Abnormal polysomnographic (PSG) features, most notably increased electromyographic (EMG) tone and eye movements during non-REM sleep have been observed during sleep in fluoxetine-treated depressed patients. However, the relationship between these PSG features and sleep disruption is unclear. Nine depressed patients treated with 10 to 80 mg of fluoxetine and six unmedicated, depressed patients were studied polysomnographically on two consecutive nights during which sleep parameters, transient arousals, and eye movements were measured. The fluoxetine group experienced a lower-average sleep efficiency index (SEI) and significantly more eye movements and arousals during non-REM sleep than the control group. Eye movement and arousal counts were significantly correlated. In addition, clinically significant periodic limb movement disorder (PLMD) was observed in 44% of the fluoxetine-treated group versus none of the control group. We conclude that a higher incidence of PLMD and frequent transient arousals associated with eye movements may be responsible in part for the complaint of insomnia made by patients treated with fluoxetine.

Supplemental Digital Content is available in the text.

Research Objective Socioeconomic disparities in mental health care use can be mitigated by interventions to systematically improve disease diagnosis and treatment. Beginning 2010, Los Angeles County Departments of Health Services (DHS) and Mental Health (DMH) collaborated to increase access to effective mental health treatment. Implemented in staggered fashion among select safety‐net clinics, the Mental Health Integration Program (MHIP) embedded behavioral health specialists to co‐manage primary care patients (e.g., psychiatric consultation support for medication prescription, brief problem‐focused psychotherapy). This study examined whether clinic MHIP implementation was associated with higher treatment initiation (medication receipt and/or specialist follow‐up [for psychotherapy]) in safety‐net clinic sites. Study Design This retrospectively cohort study examined 33,537 primary care patients who were newly diagnosed with depression from June 30, 2009 to July 1, 2014 (n = 39,066 person‐years), using merged medical and pharmacy records from DHS and DMH. Patients were seen in one of 19 clinics that were either: (1) before, (2) after, or (3) had never implemented MHIP. We followed patients for 90 days after a new diagnosis, defined as having no diagnoses in the prior year. Then, we determined whether patients received antidepressant medication (prescription filled per pharmacy records) and/or mental health specialist follow‐up (completed referral documented in medical records). In multilevel regression models, we used MHIP implementation status to predict odds of treatment initiation, controlling for time and patient demographic and clinical characteristics (e.g., comorbidity). Population Studied 255,840 safety‐net patients seen in LA County. Principal Findings In 30.1% of episodes, primary care patients received any treatment within 90 days of a new depression diagnosis (38.7% [after] vs 36.2% [before] vs 25.0% [never implemented MHIP in clinic]). 27.9% received medication within 90 days, while only 3% reached specialists. In multivariate analyses (n = 30,347 person‐years), we found that there was 2.35‐times odds of treatment initiation among patients in clinics after MHIP implementation (odds ratio [OR] = 2.35; 95% confidence interval [CI] = 2.18–2.55; p < 0.001), compared to patients in clinic before implementation. This program effect was driven by medication treatment (OR = 2.42; CI = 2.24–2.63; p < 0.001), as opposed to specialist follow‐up (OR = 1.14; CI = 0.88–1.47; p = 0.32). Odds of treatment initiation were higher among minority (Asian > Latino > Black) than White patients, Spanish‐ than English‐speaking patients, and uninsured than insured patients in the study sample. Conclusions Collaboration between two large, distinct health systems and its resulting integrated care programs were associated with greater odds of treatment initiation among patients newly diagnosed with depression, compared to patients in clinics before implementation or in clinics that did not implement. Findings suggest potential pos...