Obstructive sleep apnoea: quantifying its association with obesity and snoring.
R. Parekh,Edward Green,A. Majeed
DOI: https://doi.org/10.4104/pcrj.2012.00091
2012-10-22
Primary Care Respiratory Journal
Abstract:Obstructive sleep apnoea (OSA) is a chronic condition with an estimated prevalence of between 2-3% in the UK, although a large proportion of individuals with OSA remain undiagnosed. It is caused by repetitive pharyngeal collapse during sleep, which is restored with recurrent arousals and activation of the sympathetic nervous system. Interest around the condition has increased following evidence showing an association of OSA with hypertension, type 2 diabetes mellitus and coronary artery disease, together with adverse health outcomes. The burden of obesity remains one of the largest public health issues facing the UK. Studies suggest that 67% of men and 69% of women in England are obese or overweight. Such levels of obesity lead to wide ranging public health consequences, including significant medical, social, and financial burdens on individuals and society. Estimates suggest that the annual direct cost of obesity to the UK National Health Service (NHS) stands at £3.2 billion. The relationship between OSA and obesity remains complex and multi-factorial; a recent review by Carter & Watenpaugh demonstrates the bidirectional causality between OSA and obesity, which can lead to a cycle of worsening symptoms of both conditions. This relationship is well documented in the literature. However, there has been no previous study to quantify this association within the general population in the UK. The study by Wall and colleagues in this issue of the PCRJ is the first time the strength of association between OSA, snoring and obesity has been investigated in the UK using a large primary care database. In a cross-sectional study of records from The Health Improvement Network (THIN) database, and controlling for confounding factors such as gender, age, region, and socioeconomic status, they used logistic regression techniques to explore the relationship between body mass index (BMI), a history of snoring, and a diagnosis of OSA in the over-50s in the UK. They report that 88% of patients aged 50 and over had a recorded BMI, and that 55% of patients with a recorded BMI were overweight or obese, with the prevalence of OSA increasing progressively with increasing BMI. 96% of patients with a diagnosis of OSA had a recorded BMI. Perhaps the most important result from the study is that individuals with a BMI over 40 kg/m were 27 times more likely to have a diagnosis of OSA when compared to those with a normal BMI. 27% of people with a diagnosis of OSA are recorded as being snorers, an equivalent multivariate analysis odds ratio for OSA in snorers of 26. Interestingly, the levels of OSA decreased with increasing age and deprivation after controlling for confounding factors. When comparing these results to those from other studies, the initial striking factor is the low prevalence of a diagnosis of OSA (0.6%) in those with a recorded BMI. The reasons for this may include coding variability on primary care computer systems, under-diagnosis by clinicians, or people with OSA not presenting to primary healthcare services. The possibility of selection bias is also discussed, with regard to general practitioners (GPs) being more likely to suspect and diagnose OSA in a patient with a high BMI. This could lead to patients with OSA and a normal BMI being under-represented within this study. The inverse relationship of deprivation and age with OSA is a new finding, which requires further investigation. This research highlights some important points for primary healthcare policy and practice. The prevalence of OSA seen here is lower than previous estimates in the literature, based on studies of high-risk populations. The figure in this paper may, however, represent a more accurate reflection of the current situation in the UK. One possible reason suggested for the low OSA prevalence is that a range of computer Read-codes may be utilised for the same condition. Therefore, it is essential that clinicians accurately code diagnoses in their practice to enable a true prevalence to be calculated for allocation of resources. Clinicians should also be encouraged to ask about OSA, particularly when managing people with significant obesity. The strong degree of association of OSA with obesity and snoring indicates that these factors should be central in flagging GPs towards the condition and referring to secondary care for diagnostic studies. However, there is also a significant proportion (40%) of total OSA diagnoses in people with BMIs under 30 kg/m, and so GPs should be open-minded about this diagnosis in symptomatic individuals who are not obese. The finding that people from lower socio-economic backgrounds Obstructive sleep apnoea: quantifying its association with obesity and snoring