Depression as a risk factor for adverse outcomes and increased healthcare utilization in bariatric surgery patients

Depression is strongly associated with obesity and is common among patients undergoing bariatric surgery. Little is known about the impact of depression on early postoperative outcomes or its association with substance use.

The Michigan Bariatric Surgery Collaborative is a statewide quality improvement program that maintains a large clinical registry. We evaluated patients undergoing primary Roux-en-Y gastric bypass or sleeve gastrectomy between 2017 and 2022. Patients self-reported symptoms of depression (PHQ-8) and use of alcohol (AUDIT-C), smoking, prescription opiates, and marijuana at baseline. Preoperative PHQ-8 scores stratified patients based on severity: no depression (0-4), mild (5-9), moderate (10-14), or severe (15-24). We compared 30-day outcomes and substance use between patients with and without depression.

Among 44,301 patients, 30.8% had some level of depression, with 19.8% mild, 7.5% moderate, and 3.5% severe. Patients with depression were more likely to have an extended length of stay (LOS) (> 3 days) than those without depression (no depression 2.1% vs. severe depression 3.0%, p = 0.0452). There were no significant differences between no depression and severe depression groups in rates of complications (5.7% vs. 5.2%, p = 0.1564), reoperations (0.9%, vs. 0.8%, p = 0.7394), ED visits (7.7% vs. 7.8%, p = 0.5353), or readmissions (3.2% vs. 3.9%, p = 0.3034). Patients with severe depression had significantly higher rates of smoking (9.7% vs. 12.5%, p < 0.0001), alcohol use disorder (8.6% vs. 14.0%, p < 0.0001), opiate use (14.5% vs. 22.4%, p < 0.0001) and marijuana use (8.4%, vs. 15.5%, p = 0.0008).

This study demonstrated that nearly one-third of patients undergoing bariatric surgery have depression, with over 10% in the moderate to severe range. There was a significant association between preoperative depressive symptoms and extended LOS after bariatric surgery, as well as higher rates of smoking and use of marijuana, prescription opiates and alcohol. There was no significant effect on adverse events or other measures of healthcare utilization.