|
Abstract
|
We commend Yasmina Molero and colleagues for their study on associations between statin use and neuropsychiatric outcomes. With use of a within-individual design, the authors compare the incidence of neuropsychiatric outcomes during periods on and off statins for each participant, which avoids time-invariant confounders. However, we echo the authors' concern of reversed causality and would like to highlight the potential underuse of statins among patients with psychiatric comorbidity.
Since 2006, the Swedish National Board of Health and Welfare has regularly assessed the quality and efficiency of Swedish health care, focusing on inequalities in health. Differences have been observed in statin use in the diabetic population according to history of psychiatric disorders. For the years 2006 to 2020, we analysed filled prescriptions of statin use in the diabetic population, aged 40-79 years, and compared statin use between those with and without psychiatric comorbidity within the past 5 years. In those with no psychiatric comorbidity, statin use increased from 51·6% to 67·4% between 2006 and 2020. Corresponding numbers for those with at least two hospitalisations or visits with a diagnosis of psychosis, anxiety or depression, or substance use disorder were 35·1% to 59·7%, 47·1% to 62·6%, and 31·7% to 54·8%, respectively (appendix). The markedly increased use of statins over time among those with psychiatric comorbidity could be in part because in earlier years, mainly hospitalised patients were registered, whereas in later years, patients in psychiatric specialised outpatient care, presumably with a less severe psychiatric condition, were also registered. Nevertheless, statin use is much lower in diabetic patients with psychiatric comorbidity than those without psychiatric comorbidity.
As stated by Molero and colleagues, a reduced adherence to medication in individuals with psychiatric comorbidity could be one explanation for this finding. Also, despite Sweden having a universal health-care system, financial constraints could contribute to patients not having their prescriptions filled. However, the analyses of patients with diabetes show a clear difference in statin use between those with and without psychiatric comorbidity. Patients with diabetes and psychiatric comorbidity presumably smoke more, exercise less, have a less healthy diet and are more often overweight, than those with no psychiatric comorbidity. Hence, the benefit of statin treatment in those with psychiatric comorbidity is presumably higher. We argue that there is a systematic underuse of statins among patients with psychiatric comorbidity, because they less frequently have filled prescriptions despite them being in higher need. This reverse causality in studies of statin use and psychiatric disorders must be acknowledged and accounted for in future studies...
Language: en |