Article Text

Download PDFPDF

Association of time outdoors and patterns of light exposure with myopia in children
  1. Mijie Li1,2,
  2. Carla Lanca2,
  3. Chuen-Seng Tan1,
  4. Li-Lian Foo3,4,
  5. Chen-Hsin Sun5,
  6. Fabian Yap6,
  7. Raymond P Najjar4,7,
  8. Charumathi Sabanayagam3,4,
  9. Seang-Mei Saw1,2,4
  1. 1 Saw Swee Hock School of Public Health, National University of Singapore, Singapore
  2. 2 Myopia Research Group, Singapore Eye Research Institute, Singapore
  3. 3 Singapore Eye Research Institute, Singapore National Eye Centre, Singapore
  4. 4 Ophthalmology and Visual Sciences Academic Clinical Programme, Duke-NUS Medical School, Singapore
  5. 5 Yong Loo Lin School of Medicine, National University of Singapore, Singapore
  6. 6 Department of Maternal Fetal Medicine, KK Women’s and Children’s Hospital, Singapore
  7. 7 Visual Neuroscience Research Group, Singapore Eye Research Institute, Singapore
  1. Correspondence to Dr Seang-Mei Saw, Singapore Eye Research Institute, The Academia, 20 College Road, Discovery Tower Level 6, Singapore 169856, Singapore; ephssm{at}nus.edu.sg

Abstract

Background/aims To evaluate the association of reported time outdoors and light exposure patterns with myopia among children aged 9 years from the Growing Up in Singapore Towards Healthy Outcomes birth cohort.

Methods We assessed reported time outdoors (min/day), light exposure patterns and outdoor activities of children aged 9 years (n=483) with a questionnaire, the FitSight watch and a 7-day activity diary. Light levels, the duration, timing and frequency of light exposure were assessed. Cycloplegic spherical equivalent (SE), myopia (SE≤−0.5 D) and axial length (AL) of paired eyes were analysed using generalised estimating equations.

Results In this study, 483 (966 eyes) multiethnic children (50.0% boys, 59.8% Chinese, 42.2% myopic) were included. Reported time outdoors (mean±SD) was 100±93 min/day, and average light levels were 458±228 lux. Of the total duration children spent at light levels of ≥1000 lux (37±19 min/day), 76% were spent below 5000 lux. Peak light exposure occurred at mid-day. Children had 1.7±1.0 light exposure episodes/day. Common outdoor activities were walks, neighbourhood play and swimming. Greater reported time outdoors was associated with lower odds of myopia (OR=0.82, 95% CI 0.70 to 0.95/hour increase daily; p=0.009). Light levels, timing and frequency of light exposures were not associated with myopia, SE or AL (p>0.05).

Conclusion Reported time outdoors, light levels and number of light exposure episodes were low among Singaporean children aged 9 years. Reported time outdoors was protective against myopia but not light levels or specific light measures. A multipronged approach to increase time outdoors is recommended in the combat against the myopia epidemic.

  • epidemiology
  • public health
  • child health (paediatrics)
  • vision

Data availability statement

The data that support the findings of this study are generated from the Growing Up in Singapore Towards healthy Outcomes (GUSTO) study, but restrictions apply to the public availability of these data, due to ethical restriction (patient confidentiality) imposed by the Singhealth Centralised Institutional Review Board. Datasets are, however, available from the authors upon reasonable request and with the permission of the GUSTO study.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Myopia is a major public health problem affecting over 80% of young adults in East Asia and Southeast Asia,1 2 linked to sight-threatening pathological complications.3 Effective early interventions are crucial, with randomised controlled trials4 5 and school-based interventional trials6 7 suggesting less time outdoors as a key risk factor for myopia. However, time outdoors in outdoor trials was largely quantified using questionnaires, with scarce information on light patterns.

Animal studies suggest bright light stimulate dopamine release, thereby inhibiting ocular growth.8 Several hours of bright light exposure (≥15 000 lux) retarded form-deprivation myopia in chicks and rhesus monkeys.9 10 Besides light levels, the timing11 and frequency12 of light exposure may also regulate ocular growth. In chicks, the inhibitory effects of bright light varied across the time of the day11 and the dosage (episodic vs cumulative duration) of light exposure.12

In children of similar ages living in different geographical locations with high (58.9%, Singapore) or low (14.4%, Australia) myopia prevalence,13 14 the duration of light exposure was shown to be different. A recent study demonstrated a higher daily duration of outdoor light exposure in 43 Australian children, compared with 69 Singaporean children (105±42 vs 61±40 min/day, p=0.005).15 Furthermore, average light levels were lower among 41 myopes than 41 emmetropes (915 vs 1271 lux, p<0.01), aged 10–15 years in Australia.16 However, given few studies of small sample sizes, outdoor light exposure patterns (light levels, duration, timing and frequency of light exposure), and their effects on myopia in children are not fully understood.

We aimed to examine the association of reported time outdoors and light exposure patterns with myopia, among children aged 9 years from the Growing Up in Singapore Towards Healthy Outcomes (GUSTO) study.

Materials and methods

Study population

A cross-sectional study of the 9-year visit was nested in the GUSTO birth cohort.17 Pregnant mothers were recruited from two major public maternity hospitals in Singapore. Of 1176 children at baseline, 716 (61%) children returned at the 9-year visit. On excluding 7 children on myopia treatment (orthokeratology (n=3) and atropine (n=4)), 709 children were eligible and included. Parents and children provided written consent and assent.

Ocular examination

Cycloplegic autorefraction (Canon RK-5/RK-F2; Canon, Japan) and an optical biometer (IOL Master 500; Carl Zeiss-Meditec, Germany) were used to assess cycloplegic spherical equivalent (SE) and axial length (AL), respectively. Cycloplegia was successfully induced in 627 children (88.4%), using three drops of 1% cyclopentolate hydrochloride, instilled 5 min apart. Autorefraction was performed at least 30 min after the first drop, with pupil dilation of ≥6 mm. Myopia was defined as SE of ≤−0.5 D.

Time outdoors

Parents quantified the average duration their child spent on time outdoors per day (hours/day), in the past month, using a similar questionnaire to the Sydney Myopia Study.18 Reported time outdoors included outdoor physical or leisure activities, on weekdays and weekends (presented in min/day) (1 hour/day=60 min/day).

Light exposure patterns

The FitSight wrist-worn watch (patent WO2015152818A1)19 recorded light levels at 1 min intervals, over 14 days during daylight hours (7:00–19:00). Among 627 children with cycloplegia, light data were available from 576 children. Of these, 93 children (16.1%) were excluded due to poor compliance with FitSight: (1) missing at least 1 weekday and 1 weekend of watch wear (n=72); (2) wear days with average daily light levels of ≤100 lux or proportion of 0 lux entries of ≥60% per day (n=21), which was considered implausibly low and highly suggested the watch sensor had been covered over an extended period. Thus, 483 children were included in the analyses.

Outdoor environments were defined using a light level cut-off of ≥1000 lux, based on prior studies.5 16 19 20 Metrics for evaluating light levels included (1) average light levels while outdoors only (ie, average outdoor light level), and (2) average light level (both indoors and outdoors). We assessed the average duration of light exposure (min/day) (at ≥1000 lux) and at increasing light-level cut-offs (≥3000, ≥5000 and ≥15 000 lux). Variations in light levels and duration of light exposure were assessed in hourly intervals and across three equal 4-hour periods during daylight hours (morning: 7:00–11:00, mid-day: 11:00–15:00 and evening: 15:00–19:00). The average number of light exposure episodes (episodes/day, defined as light exposure at ≥1000 lux over a continuous period of ≥5 min15) and the average duration of light exposure episodes (min/episode) (a measure of continuous duration of light exposure) were computed.

Outdoor activity patterns

Children recorded the duration of each outdoor activity, including the activity type (eg, running or swimming) in an activity diary over 7 days (7:00–19:00). ‘Outdoor’ activities were defined as activities performed outside a building (eg, open courts or playground in parks) and were categorised into physical or non-physical activities (eg, non-sports or meals). The number of children spending the longest duration in each activity type was ranked to identify the most common outdoor activities, with ties broken at random. Of 483 children with valid light data and cycloplegia, activity diary records from 436 children who reported at least one outdoor activity with non-zero duration were analysed.

Anthropometric and questionnaire measurements

Primary caregivers reported each child’s gender, ethnicity (Chinese and non-Chinese, comprising Malays, Indians and others), near-work activities (hours/day), the number of myopic parents (none or at least one parent), maternal education levels (secondary school and lower or higher) and maternal employment status (non-working or working). Each child’s standing height (cm) was measured (Seca 213; Seca, Germany).

Statistical analysis

Daily reported time outdoors, daily duration of light exposure measures, daily duration spent on various activity types (ie, mean duration in min/day, respectively) and daily light levels (lux) across all days of the week were computed as follows:

Embedded Image

Embedded Image

Embedded Image

Light exposure measures across periods of the day were compared using the linear regression model with generalised estimating equations (GEE),21 accounting for correlations between periods. Paired eyes were analysed to assess associations between exposures with each outcome, using multivariable logistic (myopia) or linear (SE or AL) regression models with GEE, accounting for confounders and the correlation between paired eyes. We further examined associations between light exposure measures stratified into weekday and weekend, respectively (ie, mean daily minutes on weekdays or weekend, min/day), with each outcome. Final multivariable models were determined with backward manual stepwise selection. A two-sided p value of <0.05 was considered statistically significant. Estimated measures of association and 95% CIs were presented. Statistical analyses were performed using STATA V.13.

Results

Included (n=483) and excluded (n=233) children did not differ in the proportion of myopes (42.2% vs 43.1%, p=0.83), mean SE (±SD) (−0.61±1.83 vs −0.59±1.56 D, p=0.87) or mean AL (23.3±1.0 vs 23.3±0.9 mm, p=0.84), proportion of boys (50.0 vs 56.1%, p=0.13), reported time outdoors (100±93 min/day vs 98±86 min/day, p=0.76) or near-work (5.5±3.0 vs 5.2±2.8 hours/day, p=0.21). Among children included, a higher proportion of children were Chinese (59.8 vs 50.0%, p=0.013) or had myopic parent(s) (79.2 vs 67.1%, p=0.001).

Across all days, children reported time outdoors of 100±93 min/day (1.7±1.6 hours/day), average outdoor light level of 4609±1736 lux and average light level of 458±228 lux (table 1). While outdoors, children spent a large proportion (76%, 28 min/day) of time in moderate light levels of <5000 lux. Of the total duration spent <5000 lux, 82% (23 min/day) were under light levels between 1000 (inclusive) and 3000 (exclusive) lux. On weekends (vs weekdays), reported time outdoors (p<0.001) and the duration of light exposure (p=0.0038) were significantly higher. Boys had significantly higher light levels and duration of light exposure than girls (p<0.001). With regard to ethnicity and parental myopia, outdoor light levels were significantly higher in non-Chinese (versus Chinese) children (4976±2130 lux vs 4379±1373 lux, p<0.001) and children without myopic parents (versus children with myopic parents) (4957±2229 lux vs 4514±1597 lux, p=0.029), but reported time outdoors and other light exposure measures (average light levels, duration of light exposure at various light level cut-offs) did not differ significantly (p>0.05). Reported time outdoors was significantly higher in children with non-working (versus working) mothers (114±113 min/day vs 92±81 min/day, p=0.022) but did not differ by maternal education levels (p>0.05). Additionally, light exposure measures did not differ significantly by maternal education level or maternal work status (p>0.05).

Table 1

Comparisons of reported time outdoors and light exposure measures across time of the week and gender groups among children aged 9 years from the Growing Up in Singapore Towards Healthy Outcomes study (N=483)

Across all days, the mean light exposure measures differed significantly across the three time periods during daylight hours (p<0.001) (table 2). Higher average light levels and duration of light exposure occurred at mid-day, compared with the morning or evening. During daylight hours, average light levels and duration of light exposure did not exceed 693±594 lux/hour or 5.6±3.6 min/hour, respectively.

Table 2

Comparisons of light exposure measures across periods of daylight hours among children aged 9 years from the Growing Up in Singapore Towards Healthy Outcomes study (N=483)

Children exhibited 1.7±1.0 light exposure episodes/day, lasting 6.3±4.5 min/episode (table 3). The number of light exposure episodes was higher on weekends than on weekdays and in boys than in girls. The average duration of light exposure episodes was significantly higher in boys than in girls. The number and duration of light exposure episodes did not differ significantly by ethnicity, parental myopia, maternal education levels or maternal work status (p>0.05).

Table 3

Comparisons of daily number (episode/day) and duration of light exposure episodes (min/episode) among children aged 9 years from the Growing Up in Singapore Towards Healthy Outcomes study (N=483)

The three most common types of specific outdoor physical activities children spent the longest duration on were walks, neighbourhood play and swimming (table 4). Transportation was a common activity but a mixture of indoor and outdoor exposures. Consistently, walks and neighbourhood play were the most common specific outdoor activities in children, regardless of gender, ethnicity, the number of myopic parents, maternal educational levels or maternal work status. While outdoors, 56.4% of children spent a higher average daily duration in physical activities, compared with non-physical activities.

Table 4

Distribution of types of common outdoor activities (based on the longest duration spent) among children aged 9 years from the Growing Up in Singapore Towards Healthy Outcomes study (N=436)

In multivariable analyses, reported time outdoors was significantly associated with lower odds of myopia (p=0.009), but not with SE (p=0.051) and AL (p=0.17) (table 5). Average light levels and duration of light exposure were not associated with myopia, SE or AL (p>0.05). Further analyses revealed no associations between duration of light exposure at higher light-level cut-offs (≥3000, ≥5000 and ≥15 000 lux), the timing of light levels or duration of light exposure at different periods during daylight hours (morning, mid-day or evening) or the number and duration of daily light exposure episodes, with myopia, SE or AL (p>0.05). When exposures were stratified by the weekdays and weekend, average outdoor light levels (on the weekdays) were associated with lower odds of myopia (OR=0.88 per 1000 lux increase, 95% CI 0.79 to 0.98; p=0.016), but not with SE or AL (p>0.05). Additionally, the duration of light exposure episodes (on the weekdays) was associated with shorter AL (β: −0.78/hour increase daily, 95% CI −1.51 to –0.06; p=0.035). Light levels or duration of light exposure episodes on the weekend were not associated with myopia, SE or AL (p>0.05). The duration, timing or frequency of light exposure on either the weekday or weekend, respectively, was not associated with myopia, SE or AL (p>0.05).

Table 5

Association between reported time outdoors, light exposure measures with myopia, SE and AL among children aged 9 years from the Growing Up in Singapore Towards Healthy Outcomes study (N=483, 966 eyes)

Discussion

Reported time outdoors, average light levels and light exposure episodes were low among Singaporean children aged 9 years. While outdoors, Singaporean children spent most of their time under moderate light levels of <5000 lux. Reported time outdoors was higher on weekends, compared with weekdays. Average light levels were higher in boys (than in girls) and highest during mid-day (than during morning or evening). The most common outdoor activities were outdoor walks, neighbourhood play and swimming. Reported time outdoors, rather than specific light exposure patterns, was protective against myopia.

Singaporean children had shorter reported time outdoors and lower light levels, relative to their counterparts overseas. In the current study, reported time outdoors was 1.7±1.6 hours/day and average light levels were 458±228 lux. In contrast, children of similar ages in Australia and the UK reported time outdoors of at least 2–3 hours/day.18 22 Likewise, higher average light levels of 1072 and 1627 lux were reported in Australian (age 10–15 years)16 and American (age 5–10 years) children,20 respectively. Differences in educational demands (homework load and after-school enrichment classes),23 lifestyle and sociocultural factors may account for the lower reported time outdoors and light levels in Singaporean children.

Singaporean children spent a longer duration outdoors on weekends, compared with weekdays (2.7±2.2 hours/day vs 1.3±1.6 hours/day). While children aged 9–11 years from 12 low-income to high-income countries also had higher reported time outdoors on weekends compared with weekdays, a smaller difference by the time of the week (2.7 hours/day vs 2.5 hours/day) was observed.24 Average outdoor light levels and average light levels were higher in boys than in girls, possibly given boys had a longer duration of light exposure, higher number and duration of light exposure episodes. However, gender differences in light levels were not observed in an Australian cross-sectional study, comprising 49 boys and 52 girls (1204±593 lux vs 949±525 lux, p=0.25) of older age (10–15 years).16

In our study, reported time outdoors was higher than the duration of light exposure. These two measures are not analogous, given they are distinct in timescale and sampling resolution, and importantly, assess different aspects of outdoor exposure. Previous studies have demonstrated a poor correlation (as low as r=0.07) between reported time outdoors and the duration of light exposure.25 Thus, the use of both reported and objective instruments has been recommended to fully capture the risks of outdoor environments.25

Light levels and duration of light exposure were highest during mid-day, compared with the morning or evening. Among 43 Australian children, the hourly duration of light exposure similarly peaked at midday (between 13:00 and 14:00).15 Higher mid-day light levels may reflect higher ambient light levels or schedules allowing for longer duration outdoors, possibly during lunch breaks or the commute home after school.

Across daylight hours, Singaporean children exhibited few light exposure episodes, each lasting 6.3±4.5 min/episode. A study evaluating the frequency of light exposures reported higher daily light exposure episodes in 43 Australian compared with 69 Singaporean children aged 8—12 years (6.9±1.5 episodes/day vs 4.6±1.5 episodes/day, p=0.02), though the duration of light exposure episodes was comparable (18±16 min/episode vs 15±14 min/episode, p=0.54).15 In our study, the low frequency and duration of daily light exposure episodes corroborate with the low light levels and limited outdoor opportunities in Singaporean children.

Types of common outdoor activities were outdoor walks, neighbourhood play and swimming. Similarly, playtime and walking were the most common outdoor activities among 1450 children aged 6–19 years in the USA.26 Thus, group-based outdoor play activities, increasing the availability and accessibility to outdoor facilities in schools or communities (eg, age-appropriate playgrounds and park trails), may aid in promoting time outdoors.

Increasing reported time outdoors by every additional hour per day was associated with lower odds of myopia (OR=0.82, 95% CI 0.70 to 0.95; p=0.009) but was not associated with SE or AL. Spending a longer duration outdoors likely confers a greater degree of protection against myopia. This protective effect of time outdoors concurs with existing studies.4 18 22 The importance of increasing time outdoors on a daily basis received further support from two outdoor trials.4 6 Chinese schoolchildren receiving a daily 40 min outdoor intervention had lower myopia incidence (difference: −9.1%, 30.4% vs 39.5%) and less myopic SE (0.17 D, 95% CI 0.01 D to 0.33 D) over 3 years.4 An additional 80 min/day outdoors effectively reduced myopia incidence (difference: −9.3%, 8.4% vs 17.7%) over a shorter period of 1 year in a Taiwanese trial.6 Moreover, in the current study, Singaporean children are mostly exposed to light levels of <5000 lux while outdoors. Taken together, even at moderate light levels, increasing daily time outdoors seems to be protective. Moderate light-level conditions are easily achievable in most outdoor spaces during daylight hours in Singapore, under various street lighting conditions (~6000 to 16 000 lux) or even tree shade (~5556 to 7876 lux).27 Given the low daily reported time outdoors, a practical recommendation is to simply lengthen the daily duration spent outdoors as much as possible, even under moderate light levels, as a longer duration seems to be a key factor conferring myopia protection.

Light exposure measures (light levels, duration, timing and frequency) across all days were not associated with myopia. On the weekdays, lower outdoor light levels and shorter duration of light exposure episodes were associated with myopia and AL, respectively. Given there are five weekdays in a week, the effects of weekday light exposure measures might be more pronounced or similar to the true light exposure patterns, compared with weekend measures. However, overall, evidence suggests the duration spent outdoors remains the unifying feature underlying protective outdoor exposure, compared with intense light levels, specific activity types or outdoor behaviours. Beyond intense light levels, several proposed hypotheses provide alternative explanations linking time outdoors to myopia protection. These include the chromaticity of light,28 29 spatial frequencies,30 less peripheral defocus and more uniform dioptric structures of outdoor visual environments, which have been shown to inhibit ocular growth.31

School-based and community-based outdoor programmes are important to effectively increase daily time outdoors. Public education on myopia and engagement of multiple stakeholders (children, parents, educators and policymakers) promote sustainable measures,32 particularly in regions where education is prioritised. Positive parental attitudes on after-school outdoor activities and support for educators on outdoor learning initiatives, particularly for preschool children, are crucial.32 In China and Taiwan, national school-based interventions promoting 1–2 hours of daily time outdoors have been implemented.33 The predominantly indoor transmission of COVID-19 further highlights the pertinence of time outdoors, beyond myopia prevention.34

This study combines questionnaires, light metres and activity diaries to evaluate outdoor patterns, with cycloplegic refraction. There are several limitations. First, the limited sample size of 483 children (966 eyes) may be insufficient to supply adequate power to detect the effects of specific light patterns on myopia. Second, temporality needs to be assessed longitudinally. Third, while wrist-worn sensors may underestimate light levels at eye level, reasonable agreements (average differences ≤ 50 lux)35 have been reported.

In summary, this study highlighted the low reported time outdoors and low light levels in Singaporean children aged 9 years. Our findings suggest that increasing daily time outdoors, even under moderate light levels, rather than specific outdoor light levels and patterns, is protective against myopia. While longitudinal studies are needed to confirm these findings, a multipronged approach to increase daily time outdoors is recommended in our combat against the myopia epidemic.

Data availability statement

The data that support the findings of this study are generated from the Growing Up in Singapore Towards healthy Outcomes (GUSTO) study, but restrictions apply to the public availability of these data, due to ethical restriction (patient confidentiality) imposed by the Singhealth Centralised Institutional Review Board. Datasets are, however, available from the authors upon reasonable request and with the permission of the GUSTO study.

Ethics statements

Patient consent for publication

Ethics approval

Ethical approval was provided by the National Healthcare Group Domain Specific Institutional Review Board and SingHealth Centralised Institutional Review Board (IRB protocol number R1517/16/2018). The study adhered to the tenets of the Declaration of Helsinki.

Acknowledgments

The authors thank Naimah Binte Ismail, Tan Chun Wei, Neo Wei Qian, Lim Shan Shan and Andrea Ong for data collection efforts. Further acknowledgements to the GUSTO participants and the GUSTO study group, which includes Allan Sheppard, Amutha Chinnadurai, Anne Eng Neo Goh, Anne Rifkin-Graboi, Anqi Qiu, Arijit Biswas, Bee Wah Lee, Birit F P Broekman, Boon Long Quah, Borys Shuter, Chai Kiat Chng, Cheryl Ngo, Choon Looi Bong, Christiani Jeyakumar Henry, Claudia Chi, Cornelia Yin Ing Chee, Yam Thiam Daniel Goh, E Shyong Tai, Elaine Tham, Elaine Quah Phaik Ling, Evelyn Chung Ning Law, Evelyn Xiu Ling Loo, George Seow Heong Yeo, Heng Hao Tan, Hugo P S van Bever, Iliana Magiati, Inez Bik Yun Wong, Ivy Yee-Man Lau, Izzuddin Bin Mohd Aris, Jeevesh Kapur, Jenny L. Richmond, Jerry Kok Yen Chan, Joanna D. Holbrook, Joao N Ferreira, Jonathan Tze Liang Choo, Jonathan Y Bernard, Joshua J Gooley, Kenneth Kwek, Krishnamoorthy Niduvaje, Kuan Jin Lee, Leher Singh, Lieng Hsi Ling, Lin Lin Su, Ling-Wei Chen, Lourdes Mary Daniel, Marielle V. Fortier, Mark Hanson, Mary Rauff, Melvin Khee-Shing Leow, Mya Thway Tint, Neerja Karnani, Ngee Lek, Oon Hoe Teoh, P. C. Wong, Paulin Tay Straughan, Pratibha Agarwal, Queenie Ling Jun Li, Rob M van Dam, Salome A. Rebello, See Ling Loy, S. Sendhil Velan, Seng Bin Ang, Shang Chee Chong, Sharon Ng, Stella Tsotsi, Chin-Ying Stephen Hsu, Sue Anne Toh, Swee Chye Quek, Victor Samuel Rajadurai, Walter Stunkel, Wayne Cutfield, Wee Meng Han, Yap-Seng Chong, Yin Bun Cheung, Yiong Huak Chan and Yung Seng Lee.

References

Footnotes

  • Twitter @Ray_Najjar

  • Contributors ML was involved in data collection, statistical analysis, data interpretation, drafting, revision and review of the manuscript. CL served as the study coordinator and was involved in data collection, data interpretation, revision and review of the manuscript. ML and CL contributed equally to this manuscript and are joint first authors and guarantors. C-ST provided statistical advice and was involved in data interpretation, revision and review of the manuscript. L-LF was involved in data interpretation, revision and review of the manuscript. C-HS and FY served as site coordinators, advised on clinical aspects of the study, and were involved in patient recruitment, revision and review of the manuscript. RPN was involved in data interpretation, revision and review of the manuscript. CS was involved in data interpretation, revision and review of the manuscript. S-MS conceived, designed, served as the principal investigator and guarantor of the study. She was involved in data interpretation, revision and review of the manuscript.

  • Funding This study was supported by the Agency for Science Technology and Research, Singapore (A*STAR) and JANSSEN World Without Disease Grant (grant JRBMRR151701).

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • Highlights from this issue
    Frank Larkin