Marijuana Consumption Leads to Lower Weight Loss Following Laparoscopic Sleeve Gastrectomy

Moneer E Almadani, MD, Cristina R Harnsberger, MD, Ryan C Broderick, MD, Hans F Fuchs, MD, Joshua Langert, MD, Christopher Ducoin, MD, Bryan J Sandler, MD, Santiago Horgan, Alberto S Gallo, MD, Garth R Jacobsen. University of California San Diego UCSD

Introduction: Broader acceptance and legalization of both recreational and medical marijuana use is occurring in the United States. Although it is known to stimulate appetite, the influence of marijuana use on body mass index (BMI) is disputed in literature. Data is scarce regarding the effect of marijuana use in patients who undergo bariatric surgery, and thus, our aim was to determine whether its use has an effect on weight loss following laparoscopic sleeve gastrectomy (LSG).

Methods: Patients that underwent LSG were identified retrospectively using a prospectively maintained database. All patients were evaluated preoperatively by a multidisciplinary team, which includes a psychological stability evaluation to undergo surgery. Patients were included if they used marijuana preoperatively, either illicit or prescribed, and these patients were compared to those without marijuana use. An exclusion criterion was the usage of other illicit substance. Primary endpoint was percent excess weight loss (%EWL) at 3, 6, and 12 months, and secondary endpoints included intraoperative blood loss, operative time, length of hospitalization, and complications.

Results: Between May, 2006 and August, 2014, 250 patients underwent LSG. Of this cohort, 8 patients (3.2%) were identified that used marijuana preoperatively. Eleven patients were excluded as they abuse other than marijuana. BMI at initial clinic visit was significantly higher in those who used marijuana compared to those who did not, with a mean of 55.93 vs. 46.93 (p=0.010). %EWL was significantly lower in those who used marijuana compared to those who did not at 6 and 12 months follow-up with means of 31.0% vs. 47.1% (p=0.042), and 33.3% vs. 54.9% (p= 0.002), respectively (Figure 1). All secondary endpoints were similar between the two groups.