Introduction It is well established that asthma and obstructive sleep apnoea (OSA) are significantly more prevalent in obese vs. non-obese populations. To date however there is limited data on whether this risk is increased with severity of obesity as most studies classify all patients >30 Kg/m² simply as ‘obese’. In addition, many existing studies’ obese cohorts have fairly low BMI scores compared to patients attending specialist medical obesity services. Our study aims to examine the prevalence of these diseases in higher BMI groups, compare the relative risk of increasing obesity on prevalence of respiratory disease and investigate whether there is a synergistic effect of multiple demographic factors and severity of obesity.

Methods Data was collected from a total of 367 (of whom 159 had a BMI recorded) patient records attending a tier 4 obesity clinic over an 8 month period. Patients were divided into three groups according to severity of obesity, BMI 30–40, 40–50 and >50 Kg/m². Index of multiple deprivation (IMD) scores (mapped to postcodes) were used as a proxy of socioeconomic status.

Results 43% of our total cohort had OSA, including 75.7% of those with a BMI >50 (Multivariate logistic regression OR 10.4 (95% CI 3.33 – 32.7, p < 0.001). In a chi-square analysis, this association was significant in both genders but stronger in males (Cramer’s V 0.481 vs 0.305) and was significantly associated with a worse IMD score, being white and increasing age. 11.6% of the cohort were asthmatic however there was no difference in prevalence between the groups OR 0.175 (95% CI 0.019 – 1.631, p = 0.126). There was however a significant co-effect of being male and increasing BMI in a multi-layer chi-square analysis p = 0.044.

Conclusions Our study highlights a very high prevalence of major respiratory diseases as co-morbidities in a severely obese population. Early data suggests a synergistic effect of Caucasian ethnicity, male gender and IMD score with increasing BMI on the risk of developing OSA (and Asthma for male gender). This is in contrast with our initial findings for Diabetes and Cardiovascular disease where the association is with Asian ethnicity.