Background The degree to which the number of comorbidities and the presence of depression impact health care use and costs has been unexplored employing actual resource use data representing a general population. We aimed to evaluate the impact of number and type of comorbidities on depression prevalence, health care utilisation and health care costs.

Methods Cohort study and included in the study were 300,020 participants aged 30 years of age registered with the UK General Practice Research Database. We used the UK General Practice Research Database to examine participants’ diagnoses and resource use in primary and secondary care from 2005 to 2009. Healthcare unit costs and the costs of each individual prescription item were analysed.

Results In participants with no comorbidity, the age-standardised prevalence of depression was 7% in men and 14% in women. The likelihood of having depression increased in participants with single comorbidities including diabetes mellitus (men 13%, women 22%), CHD (men 15%, women 24%), stroke (men 14%, women 26%) or colorectal cancer (men 10%, women 21%). Patients with concurrent diabetes, CHD and stroke had a very high prevalence of depression (men 23%, women 49%) with women being more likely to suffer depression. Patients with a single comorbidity were 1.63 (95% confidence interval 1.59 to 1.66) times more likely to be depressed than those without comorbidity while those with two and three comorbidities were 1.96 (1.89 to 2.03) and 2.35 (2.03 to 2.59) times more likely. Depression increased the total costs of resources utilised per year in all participants across both genders, all age groups and across all comorbidities analysed in this study. Individuals with depression had higher total annual health care costs (males = £1014, females=£1212) than those without comorbidity or depression (males = £380, females = £517). Those with diabetes alone had £1144 for males and £1393 for females but £ 2534 for males and £ 3017 for females when depression was present alongside diabetes. When patients had diabetes, CHD and had suffered a stroke these patients’ costs were £1541 for males and £1879 for females without depression and £3420 for males and £4072 for females with depression. Depression increased the associated cost of any comorbidity.

Conclusion The prevalence of depression appears to be more strongly determined by the number of comorbidities rather than the precise nature of the comorbid diagnoses. Additional costs of health care utilisation are considerably higher when depression is associated with single or multiple comorbidities.