AUTHOR=Puzhko Svetlana , Schuster Tibor , Barnett Tracie A. , Renoux Christel , Rosenberg Ellen , Barber David , Bartlett Gillian TITLE=Evaluating Prevalence and Patterns of Prescribing Medications for Depression for Patients With Obesity Using Large Primary Care Data (Canadian Primary Care Sentinel Surveillance Network) JOURNAL=Frontiers in Nutrition VOLUME=7 YEAR=2020 URL=https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2020.00024 DOI=10.3389/fnut.2020.00024 ISSN=2296-861X ABSTRACT=

Introduction: Depression is a serious disorder that brings a tremendous health and economic burden. Many antidepressants (AD) have obesogenic effects, increasing the population of obese patients at increased risk for a more severe disease course and poor treatment response. In addition, obese patients with depression may not be receiving the recommended standard of care due to “obesity bias.” It is important to evaluate prescribing pharmacological treatment of depression in patients with obesity.

Objectives: To describe the prevalence and patterns of AD prescribing for patients with depression and comorbid obesity compared with normal weight patients, and to examine the association of prescribing prevalence with obesity class.

Methods: Study sample of adult patients (>18 years old) with depression was extracted from the national Canadian Primary Care Sentinel Surveillance Network (CPCSSN) Electronic Medical Records database for 2011–2016. Measures were prescribing of at least one AD (outcome) and body mass index (BMI) to categorize patients into weight categories (exposure). Data were analyzed cross-sectionally using descriptive statistics and mixed effects logistic regression model with clustering on CPCSSN networks and adjusting for age, sex, and the comorbidities.

Results: Of 120,381 patients with depression, 63,830 patients had complete data on studied variables (complete cases analysis). Compared with normal weight patients, obese patients were more likely to receive an AD prescription (adjusted Odds Ratio [aOR] = 1.17; 95% Confidence Interval [CI]: 1.12–1.22). Patients with obesity classes II and III were 8% (95% CI: 1.00, 1.16) and 6% (95% CI: 0.98, 1.16) more likely, respectively, to receive AD. After imputing missing data using Multiple Imputations by Chained Equations, the results remained unchanged. The prevalence of prescribing >3 AD types was higher in obese category (7.27%, [95% CI: 6.84, 7.73]) than in normal weight category (5.6%; [95% CI: 5.24, 5.99]).

Conclusion: The association between obesity and high prevalence of AD prescribing and prescribing high number of different AD to obese patients, consistent across geographical regions, raises a public health concern. Study results warrant qualitative studies to explore reasons behind the difference in prescribing, and quantitative longitudinal studies evaluating the association of AD prescribing patterns for obese patients with health outcomes.