There is growing evidence that daylight saving clock changes (the clocks moving one hour forward in spring and one hour back in autumn) may have adverse effects on people’s health due to sleep deprivation and the disruption of biological rhythms. For example, studies outside of England have shown that the number of depressive episodes, heart attacks, strokes and fatal road traffic accidents increases in the weeks after the clock changes. This has prompted countries such as the US and the EU to vote to end DST clock changes. However, it is unclear whether DST will be abolished in England.
The purpose of this research is to estimate the effect of daylight savings time clock changes on the health of the English population. To do this we will compare the number of GP and hospital visits for cardiovascular disease, depression and road traffic injuries in the 4 weeks directly before and after the clock changes.
This data will provide evidence as to whether England should abolish DST clock changes. It will also offer insights into the wider effects of sleep and circadian disruption on mental and physical health.
Growing evidence suggests that daylight saving time (DST) clock changes (one hour forward in spring and one hour back in autumn) may have adverse effects on population health, likely via sleep deprivation and circadian disruption. For example, several studies using small datasets from outside England have reported increased incidence of depressive episodes, myocardial infarction, atrial fibrillation, strokes and fatal traffic accidents in the weeks immediately following the clock changes. This has prompted the US and the EU to vote to end DST clock changes. However, it is currently unclear whether England will do the same.
The aim of this study is to estimate the effects of DST transitions on cardiovascular disease, depression and road traffic injuries in the English population. Using CPRD and HES data for 2008-2022, regression discontinuity analysis will be used to compare the average number of primary and secondary care visits for these health outcomes in the 4 weeks before and after the spring and autumn clock changes. Regression discontinuity designs are a statistical method that uses a specific threshold or cut-off point (here the time of DST clock change) to estimate the causal, real-world, effects of policies. Differences in the number of visits will be compared in the 1- (modelled daily), 2- and 4-week periods before and after the clock changes.
Research findings will help to quantify the effects of DST clock changes on English population health by estimating the number of incident cases of disease and recurrent events attributable to DST clock changes. The results will be triangulated to formulate policy recommendations as to whether the English government should abolish DST clock changes. More broadly, the research will also offer insights into the wider effects of sleep and circadian disruption on mental and physical health.
Primary outcomes
All primary/secondary care visits (incident & prevalent cases) in the 1- (modelled daily), 2- and 4-week periods before and after the clock changes, stratified by age (10 year age bands), for:
o Depression (aged ≥10*).
o Road traffic injuries (all ages).
o Cardiovascular disease (aged ≥40**).
*Age ≥10 was chosen to ensure we capture adolescents who may be disproportionately affected by the sleep deprivation caused by the clock changes due to a shift towards having a later chronotype (going to bed and getting up later) during adolescence.
**The age at which people become eligible for their NHS health check and are considered at higher risk of cardiovascular disease.
Secondary outcomes
All primary/secondary care visits (incident & prevalent cases) in the 1- (modelled daily), 2- and 4-week periods before and after the clock changes, aged ≥10, stratified by age (10 year age bands), for:
o Sleep disorders. This analysis will be exploratory, and will only be conducted if it is possible to adequately define sleep disorders within CPRD data.
o Other mental health (anxiety & self-harm) & eating disorders.
The number of subsequent hospitalisations/referrals for depression/cardiovascular disease, or deaths amongst those visiting primary/secondary care for depression/cardiovascular disease in the 1-, 2- and 4-week periods before and after the clock changes.
Detailed definition of outcomes:
Cardiovascular disease, road traffic injuries, self-harm and eating disorders will be defined using ICD10, Read codes and HES A&E codes.
Depression and anxiety will be defined using ICD-10, Read codes and HES A&E codes:
• Read codes alone will be used to define depression/anxiety if the code is specific enough (see code lists C1 and C6 for depression/anxiety diagnoses).
•However, where Read codes relate to symptoms of depression/anxiety (see code lists C2 and C7 for depression/anxiety symptoms) that are not considered adequate to define depression/anxiety, patients will only be counted as having depression/anxiety if they have been prescribed a drug used to treat depression within 90 days.
Sleep disorders will be defined using ICD-10 and Read codes:
• Sleep disorders are difficult to define which means that most Read codes are not specific and relate to symptoms rather than diagnoses. Patients will therefore only counted as having a sleep disorder if they have a Read code for a sleep disorder diagnosis or symptom and have been prescribed a drug used to treat sleep disorders within 90 days.
See Appendix A for dates of clock changes, Appendix B for preliminary ICD10 codes, Appendix C for preliminary Read v2 codes, Appendix D for preliminary prescription codes and Appendix E for HES A&E codes.
*Notes to reviewer:
• Having death as an outcome would be interesting, but unfortunately we do not have the budget to buy the linked ONS data.
• Looking at the different severities of presentation would also be interesting but would be challenging to assess and other studies have struggled to do this using CPRD data.
Kate Tilling - Chief Investigator - University of Bristol
Melanie de Lange - Corresponding Applicant - University of Bristol
Kate Birnie - Collaborator - University of Bristol
Neil Davies - Collaborator - University of Bristol
Sophie Eastwood - Collaborator - University College London ( UCL )
HES Accident and Emergency;HES Admitted Patient Care;Patient Level Index of Multiple Deprivation