study sessions.

Phase 1 (Winter) Measurement session

Phase 2 (Summer)

Phase 3 (Winter)

Mean ± SD measurement time (hh:mm) sample size (emmetropes, myopes) Weekday

Weekend

Weekday

Weekend

Weekday

Session 1

09:28 ± 00:16 (18, 16)

09:21 ± 00:16 (16, 12)

09:10 ± 00:11 (17, 12)

09:17 ± 00:23 (15, 10)

09:59 ± 00:34 (17, 14)

Session 2

12:09 ± 00:10 (18, 16)

12:15 ± 00:15 (16, 12)

12:01 ± 00:21 (17, 12)

12:10 ± 00:22 (15, 10)

Session 3

15:07 ± 00:18 (18, 16)

15:11 ± 00:08 (16, 12)

15:02 ± 00:09 (17, 12)

15:10 ± 00:18 (15, 10)

Session 4

18:11 ± 00:13 (18, 15)

18:09 ± 00:13 (16, 12)

17:54 ± 00:11 (17, 12)

17:57 ± 00:23 (15, 10)

Session 5

21:00 ± 00:18 (18, 16)

20:56 ± 00:16 (16, 12)

20:40 ± 00:14 (17, 12)

20:41 ± 00:13 (15, 10)

To investigate the daily ocular variations in each subject, on each of the measurement days, ocular parameters (including AL and SFCT) were assessed at five measurement sessions, starting at ~9:00 am in the morning until ~9:00 pm at night with each measurement session separated by 2:30 to 3:00

7

hours. Since contact lens wear can affect corneal thickness and curvature (Liu & Pflugfelder 2000), contact lens wearing subjects (n = 2) were instructed to discontinue contact lens use on the day of measurements. All participants were instructed to avoid caffeine and alcohol from the evening before and on the measurement day to avoid potential confounding effects of these beverages on ocular measures (Altinkaynak et al. 2016; Kang et al. 2016). The first measurement in the morning was collected at least 1 – 2 hours after waking, to avoid the potential confounding effects of overnight corneal and anterior segment changes (Read et al. 2008) influencing AL measurements. Participants were asked to view a distant object (5 m distance) binocularly with their optimal distance correction for a period of 10 minutes before each measurement session, to avoid the influence of prior tasks (e.g. near-work) upon the measurements. Participants continued their regular daily activities between measurement sessions each day. 2.3 Ocular measurements: AL was measured on each subject’s right eye using the Lenstar LS 900 (Haag Streit AG, Koeniz, Switzerland) optical biometer, with 5 repeated measurements acquired at each session and then averaged for analyses. Choroidal thickness measurements from the right eye were obtained from cross-sectional chorioretinal images captured with the Heidelberg Spectralis Spectral Domain Optical Coherence Tomography (OCT) instrument (Heidelberg Engineering, Heidelberg, Germany). The instrument uses an 870 nm super luminescent diode for OCT imaging with an axial resolution of 3.9 µm and a transverse resolution of 14 µm. Three repeated chorioretinal images (each the average of 40 B-Scans) were obtained at each session using a single horizontal 30° line scan centred on the fovea with a minimum image quality index of 30 dB. The instrument’s enhanced depth imaging (EDI) mode was used to enhance choroidal visibility, and the follow-up feature was also used (with the first image at baseline as a reference) to ensure that the same retinal location was imaged throughout the study. 2.4 Objective light exposure measurements: Each participant’s daily objective ambient light exposure was also assessed in winter and summer by a light sensor (Actiwatch 2; Philips Respironics, Pittsburgh, PA, USA), worn on the nondominant wrist and configured to record instantaneous measures of personal ambient light exposure every 30 seconds, 24 hours a day, continuously for 14 days. Details regarding the sensors, betweendevice correlations, and the measurement protocol and data screening and analysis procedures for this data have been described in detail previously (Ulaganathan et al. 2017, 2019). Each

8

participant’s valid light exposure measurements were analysed to determine their average daily bright outdoor light exposure (daily minutes of exposure to light > 1000 lux) in winter and summer. On average, each participant had 13.5 ± 2.0 days of valid light exposure measures in winter and 13.3 ± 1.8 days of valid light exposure measures in summer out of the 14 day measurement periods in each season. In each season for all participants, the ocular diurnal variation measurements were conducted during the 14-day light exposure measurement period. 2.5 Data analysis: Following data collection, the cross-sectional chorioretinal images from the OCT instrument were exported for further analysis to obtain measures of choroidal thickness. Initially, the images were analysed using custom written software to automatically segment the outer boundary of the retinal pigment epithelium (RPE) and the chorioscleral interface (CSI) in each of the OCT images (AlonsoCaneiro et al. 2013). Following the automated segmentation, an experienced masked observer manually corrected any RPE and CSI segmentation errors. SFCT, defined as the axial distance from the RPE to CSI at the deepest point of the foveal pit, was then derived from all OCT images. Two participants’ choroidal images were excluded from the analysis due to poor image quality (poor visibility of CSI). All statistical analyses were performed using SPSS version 21 (IBM, Armonk, NY, USA). The normality of the data was confirmed by the Kolmogorov-Smirnov test (all p>0.05). Within-session repeatability was assessed for the AL and SFCT measurements by calculating the standard deviation, the coefficient of variation and intraclass correlation coefficient (ICC) from the 5 repeated AL and 3 repeated choroidal thickness measurements from each session. The daily variations in AL (measured from 9 am to 9 pm) and the influence of day of the week (weekday versus weekend), refractive group (emmetropes versus myopes) and season (winter versus summer) upon these variations were examined using linear mixed model (LMM) analysis. LMM analysis was chosen to examine the daily variations in this study since the LMM analysis can reliably accommodate missing data points which are often encountered in a longitudinal study (Krueger & Tian 2004). In order to avoid longitudinal changes in AL influencing the comparisons of diurnal variations between seasons, for each subject, the daily variations in AL were normalised to the mean of all AL measurements on each measurement day. Time of day, day of the week, and season were specified as repeated factors and a compound symmetry covariance structure was specified for the repeated effects. The change in AL was defined as the dependent variable and time of the day (expressed as a categorical factor), day of the week, season and refractive group were included as

9

fixed effects. Bonferroni adjusted pairwise comparisons were carried out for any significant main effects and interactions. Similar LMM analyses were used to examine the diurnal SFCT variations. To investigate the association between the diurnal changes in AL and SFCT, repeated measures analysis of covariance (ANCOVA) was performed (Bland & Altman 1995). The longitudinal changes in AL and SFCT over the 12-month study period were also examined using LMM analyses. A first-order autoregressive covariance structure was specified as the repeated covariance type. AL was defined as the dependent variable and refractive group was included as a fixed factor, and time from the baseline visit was included as a covariate in the model. Individual subject’s slopes and intercepts were included as random effects in the model, assuming a variance components covariance structure for the random effects. A similar approach was used for the analysis of the longitudinal SFCT changes. A separate LMM analysis was used to examine the relationship between the longitudinal changes in AL (dependent variable) and SFCT (covariate). Baseline SFCT was also included as a covariate in this model to analyse the association between baseline SFCT and the longitudinal AL changes. Partial correlation analysis was used to examine the influence of light exposure (daily time exposed to bright light (>1000 lux) averaged over both seasons) upon the daily variations in AL, controlling for the baseline AL. The difference between the daily peak and trough in AL (i.e. amplitude of variations) and the timing of the daily peak in AL (actual time of peak AL measurement which provides an estimate of the phase of the daily variations) for each subject averaged across all of the measurement days were used in this analysis. These analyses were also repeated to examine the SFCT measurements. Finally, the relationship between the daily variations and the longitudinal changes occurring in AL and SFCT over the 12-month study period was also examined using partial correlation analysis. 3. Results: The descriptive data of the average day length and climate conditions in the two seasons over which light exposure measures were collected has been summarized previously (Ulaganathan et al. 2019). Briefly, the days were significantly shorter and cooler in winter (p<0.05), and there were no significant differences in climate conditions experienced by emmetropes and myopes (p>0.05). 3.1 Within-session repeatability:

10

The mean within-session standard deviation for the AL and SFCT measures were 0.006 mm and 0.007 mm respectively, and the mean within-session range was 0.015 mm and 0.014 mm respectively, and the mean coefficient of variation was 0.025% and 1.79% respectively. The withinsession intraclass correlation coefficient was 0.999 for both the AL and SFCT measurements. These data indicate a high level of precision for both the AL and SFCT measurements used in the study. 3.2 Diurnal variation in axial length: LMM analysis revealed that AL underwent significant variations as a function of time of the day (p<0.001, n=34). On average, the daily peak in AL was observed at the 2nd measurement session (mean measurement time: 12:09 ± 0:02) and the trough at the final measurement session (20:51 ± 0:02), and post-hoc comparisons indicated that the AL at these two measurement sessions was significantly different (mean difference 0.015 mm, 95% CI: 0.011 to 0.019 mm, p<0.001). The overall (averaged from all measurement days in different seasons for all subjects) mean amplitude (peak to trough difference) of daily variations in AL was 0.029 ± 0.007 mm (range: 0.019 to 0.045 mm) (Table 2). There was no significant effect of day of the week (p=0.943, n=34), or a time of day by day of the week interaction (p=0.116), indicative of a similar pattern of daily variations in AL on weekdays and weekends in these young adults (Figure 1A). There was also no significant difference in the daily changes in AL observed between summer and winter (effect of season, p=0.964, n=34) and no significant interaction between time of day and season in AL (p=0.898) (Figure 1B). The mean amplitude of daily variations in AL was 0.031 ± 0.010 mm (range: 0.014 to 0.055 mm, n=34) in winter and 0.027 ± 0.010 mm (range: 0.011 to 0.054 mm, n=29) in summer (p=0.085).

11

Table 2: The mean ± SD baseline axial length (AL) and subfoveal choroidal thickness (SFCT), and the mean ± SD amplitude (mm) of daily AL and SFCT variations on weekdays and weekends in the emmetropes and myopes over winter and summer.

Seasons Refractive group

Baseline Winter

Day AL

SFCT

Summer

AL

SFCT

AL

SFCT

0.028 ± 0.011 (n=18)

0.018 ± 0.008 (n=18)

0.027 ± 0.013 (n=17)

0.017 ± 0.018 (n=16)

Weekend

0.028 ± 0.007 (n=16)

0.017 ± 0.006 (n=16)

0.025 ± 0.013 (n=15)

0.014 ± 0.006 (n=14)

Weekday

0.033 ± 0.010 (n=16)

0.016 ± 0.008 (n=15)

0.029 ± 0.011 (n=12)

0.018 ± 0.009 (n=12)

0.036 ± 0.018 (n=12)

0.016 ± 0.008 (n=11)

0.028 ± 0.007 (n=10)

0.015 ± 0.006 (n=10)

Weekday Emmetropes (mm)

23.79 ± 0.56

Myopes (mm)

24.94 ± 1.18 Weekend

0.379 ± 0.072

0.275 ± 0.094

The daily variations in AL exhibited a significant time of day by refractive group interaction (p=0.001), suggesting significant differences in the pattern of daily AL changes between emmetropes and myopes (Figure 1C). The overall mean amplitude of daily variations in AL was 0.027 ± 0.007 mm (range: 0.019 to 0.043 mm, n=18) in emmetropes and 0.032 ± 0.007 mm (range: 0.022 to 0.045 mm, n=16) in myopes (Table 2). Both emmetropes and myopes on average displayed a peak in AL at the second session and a trough at the final session of the day, but post-hoc comparisons revealed that the magnitude of difference between these two measurement sessions was significantly greater in the myopes (mean difference 0.020 mm; 95% CI: 0.014 to 0.026 mm) than the emmetropes (mean difference 0.010 mm; 95% CI: 0.005 to 0.015 mm) (p<0.001). These differences between refractive groups appear to have been driven by the myopes exhibiting a marginally greater increase (0.004 mm greater; 95% CI: 0.000 to 0.008 mm) in AL than the emmetropes (p=0.07) at the second measurement session, and a significantly greater decrease (0.006 mm shorter; 95% CI: 0.002 to 0.010 mm) in AL than the emmetropes during the final session of the day (p=0.001). There was no significant time of day by day of the week by refractive group (p=0.105), or time of day by season by refractive group interaction (p=0.262), suggesting that

12

the differences in the daily AL variations associated with refractive group were not different between weekdays and weekends or summer and winter. 3.3 Diurnal variations in subfoveal choroidal thickness: LMM analysis revealed that SFCT underwent significant variations as a function of time of the day (p<0.001, n=33). On average, the daily trough in SFCT was observed at the 3rd measurement session (mean measurement time: 15:07 ± 0:10) and the peak observed at the final measurement session (20:50 ± 0:10), and post-hoc comparisons indicated that the SFCT at these two measurement sessions was significantly different (mean difference 0.007 mm, 95% CI: 0.002 to 0.012 mm, p<0.001). The overall mean amplitude (peak to trough difference) of daily variations in SFCT was 0.012 ± 0.006 mm (range: 0.004 to 0.027 mm) (Table 2). There was no significant effect of day of the week (p=0.446, n=33), or time of day by day of the week interaction (p=0.151) or seasons (p=0.446, n=33), or time of day by season interaction (p=0.655) observed in the daily variations in SFCT, indicative of a similar pattern of daily variations in SFCT on weekdays and weekends and across the two seasons in these young adults (Figure 2A and 2B). The mean amplitude of daily variations in SFCT was 0.015 ± 0.006 mm (range: 0.003 to 0.029 mm, n=33) in winter and 0.013 ± 0.006 mm (range: 0.002 to 0.025 mm, n=28) in summer (p=0.958). The daily variations in SFCT were not significantly different between the refractive groups (p=0.45) (Figure 2C). The overall mean amplitude was 0.012 ± 0.006 mm (range: 0.004 to 0.027 mm, n=18) in emmetropes and 0.012 ± 0.006 mm (range: 0.005 to 0.026 mm, n=15) in myopes. Both emmetropes and myopes on average showed their trough in SFCT at the third session and a peak at the final session of the day. Repeated measures ANCOVA revealed a significant moderate negative association between the diurnal changes in AL and SFCT (slope = -0.338, p<0.001), and diurnal changes in SFCT could explain 36% of the variance in the AL changes (r = -0.603, p<0.001). 3.4 Light exposure and daily variations: The results of the analysis of daily bright light exposure (>1000 lux) has been presented in detail previously (Ulaganathan et al. 2019). Since both the daily AL and SFCT variations were not significantly different between seasons, we averaged the daily AL and SFCT variations across all measurement days to conduct further analyses. Partial correlation analysis indicated that there was a significant moderate negative association between the daily minutes of bright light exposure (averaged over both seasons) and the amplitude of daily AL variations (averaged across all measurement days) after adjusting for baseline AL (r = -0.511, p=0.002, n=34), with greater time

13

spent in bright light associated with a lower daily amplitude of change in AL (Figure 3). When these analyses were repeated for each of the refractive groups separately, a moderate negative association was found between daily minutes of bright light exposure and the amplitude of AL variations in both groups, but the association reached statistical significance only in the myopes (r = -0.465, p=0.06 in emmetropes, n=18; r = -0.577, p=0.024 in myopes, n=16). However, there was no significant association between daily bright light exposure and the amplitude of daily SFCT variations after adjusting for baseline SFCT (r = -0.07, p=0.713, n=33). We also examined the relationship between the timing of the daily peak in AL (and SFCT) and the daily bright light exposure, using a similar correlation analysis approach, and this did not reveal a significant association for either AL or SFCT (both p > 0.05), suggesting that the magnitude of daily light exposure did not significantly influence the timing/phase of the daily variations. 3.5 Daily variations, light exposure and longitudinal changes in axial length and SFCT: Examining the longitudinal changes in AL revealed significant increases over the 12-month period, with a mean change of 0.04 ± 0.07 mm (effect of time: p<0.001, LMM, n=34) (Figure 4A). Significantly larger magnitude changes were observed in myopes (β = 0.066 mm/year; 95% CI: 0.040 to 0.091, n=16) compared to emmetropes (β = 0.008 mm/year; 95% CI: -0.046 to 0.069, n=18) (p=0.005). SFCT did not change significantly over time (p=0.183, LMM, n=33). The mean change in SFCT over the 12-month period was -0.004 ± 0.021 mm (n=30). Myopes (β = -0.006 mm/year; 95% CI: -0.016 to 0.003, n=15) experienced slightly greater choroidal thinning over the 12-month period than emmetropes (β = -0.001 mm/year; 95% CI: -0.008 to 0.015, n=18), but the difference between the refractive groups was not statistically significant (p=0.433) (Figure 4A). Although on average the changes in choroidal thickness over 12 months were not statistically significant, a wide range of changes in choroidal thickness were observed in individual subjects (the observed changes ranged from -0.046 mm to 0.038 mm in individuals over the 12 months). Since measurable increases and decreases in choroidal thickness were observed in individual young adults over the course of the study, analyses were conducted to examine the potential association between the longer term changes in choroidal thickness and the 12-monthly changes in axial length. LMM analysis was used to examine the relationship between the longitudinal changes in AL and SFCT. A significant negative association between these two factors was observed (β = -0.25, 95% CI: -0.41 to -0.09, p=0.011, n=34) with the longitudinal changes in SFCT explaining 12.4% of the variance in the longitudinal changes in AL (r = -0.352, p=0.002). Figure 4B illustrates the inverse relationship between the longitudinal changes in AL and SFCT in each of the refractive groups, and

14

demonstrates that young adults with greater choroidal thickening exhibited less axial elongation over the study period. We also examined the influence of baseline SFCT on the longitudinal AL changes and found a significant inverse association between the two variables (β = -0.25, 95% CI: 0.41 to -0.09, p=0.002), suggesting that the young adults with thicker baseline SFCT exhibited significantly less axial elongation (Pearson correlation, r = -0.401, p=0.011) (Figure 4C). However, the association was not significant after adjusting for the refractive group (p=0.508), suggesting that the baseline choroidal thickness may not have an independent effect on the longitudinal AL changes. Partial correlation analysis revealed a significant moderate positive association between the amplitude of daily AL variations (averaged across all measurement days) and the longitudinal changes in AL over 12 months (r = 0.381, p=0.04, n=31). A greater daily amplitude of change in AL was associated with greater longitudinal changes in AL (after adjusting for baseline AL) (Figure 5A). However, the relationship between the daily amplitude of change and the longitudinal changes in AL was not statistically significant, after adjusting for both the baseline AL and the daily minutes of exposure to bright light (r = 0.340, p=0.07) (Figure 5B). The daily SFCT variations were not associated with the longitudinal changes in SFCT (r = 0.016, p=0.933, n=31) and this association remained unaltered after adjusting for both the baseline SFCT and the daily minutes of exposure to bright light (r = 0.016, p=0.934). There was also no significant association between the amplitude of daily SFCT variations and the longitudinal AL changes (r = 0.106, p=0.57, n=31). The phase of daily AL and SFCT variations (or the difference in phase between AL and SFCT) were also not significantly associated with the longitudinal changes in AL or SFCT (all p>0.05, n=31). 4. Discussion: This study provides evidence that there are differences in the daily variations of AL in young adults associated with refractive error, with myopes exhibiting a greater magnitude of change in AL during the day between 12 noon and 9 pm compared to emmetropes. The longitudinal nature of this study also allowed us to examine for the first time in humans, the relationship between the shortterm daily variations in AL and longer-term axial eye growth, which demonstrated that a larger amplitude of daily AL variations was significantly associated with more rapid longer-term axial eye growth, irrespective of the underlying baseline AL.

15

The overall mean amplitude of daily variations in AL observed is consistent with previous studies reporting on daily variations of AL in human eyes that have ranged from 0.015 to 0.046 mm (Stone et al. 2004; Wilson et al. 2006; Read et al. 2008; Chakraborty et al. 2011; Burfield et al. 2018). Previous animal studies have indicated that the amplitude of diurnal AL variation is greater in eyes that are growing more rapidly (i.e. in those eyes developing myopia) (Nickla et al. 1998b; Papastergiou et al. 1998; Nickla et al. 2002; Nickla 2006), which is consistent with our finding of greater amplitude of diurnal AL change associated with faster eye growth over a 12 month period. Previous work examining the effects of optical blur on diurnal AL changes in humans has also found an increase in the amplitude of daily variations following the introduction of monocular hyperopic defocus (an optical treatment known to stimulate axial eye growth in animal models) (Chakraborty et al. 2013). From this observational study, we cannot be certain that the increased amplitude of AL change associated with greater eye growth is a cause or consequence of myopia, since the myopes had established myopia at the beginning of the study. Nonetheless, this significant association (and the differences in diurnal AL variations observed between emmetropes and myopes) suggests a potentially important role for diurnal variation in the regulation of human eye growth. Previous research examining diurnal changes in AL have also included myopic and emmetropic subjects, but in contrast to our current findings these studies have not reported a significant difference between the diurnal variations of myopes and emmetropes (Stone et al. 2004; Chakraborty et al. 2011; Burfield et al. 2018). These previous studies have measured diurnal variations over only one or two days, whereas the diurnal variations in our study were derived from four measurement days across two different seasons. The older average age of participants examined in past studies may also have contributed towards the lack of significant difference observed between myopic and emmetropic participants (Stone et al. 2004; Chakraborty et al. 2011; Burfield et al. 2018). Furthermore, the longer-term changes in AL was not assessed prospectively in these studies, whereas in our current study axial elongation was assessed prospectively and confirmed significant progression (i.e. axial elongation) in our myopes (0.066 mm change in AL over 12 months of study), and allowed us to detect a significant association between long-term axial elongation and diurnal variations. Given that human studies have primarily assessed young adults, and since eye growth is known to occur more rapidly in children, additional longitudinal research exploring the relationship between diurnal variations of AL and longitudinal eye growth in children will help to further understand the association between diurnal rhythms and eye growth in humans.

16

There was a significant daily SFCT variation in these young adults with a mean amplitude of 0.012 ± 0.006 mm (range: 0.004 to 0.027 mm). Previous studies on humans have reported diurnal choroidal variations ranging from 0.018 to 0.060 mm (Brown et al. 2009; Chakraborty et al. 2011; Tan et al. 2012; Usui et al. 2012; Lee et al. 2014; Seidel et al. 2015; Kinoshita et al. 2016; Burfield et al. 2018). The lower amplitude found in our study could be due to differences in the duration of diurnal measurements (ranging from 12 hours to 24 hours in previous work). Our last measurement of choroidal thickness was at ~9 pm, which means that we might not have captured the true peak in SFCT which is known to occur around mid-night or early morning (Usui et al. 2012; Burfield et al. 2018). We also found a significant negative association (i.e. anti-phase relationship) between the pattern of daily rhythms in AL and SFCT, in agreement to previous human and animal studies (Weiss & Schaeffel 1993; Nickla et al. 1998a, 1998b; Papastergiou et al. 1998; Nickla et al. 2001, 2002; Stone et al. 2004; Wilson et al. 2006; Brown et al. 2009; Chakraborty et al. 2011, 2012, 2013; Nickla 2013; Lee et al. 2014; Kinoshita et al. 2016; Burfield et al. 2018). Given that the diurnal choroidal changes could not explain all of the variance in the diurnal AL changes in these young adults, it is possible that other factors such as scleral thickness changes may be involved in the diurnal rhythms of AL. Posterior scleral thickness is also known to be associated with refractive error, being thinner in myopic eyes (Rada et al. 2006). Nickla et al. (1999) have also observed diurnal variations in scleral proteoglycan synthesis pattern in chicks, analogous to AL, with a higher synthesis during the day than night, and hypothesized that rhythms in scleral proteoglycan synthesis may influence the rhythm in AL. It will therefore be of interest for future studies to investigate the presence of diurnal variations in posterior scleral thickness in humans to understand the potential role of scleral thickness changes in short- and longer-term AL and choroidal changes. Although diurnal SFCT variations were significantly associated with the AL variations, unlike the diurnal AL variations, the amplitude of diurnal SFCT variations were not significantly different between the refractive groups in this study. Previous studies (Chakraborty et al. 2011; Burfield et al. 2018) also did not find a significant difference in the amplitude of diurnal choroidal thickness variations between emmetropes and myopes. The relatively lesser magnitude of diurnal SFCT variations (compared with the AL variations) together with the small sample size may have contributed to the lack of significant difference between emmetropes and myopes in our study. We also assessed daily variation in winter and summer and did not find any statistically significant seasonal difference in the observed daily AL and SFCT changes (considering all subjects). There

17

was a small reduction in daily AL variations during summer (0.023 ± 0.010 mm) compared to winter (0.027 ± 0.010 mm), but the difference did not reach statistical significance. The modest seasonal differences in climate (~2 hours difference in day length), coupled with the relatively small sample size may have reduced our study’s ability to detect a significant seasonal difference in daily AL and choroidal variations. Our study provides evidence of an association between natural daily AL variations and objective ambient light exposure in humans, demonstrating that greater daily bright light exposure is associated with a lower amplitude of daily AL variations. However, a recent study on humans did not find a significant association between objective ambient light exposure (over 7 days) and natural diurnal AL rhythms in adults over a single day (Burfield et al. 2019). We captured ambient light exposure over two 14-day periods and measured diurnal variations over 4 days in different seasons and these differences in measurement protocols may have contributed to the variation in the findings between the two studies. Although we have reported an association between diurnal AL variations and ambient light exposure, the mechanism underlying this association remains unclear. Animal studies indicate that light stimulates the release of dopamine in the retina, and dopaminergic agonists have also been shown to block or reduce experimental myopia (Stone et al. 1989; Ashby & Schaeffel 2010; Feldkaemper & Schaeffel 2013; Lan et al. 2016). Dopamine is generally released only during the day when there is light, and melatonin (a molecule produced during the night) inhibits its production at night (Adachi et al. 1998). It has also been suggested that retinal dopamine release may induce choroidal thickening by stimulating the release of nitric oxide (Nickla & Wildsoet 2004; Sekaran et al. 2005; Nickla et al. 2009) from either the choroid or the retina. We hypothesise, based upon this evidence, that light induced dopamine release may limit the axial elongation during the day time and thereby reduce the amplitude of diurnal AL changes. This is the first study to report the longitudinal choroidal changes in young adults. Overall, the longitudinal choroidal change over the 12 months was not significantly different from the baseline choroidal thickness (mean of -0.004 mm; ranging from -0.046 mm to 0.038 mm). The betweensubject variation in SFCT changes was relatively large (0.021 mm), and greater than the withinsession measurement variability of 0.007 mm, suggesting that measurable changes in SFCT occur over time in young adults, but the direction of change was not consistent across the participants in this study. The longitudinal SFCT changes were inversely associated with the longitudinal AL changes suggesting that longitudinal choroidal thickening is associated with a lesser axial

18

elongation or slowing of eye growth in young adults. Previous animal studies have shown choroidal thickening and a reduction in AL in response to experimentally induced myopic defocus (Wallman et al. 1995; Hung et al. 2000; Troilo et al. 2000) suggesting that the changes in choroidal thickness may be directly associated with the changes in AL. Studies that investigated longitudinal eye growth in children have also demonstrated that subjects who experienced greater choroidal thickening exhibited lesser axial elongation (Read et al. 2015a; Fontaine et al. 2017). Taken together, the findings suggest that choroidal thickness may significantly contribute towards the regulation of eye growth in young adults. We also found that the amplitude of daily AL variations was positively associated with the longitudinal changes in AL, with a higher amplitude of daily change associated with greater axial elongation over 12 months (r = 0.381). However, the association was only modest and the strength of the association reduced further after adjusting for individual objective light exposure, which raises the possibility that changes in diurnal rhythms may not be independently associated with the longitudinal AL changes, but may be due to daily light exposure being associated with both longitudinal and diurnal changes in AL. Given the relatively small sample size in the current study, it appears that further research with larger numbers of emmetropes and myopes is required to confirm whether diurnal AL variations and longitudinal changes in AL are independently associated. A limitation of this study is that the daily ocular measurements were only captured during 12 hours of each 24-hour day. We used this 12-hour protocol with measurements every 3 hours in order to capture a large proportion of the diurnal variations without disturbing the routine daily activities of the participants. However, this approach does limit our study’s ability to reliably assess the phase of the 24 hour diurnal rhythm, and may slightly underestimate the amplitude of 24 hour change (since the trough in AL (Read et al. 2008; Burfield et al. 2018) and the peak in choroidal thickness (Usui et al. 2012; Burfield et al. 2018) has previously been documented in 24 hour studies to occur close to midnight). Further studies examining 24 hour AL and SFCT variations and longitudinal eye growth may therefore provide a more comprehensive understanding of the association between diurnal ocular rhythms and the longitudinal eye growth. Another limitation of the current study is that the light sensors measured only white light and the spectral composition of light exposure was not assessed. Several animal studies have shown that the spectral composition of ambient lighting can influence the course of refractive development (Rohrer et al. 1992; Seidemann & Schaeffel 2002; Rucker 2013; Smith et al. 2015), and exposure to

19

different wavelengths of light are also known to influence diurnal rhythms (Lee et al. 1997; Warman et al. 2003; Glickman et al. 2006). This suggests that the spectral content of light exposure may be involved in the association between light exposure and daily ocular variations documented in this study, and additional research examining the spectral composition of light exposure and diurnal ocular rhythms therefore appears warranted. Such work may help to better understand the potential role of intrinsically photosensitive retinal ganglion cells (ipRGCs), in the changes we have observed. This subset of retinal ganglion cells have been implicated in the control of circadian rhythms (Berson et al. 2002), and are known to exhibit their peak sensitivity to short wavelength light (446 – 480 nm) and are insensitive to long wavelengths (Brainard et al. 2001; Panda et al. 2005; Wang et al. 2011). However, in addition to their intrinsic photosignal, ipRGCs have also been shown to be strongly activated by rods and cones suggesting that long wavelengths indirectly stimulate (extrinsically signal) ipRGCs through cones (Dacey et al. 2005). 5. Conclusions: In summary, this study provides objective evidence demonstrating a significant association between light exposure, daily AL variations and longitudinal AL changes in human eyes. A significant association between the daily and longitudinal AL and SFCT variations was also found. Slower axial eye growth in subjects with thicker choroids and in subjects exhibiting longitudinal choroidal thickening was another important finding, which implies a potential role for the choroid in eye growth. Although we can only speculate upon the underlying mechanisms of these associations, these findings add new evidence to the existing literature supporting an important role of light exposure and the choroid in the regulation of human eye growth. While this observational study allowed us to establish an association between the AL and choroidal thickness changes, and the association between light exposure, and the short-term and longer-term changes in AL using detailed objective measures, the causative nature of these associations are not clear. Future prospective experimental interventional studies are required to further understand the associations between ambient light exposure, ocular diurnal variations and myopia development and progression.

Acknowledgements: The authors thank David Alonso Caneiro for his assistance with data analysis procedures.

20

6. References: Adachi A, Nogi T & Ebihara S (1998): Phase-relationship and mutual effects between circadian rhythms of ocular melatonin and dopamine in the pigeon. Brain Res. 792 (2): 361-369. Alonso-Caneiro D, Read S A & Collins M J (2013): Automatic segmentation of choroidal thickness in optical coherence tomography. Biomed. Opt. Express 4 (12): 2795-2812. doi: 10.1364/BOE.4.002795. Altinkaynak H, Ceylan E, Kartal B, Keles S, Ekinci M & Olcaysu O O (2016): Measurement of choroidal thickness following caffeine intake in healthy subjects. Curr. Eye Res. 41 (5): 708-714. doi: 10.3109/02713683.2015.1020168. Ashby R, Ohlendorf A & Schaeffel F (2009): The effect of ambient illuminance on the development of deprivation myopia in chicks. Invest. Ophthalmol. Vis. Sci. 50 (11): 5348-5354. doi: 10.1167/iovs.09-3419. Ashby R S & Schaeffel F (2010): The effect of bright light on lens compensation in chicks. Invest. Ophthalmol. Vis. Sci. 51 (10): 5247-5253. doi: 10.1167/iovs.09-4689. Berson D M, Dunn F A & Takao M (2002): Phototransduction by retinal ganglion cells that set the circadian clock. Science 295 (5557): 1070-1073. doi: 10.1126/science.1067262. Bland J M & Altman D G (1995): Calculating correlation coefficients with repeated observations: Part 1correlation within subjects. BMJ. 310 (6977): 446. Brainard G C, Hanifin J P, Greeson J M, Byrne B, Glickman G, Gerner E & Rollag M D (2001): Action spectrum for melatonin regulation in humans: evidence for a novel circadian photoreceptor. J. Neurosci. 21 (16): 6405-6412. Brown J S, Flitcroft D I, Ying G S, Francis E L, Schmid G F, Quinn G E & Stone R A (2009): In vivo human choroidal thickness measurements: evidence for diurnal fluctuations. Invest. Ophthalmol. Vis. Sci. 50 (1): 512. doi: 10.1167/iovs.08-1779. Burfield H J, Patel N B & Ostrin L A (2018): Ocular biometric diurnal rhythms in emmetropic and myopic adults. Invest. Ophthalmol. Vis. Sci. 59 (12): 5176-5187. doi:10.1167/iovs.18-25389. Burfield H J, Carkeet A & Ostrin L A (2019): Ocular and systemic diurnal rhythms in emmetropic and myopic adults. Invest. Ophthalmol. Vis. Sci. 60 (6): 2237-2247. doi:10.1167/iovs.19-26711. Chakraborty R, Read S A & Collins M J (2011): Diurnal variations in axial length, choroidal thickness, intraocular pressure, and ocular biometrics. Invest. Ophthalmol. Vis. Sci. 52 (8): 5121-5129. doi: 10.1167/iovs.11-7364. Chakraborty R, Read S A & Collins M J (2012): Monocular myopic defocus and daily changes in axial length and choroidal thickness of human eyes. Exp. Eye Res. 103: 47-54. doi: 10.1016/j.exer.2012.08.002. Chakraborty R, Read S A & Collins M J (2013): Hyperopic defocus and diurnal changes in human choroid and axial length. Optom. Vis. Sci. 90 (11): 1187-1198. doi: 10.1097/OPX.0000000000000035. Cohen Y, Belkin M, Yehezkel O, Solomon A S & Polat U (2011): Dependency between light intensity and refractive development under light-dark cycles. Exp. Eye Res. 92 (1): 40-46. doi: 10.1016/j.exer.2010.10.012. Dacey D M, Liao H W, Peterson B B, Robinson F R, Smith V C, Pokorny J, Yau K W & Gamlin P D (2005): Melanopsin-expressing ganglion cells in primate retina signal colour and irradiance and project to the LGN. Nature 433 (7027): 749-754. Feldkaemper M & Schaeffel F (2013): An updated view on the role of dopamine in myopia. Exp. Eye. Res. 114: 106-119. doi: 10.1016/j.exer.2013.02.007. Fontaine M, Gaucher D, Sauer A & Speeg-Schatz C (2017): Choroidal thickness and ametropia in children: A longitudinal study. Eur. J. Ophthalmol. 27 (6): 730-734. doi: 10.5301/ejo.5000965.

21

Glickman G, Byrne B, Pineda C, Hauck W W & Brainard G C (2006): Light therapy for seasonal affective disorder with blue narrow-band light-emitting diodes (LEDs). Biol. Psychiatry 59 (6): 502-507. doi: 10.1016/j.biopsych.2005.07.006. Hung L F, Wallman J & Smith E L, 3rd (2000): Vision-dependent changes in the choroidal thickness of macaque monkeys. Invest. Ophthalmol. Vis. Sci. 41 (6): 1259-1269. Kang H M, Woo Y J, Koh H J, Lee C S & Lee S C (2016): The effect of consumption of ethanol on subfoveal choroidal thickness in acute phase. Br. J. Ophthalmol. 100 (3): 383-388. doi: 10.1136/bjophthalmol-2015306969. Kinoshita T, Mitamura Y, Shinomiya K, et al. (2017): Diurnal variations in luminal and stromal areas of choroid in normal eyes. Br. J. Ophthalmol. 101 (3): 360-364. doi: 10.1136/bjophthalmol-2016-308594. Krueger C & Tian L (2004): A comparison of the general linear mixed model and repeated measures ANOVA using a dataset with multiple missing data points. Biol. Res. Nurs. 6 (2): 151-7. Lan W, Yang Z, Feldkaemper M & Schaeffel F (2016): Changes in dopamine and ZENK during suppression of myopia in chicks by intense illuminance. Exp. Eye Res. 145: 118-124. doi: 10.1016/j.exer.2015.11.018. Lee S W, Yu S Y, Seo K H, Kim E S & Kwak H W (2014): Diurnal variation in choroidal thickness in relation to sex, axial length, and baseline choroidal thickness in healthy Korean subjects. Retina. 34 (2): 385-393. doi: 10.1097/IAE.0b013e3182993f29. Lee T M, Chan C C, Paterson J G, Janzen H L & Blashko C A (1997): Spectral properties of phototherapy for seasonal affective disorder: a meta-analysis. Acta Psychiatr. Scand. 96 (2): 117-121. LeGates T A, Fernandez D C & Hattar S (2014): Light as a central modulator of circadian rhythms, sleep and affect. Nat. Rev. Neurosci. 15 (7): 443-454. doi: 10.1038/nrn3743. Liu Z & Pflugfelder S C (2000): The effects of long-term contact lens wear on corneal thickness, curvature, and surface regularity. Ophthalmology 107 (1): 105-111. Nickla D L (2006): The phase relationships between the diurnal rhythms in axial length and choroidal thickness and the association with ocular growth rate in chicks. J. Comp. Physiol. A. 192 (4): 399-407. doi: 10.1007/s00359-005-0077-2. Nickla D L (2013): Ocular diurnal rhythms and eye growth regulation: where we are 50 years after Lauber. Exp. Eye Res. 114: 25-34. doi: 10.1016/j.exer.2012.12.013. Nickla D L, Damyanova P & Lytle G (2009): Inhibiting the neuronal isoform of nitric oxide synthase has similar effects on the compensatory choroidal and axial responses to myopic defocus in chicks as does the non-specific inhibitor L-NAME. Exp. Eye Res. 88 (6): 1092-1099. doi: 10.1016/j.exer.2009.01.012. Nickla D L, Rada J A & Wallman J (1999): Isolated chick sclera shows a circadian rhythm in proteoglycan synthesis perhaps associated with the rhythm in ocular elongation. J. Comp. Physiol. A. 185 (1): 81-90. Nickla D L, Wildsoet C & Wallman J (1998a): The circadian rhythm in intraocular pressure and its relation to diurnal ocular growth changes in chicks. Exp. Eye Res. 66 (2): 183-193. doi: 10.1006/exer.1997.0425. Nickla D L, Wildsoet C & Wallman J (1998b): Visual influences on diurnal rhythms in ocular length and choroidal thickness in chick eyes. Exp. Eye Res. 66 (2): 163-181. doi: 10.1006/exer.1997.0420. Nickla D L & Wildsoet C F (2004): The effect of the nonspecific nitric oxide synthase inhibitor NG-nitro-Larginine methyl ester on the choroidal compensatory response to myopic defocus in chickens. Optom. Vis. Sci. 81 (2): 111-118. Nickla D L, Wildsoet C F & Troilo D (2001): Endogenous rhythms in axial length and choroidal thickness in chicks: implications for ocular growth regulation. Invest. Ophthalmol. Vis. Sci. 42 (3): 584-588. Nickla D L, Wildsoet C F & Troilo D (2002): Diurnal rhythms in intraocular pressure, axial length, and choroidal thickness in a primate model of eye growth, the common marmoset. Invest. Ophthalmol. Vis. Sci. 43 (8): 2519-2528.

22

Panda S, Nayak S K, Campo B, Walker J R, Hogenesch J B & Jegla T (2005): Illumination of the melanopsin signaling pathway. Science. 307 (5709): 600-604. doi: 10.1126/science.1105121. Papastergiou G I, Schmid G F, Riva C E, Mendel M J, Stone R A & Laties A M (1998): Ocular axial length and choroidal thickness in newly hatched chicks and one-year-old chickens fluctuate in a diurnal pattern that is influenced by visual experience and intraocular pressure changes. Exp. Eye Res. 66 (2): 195-205. doi: 10.1006/exer.1997.0421. Rada J A, Shelton S & Norton T T (2006): The sclera and myopia. Exp. Eye Res. 82 (2): 185-200. doi: 10.1016/j.exer.2005.08.009. Read S A, Collins M J & Vincent S J (2014). Light exposure and physical activity in myopic and emmetropic children. Optom. Vis. Sci. 91 (3): 330-341. doi: 10.1097/OPX.0000000000000160. Read S A, Alonso-Caneiro D, Vincent S J & Collins M J (2015a): Longitudinal changes in choroidal thickness and eye growth in childhood. Invest. Ophthalmol. Vis. Sci. 56 (5): 3103-3112. doi: 10.1167/iovs.15-16446. Read S A, Collins M J & Iskander D R (2008): Diurnal variation of axial length, intraocular pressure, and anterior eye biometrics. Invest. Ophthalmol. Vis. Sci. 49 (7): 2911-2918. doi: 10.1167/iovs.08-1833. Read S A, Collins M J & Vincent S J (2015b): Light exposure and eye growth in childhood. Invest. Ophthalmol. Vis. Sci. 56 (11): 6779-6787. doi: 10.1167/iovs.14-15978. Rohrer B, Schaeffel F & Zrenner E (1992): Longitudinal chromatic aberration and emmetropization: results from the chicken eye. J. Physiol. 449: 363-376. Rucker F J (2013): The role of luminance and chromatic cues in emmetropisation. Ophthalmic. Physiol. Opt. 33 (3): 196-214. doi: 10.1111/opo.12050. Seidel G, Hausberger S, Herzog S A, Palkovits S, Poschl E M, Wackernagel W & Weger M (2015): Circadian macular volume changes in the healthy human choroid. Am. J. Ophthalmol. 159 (2): 365-371 e362. doi: 10.1016/j.ajo.2014.11.002. Seidemann A & Schaeffel F (2002): Effects of longitudinal chromatic aberration on accommodation and emmetropization. Vision Res. 42 (21): 2409-2417. Sekaran S, Cunningham J, Neal M J, Hartell N A & Djamgoz M B (2005): Nitric oxide release is induced by dopamine during illumination of the carp retina: serial neurochemical control of light adaptation. Eur. J. Neurosci. 21 (8): 2199-2208. doi: 10.1111/j.1460-9568.2005.04051.x. Smith E L, 3rd, Hung L F, Arumugam B, Holden B A, Neitz M & Neitz J (2015): Effects of long-wavelength lighting on refractive development in infant rhesus monkeys. Invest. Ophthalmol. Vis. Sci. 56 (11): 64906500. doi: 10.1167/iovs.15-17025. Stone R A, Lin T, Laties A M & Iuvone P M (1989): Retinal dopamine and form-deprivation myopia. Proc. Natl. Acad. Sci. U S A. 86 (2): 704-706. Stone R A, Quinn G E, Francis E L, et al. (2004): Diurnal axial length fluctuations in human eyes. Invest. Ophthalmol. Vis. Sci. 45 (1): 63-70. Tan C S, Ouyang Y, Ruiz H & Sadda S R (2012): Diurnal variation of choroidal thickness in normal, healthy subjects measured by spectral domain optical coherence tomography. Invest. Ophthalmol. Vis. Sci. 53 (1): 261-266. doi: 10.1167/iovs.11-8782. Troilo D, Nickla D L & Wildsoet C F (2000): Choroidal thickness changes during altered eye growth and refractive state in a primate. Invest. Ophthalmol. Vis. Sci. 41 (6): 1249-1258. Ulaganathan S, Read S A, Collins M J & Vincent S J (2017): Measurement duration and frequency impact objective light exposure measures. Optom. Vis. Sci. 94 (5): 588-597. doi: 10.1097/OPX.0000000000001041. Ulaganathan S, Read S A, Collins M J & Vincent S J (2019): Influence of seasons upon personal light exposure and longitudinal axial length changes in young adults. Acta. Ophthalmol. 97 (2): e256-e265. doi: 10.1111/aos.13904.

23

Usui S, Ikuno Y, Akiba M, Maruko I, Sekiryu T, Nishida K & Iida T (2012): Circadian changes in subfoveal choroidal thickness and the relationship with circulatory factors in healthy subjects. Invest. Ophthalmol. Vis. Sci. 53 (4): 2300-2307. doi: 10.1167/iovs.11-8383. Wallman J, Wildsoet C, Xu A, Gottlieb M D, Nickla D L, Marran L, Krebs W & Christensen A M (1995): Moving the retina: choroidal modulation of refractive state. Vision Res. 35 (1): 37-50. Wang F, Zhou J, Lu Y & Chu R (2011): Effects of 530 nm green light on refractive status, melatonin, MT1 receptor, and melanopsin in the guinea pig. Curr. Eye Res. 36 (2): 103-111. doi: 10.3109/02713683.2010.526750. Warman V L, Dijk D J, Warman G R, Arendt J & Skene D J (2003): Phase advancing human circadian rhythms with short wavelength light. Neurosci. Lett. 342 (1-2): 37-40. Weiss S & Schaeffel F (1993): Diurnal growth rhythms in the chicken eye: relation to myopia development and retinal dopamine levels. J. Comp. Physiol. A. 172 (3): 263-270. Wilson L B, Quinn G E, Ying G S, Francis E L, Schmid G, Lam A, Orlow J & Stone R A (2006): The relation of axial length and intraocular pressure fluctuations in human eyes. Invest. Ophthalmol. Vis. Sci. 47 (5): 1778-1784. doi: 10.1167/iovs.05-0869.

Contributors: Design of the research: All authors; Data collection: SU; Analysis and interpretation of data: All authors; Drafting the manuscript: All authors; All authors reviewed and approved the final manuscript. Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

24

Figure Captions: Figure 1: The mean daily variations in axial length (AL) (9 am to 9 pm) during weekdays (n=34) (dark red line) and weekends (n=28) (purple line) (averaged across the two seasons) for all subjects (A), during winter (n=34) (orange line) and summer (n=29) (turquoise line) (averaged across weekdays and weekends for all subjects) (B) and in emmetropes (n=18) (blue line) and myopes (n=16) (red line) (averaged across weekdays and weekends and summer and winter) (C). To highlight the daily variations in AL, all values are expressed as the difference from the mean of all sessions each day (i.e. all values are normalised to the mean). Vertical error bars represent the standard error of the mean. Horizontal error bars indicate standard error in the mean measurement time at each session (in minutes). * indicates significant difference in the change in AL between the refractive groups (p<0.05). Figure 2: The mean daily variations in subfoveal choroidal thickness (SFCT) (9 am to 9 pm) during weekdays (n=33) (dark red line) and weekends (n=27) (purple line) (averaged across the two seasons) for all subjects (top), during winter (n=33) (orange line) and summer (n=28) (turquoise line) (averaged across weekdays and weekends for all subjects) (middle) and in emmetropes (n=18) (blue line) and myopes (n=15) (red line) (averaged across weekdays and weekends and summer and winter) (bottom). To highlight the daily variations in SFCT, all values are expressed as the difference from the mean of all sessions each day (i.e. all values are normalised to the mean). Vertical error bars represent the standard error of the mean. Horizontal error bars indicate standard error in the mean measurement time at each session (in minutes). Figure 3: Association between the amplitude of daily variations in axial length (AL) (averaged across all days and seasons) and daily time exposed to light levels >1000 lux (averaged across all days and seasons) for all subjects (n=34). r value indicates the correlation coefficient after adjusting for baseline axial length. Solid line indicates the best fit regression line. p value indicates significance value. Figure 4: (A) Mean change in axial length (AL) (solid lines) and subfoveal choroidal thickness (SFCT) (dashed lines) over the 12 months of the study for emmetropes (n=17) (green) and myopes (n=14) (red). Vertical error bars indicate the standard error of the mean change. Horizontal error bars indicate the standard error of the study visit time. (B) Relationship between 12-monthly changes in axial length and 12-monthly changes in subfoveal choroidal thickness for emmetropes (n=17) (green) and myopes (n=14) (red). (C) Relationship between 12-monthly changes in axial length and baseline subfoveal choroidal thickness for all subjects (n=31). r values indicate the correlation coefficient. Dashed lines indicate the best fit regression line. p value indicates significance value. Figure 5: Relationship between 12-monthly changes in axial length (AL) and amplitude of daily axial length variations (averaged across all days and seasons) for all subjects, after adjusting for baseline axial length (n=31) (top), and after adjusting for baseline axial length and objectively measured daily time exposed to light levels >1000 lux (averaged across all days and seasons) (bottom). r values indicate the correlation coefficient. Solid lines indicate the best fit regression line for all subjects. p value indicates significance value.

0.015

A

Weekday Weekend

B

Winter Summer

C

Emmetropes

Change in AL (mm)

0.010 0.005 0.000 -0.005 -0.010 -0.015 0.015

Change in AL (mm)

0.010 0.005 0.000 -0.005 -0.010 -0.015 0.015

Myopes

Change in AL (mm)

0.010 0.005 0.000

*

-0.005 -0.010 -0.015 9

12

15 Time of day

18

21

1

Change in SFCT (mm)

0.010

A

Weekday Weekend

B

Winter Summer

0.005

0.000

-0.005

Change in SFCT (mm)

-0.010 0.010

0.005

0.000

-0.005

-0.010 0.010

Emmetropes

C Change in SFCT (mm)

Myopes 0.005

0.000

-0.005

-0.010 9

12

15 Time of day

18

21

1

Highlights: •

Greater bright light exposure is linked with lower amplitude diurnal variations in eye length

•

Rapid axial eye growth is associated with larger magnitude daily variations in eye length

•

Both diurnal and long-term changes in eye length are linked with choroidal thickness changes