In Scotland, major mental illness reduces life expectancy by approximately 17 years, with cardiovascular disease (CVD) the leading cause of death. We aimed to investigate whether history of major mental illness affects CVD risk in people with type 2 diabetes mellitus (T2DM) in Scotland.Methods
We identified adults diagnosed with T2DM between 2004 and 2016 from a national diabetes register (the Scottish Care Information – Diabetes database). We ascertained history of mental illness from psychiatric and acute hospital admission records and incident/recurrent stroke or myocardial infarction (abbreviated to CVD) from acute hospital admission and mortality records. Using Cox regression analysis, we obtained hazard ratios (HRs) for CVD risk among people with a history of schizophrenia, bipolar disorder or depression, compared to those with T2DM but without a history of mental illness.Results
We included 2 12 011 people with T2DM. Of these, 2107 (1.0%) had schizophrenia, 1521 (0.7%) had bipolar disorder and 5288 (2.5%) had depression. People with these major mental disorders were younger at diagnosis of T2DM and had higher prevalence of smoking and history of alcohol use disorders than the comparison group. After adjusting for age, sex, area-based deprivation, hypertension and previous history of CVD, HRs (95% confidence interval) were 1.27 (1.05, 1.53) for schizophrenia, 1.45 (1.23, 1.70) for bipolar disorder and 1.50 (1.37, 1.65) for depression compared to those without a history of mental illness. Additional adjustment for smoking and alcohol use disorder attenuated effect estimates to 1.12 (0.93, 1.36) for schizophrenia, 1.35 (1.15, 1.59) for bipolar disorder and 1.37 (1.25, 1.51) for depression.Conclusion
Schizophrenia, bipolar disorder and depression are associated with increased CVD risk among people with T2DM and this is partially explained by higher prevalence of smoking and alcohol use disorders in people with these conditions. This highlights the need for better CVD prevention in people with T2DM and major mental illness especially with respect to modifiable risk factors. Major strengths are that the diabetes register includes 99% of all people with diabetes in Scotland and the follow-up period is relatively long. However, our definition of mental illness was limited to hospital admission data only and so findings may not apply to people with mental illness not admitted to hospital. Future analyses will also include adjustment for psychotropic medication use.