Despite their cardiovascular benefits, statin use has been associated with a wide array of actual or perceived psychological and cognitive adverse events. The objective of this study was to compare baseline characteristics and the risk of developing psychological and cognitive disorders between persistent and nonpersistent statin users. We performed a retrospective cohort study (October 1, 2003, to March 1, 2010) of 13,626 statin users in a regional US military health-care system. The persistence of statin use was defined by cumulative pharmacy fill data. Outcomes were the occurrence of psychological diseases during follow-up using prespecified groups based on International Classification of Diseases, Ninth Revision , codes: (1) schizophrenia and psychosis, (2) major depression and bipolar disorders, (3) all psychological diseases, and (4) dementia and cognitive disorders. Statin users who were nonpersistent at 2 years were younger, less likely to be men, and had fewer co-morbidities than persistent users. They were also more likely to be diagnosed with schizophrenia or psychosis (odds ratio [OR] 1.58, 95% confidence interval [CI] 1.20 to 2.10) and cognitive disorders (OR 1.56, 95% CI 1.19 to 2.03) during follow-up compared with persistent users. There was not an association between nonpersistence at 2 years and the development of depression and bipolar disorders (OR 0.99, 95% CI 0.85 to 1.15) or combined psychological diseases (OR 0.97, 95% CI 0.86 to 1.09). Cumulative persistence with statin therapy as a continuous measure was associated with less risk of all outcomes. In conclusion, persistent statin users did not demonstrate an increase in the diagnosis of psychological disorders compared with nonpersistent users. Nonpersistent statin use was associated with a greater likelihood of being diagnosed with psychotic or cognitive disorders.
Highlights
- •
This is a retrospective cohort study of 13,626 statin users who were followed longitudinally.
- •
We compared baseline characteristics and risks of psychological and cognitive disorders between persistent and nonpersistent statin users.
- •
Nonpersistent statin users were younger and have fewer co-morbidities than persistent users.
- •
Nonpersistent statin users were more likely to be diagnosed with psychological and cognitive disorders.
Several conflicting reports have examined the association of statins (hydroxymethylglutaryl coenzyme-A reductase inhibitors) and the risk of psychological and cognitive disorders. Beyond the issue of whether statin use is associated with psychological side effects, the relation between persistence with statin therapy and the development of these disorders is unclear. Understanding such a relation, if one exists, is important for both providers and patients to improve guideline-based adherence to statin therapy. To our knowledge, no cohort or randomized controlled study has compared the prevalence of psychological disorders in statin users based on patient persistence with therapy. The objective of this study was to compare the baseline characteristics and risk of developing psychological and cognitive disorders between statin users who are persistent and nonpersistent with statin therapy.
Methods
This study was approved by the Institutional Review Board at the Brooke Army Medical Center. All inpatient and outpatient medical encounters, diagnoses, and medication fill histories were retrieved using the Military Health System Management Analysis and Reporting Tool (M2), as previously published. The study was divided into 2 periods: baseline period (October 1, 2003, to September 30, 2005) and follow-up period (October 1, 2005, to March 5, 2010). We included all patients aged 30 to 85 years who were enrolled in the San Antonio Military Area as Tricare Prime or Plus, who had at least 1 encounter during the baseline and follow-up periods and received a statin during the fiscal year 2005 for ≥90 days. Using International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] codes, we excluded trauma and burn patients (as defined by the Agency for Health Research Quality Clinical Classifications Software [AHRQ-CCS] category 240) and previous publications. AHRQ-CCS is a diagnosis and procedure categorization scheme that can be used in analyzing data on diagnoses and procedures using ICD-9-CM codes. AHRQ-CCS collapses multitude of diagnoses and procedures into a smaller number of clinical categories of diseases.
We calculated the cumulative days of statin use as the sum of statin supply days that were dispensed from the pharmacy. Our data captured all medication supplies regardless of pharmacy site or affiliation. Because there is no standard definition for persistence with statin therapy, we used the following definitions: (1) 1-year nonpersistent use: defined as patients who used statins for a cumulative period of 90 to 364 days, whereas persistent users had a cumulative statin use of ≥365 days; (2) 2-year nonpersistent use: defined as patients who used statins for a cumulative period of 90 to 729 days, whereas persistent users at 2 years had a cumulative statin use of ≥730 days; (3) 4-year nonpersistent use: defined as patients who used statins for a cumulative period of 90 to 1,459 days, whereas persistent users at 4 years had a cumulative statin use of ≥1,460 days; and (4) cumulative years of persistence with statin therapy as a continuous measure for persistence.
The occurrence of an ICD-9-CM code during the follow-up period, as defined later, in either the inpatient or outpatient setting constituted an outcome event. We used prespecified outcome diagnosis groups to define psychological diseases as described in a previous publication. Briefly, these outcomes were (1) schizophrenia and other psychosis: including codes for schizophrenia, schizoaffective disorders, and other psychosis as identified in the Veterans Aging Cohort Study ; (2) depression and bipolar disorder: including codes for major depression, bipolar disorders, and post-traumatic stress disorder as identified in Veterans Aging Cohort Study; (3) all psychological diseases and disorders: consisting of all ICD-9 codes for psychological diseases as identified by AHRQ-CCS, except for categories of childhood or developmental psychiatric disorders (categories 654 and 655); this group included adjustment disorders (category 650), anxiety disorders (category 651), attention-deficit and disruptive behavior disorders (category 652), impulse control disorders (category 656), mood disorders (category 657), personality disorders (category 658), schizophrenia and other psychotic disorders (category 659), and miscellaneous disorders (category 670); and (4) cognitive disorders: including codes for delirium, dementia, and amnestic and other cognitive disorders as identified by AHRQ-CCS (category 653).
Patients’ baseline characteristics are described using the Charlson comorbidity index described by Deyo et al, social habits, presence of psychological diseases at baseline, and other comorbid conditions and medication use as in shown in Table 1 . Baseline characteristics were compared using chi-square and Student t tests as appropriate. Comparisons were considered statistically significant if the calculated 2-tailed p value was ≤0.05. We used separate multivariate logistic regression models to examine the risk of each outcome in which each psychological/cognitive outcome was a dependent variable, and persistence measure was an exposure variable (each of 1-year nonpersistent, 2-year nonpersistent, 4-year nonpersistent, and cumulative years of persistence with statin therapy in a separate model), adjusting for potential confounders. Statistical analyses were performed using SPSS statistical software, version 19 (IBM Corp, Armonk, New York).
