GP data can provide unique insights into common health conditions, new research looking at insomnia symptom prevalence in England has shown. The University of Bristol-led study, published in BMJ Open today [8 May], also highlights the value of improving access to this data for future health research.
The study aimed to explore how useful GP records are in measuring how many people experience insomnia symptoms. The researchers used data from the UK Biobank, a database which contains health, genetic and lifestyle information on around half a million participants.
The research team directly compared the proportion of English participants reporting insomnia symptoms in a questionnaire completed when they signed up to the UK Biobank to the same people's linked GP records.
The study found 29 per cent of the sample reported having insomnia symptoms. Of those, only 10 per cent had insomnia symptoms documented in their GP records
Most notably, the research also revealed that over a quarter of people who had insomnia symptoms noted in their GP record in the month immediately before they completed the UK Biobank questionnaire, and had an accompanying prescription for insomnia medication, didn't report having insomnia symptoms in the questionnaire.
This could be due to the stigma associated with having insomnia or because medication reduced the symptoms. Either way, it suggests that GP data can play a vital role in identifying people with insomnia who are not captured by questionnaire data.
The researchers found that many of the characteristics of people visiting their GP with insomnia symptoms were similar to those reporting insomnia symptoms in the questionnaire. These included being female, older, having poorer physical and mental health, smoking, having a high caffeine intake and doing little exercise.
However, insomnia cases identified in the GP data were more likely to report snoring. This could be due to the fact that insomnia and obstructive sleep apnoea - of which snoring is a symptom - often occur together.
Melanie de Lange, a Wellcome Trust-funded epidemiology PhD student in the Bristol Medical School: Population Health Sciences (PHS) and MRC Integrative Epidemiology Unit, and study author, said: "GP records provide rich longitudinal data on common health conditions that are not necessarily picked up by other data sources. Using GP data that was directly linked to the UK Biobank enabled us to gain detailed insights into the characteristics of those suffering from insomnia. This data has huge potential to improve our understanding of numerous other health conditions treated by GPs, including depression, diabetes, anxiety and arthritis."
The study findings suggests that, in addition to providing a useful measure of the impact of insomnia on NHS resources, UK Biobank GP data are a valuable source of information for researchers wanting to explore the risk factors for insomnia.
This is particularly useful when these risk factors are lifestyle choices, such as smoking or exercising, that could be targets for interventions to help prevent or reduce insomnia symptoms. As insomnia itself has been linked to a range of other health problems, including depression, dementia and diabetes, this could have a beneficial impact on people's wider health.
Currently linked GP data is only available for less than half of UK Biobank participants. This is despite participants giving consent for UK Biobank to access their medical records when they signed up over 15 years ago.
The release of data is, at present, controlled by individual GPs who may not have the time to do so or may be worried about regulations surrounding sharing the data. It is hoped that in the future all participants' GP data may become available centrally via NHS Digital. This would help to accelerate research into the wide range of common health conditions managed in primary care.
Source:
Journal reference:
de Lange, M. A., et al. (2024). Insomnia symptom prevalence in England: a comparison of cross-sectional self-reported data and primary care records in the UK Biobank. BMJ Open. doi.org/10.1136/bmjopen-2023-080479.