OBJECTIVES : Higher body mass index (BMI) relates to increased risk of nonalcoholic fatty liver disease (NAFLD)/nonalcoholic steatohepatitis (NASH). We sought to assess the association of bariatric surgery in obese individuals on NAFLD/NASH risk.

METHODS : Patients who underwent bariatric surgery or cholecystectomy (control surgery) were identified in TriNetX “Analytics” (TriNetX, Cambridge, MA), a live electronic medical record (EMR) database of >50M patients, mostly from the US (analysis 13-May-2019). Patients had to have a record of obesity (BMI measurement ≥30 kg/m² or ICD-10 code) within 6 months before surgery. A third group consisted of obese patients without bariatric surgery or cholecystectomy. The index event in the surgery groups was the date of surgery and in the non-surgery group the first record of obesity. Patients had to have EMR data ≥1 year before and ≥2 years after index event. Those with histories of NAFLD/NASH (ICD-10 K76.0/K75.81) or gastrointestinal cancer were excluded.

RESULTS : Incidence of NALFD/NASH per 1000 population was 18.8 following bariatric surgery (N=2,233 patients), 27.4 following cholecystectomy (N=4,641) and 32.0 in the non-surgery cohort (N=1,389,271). Compared to bariatric surgery, the risk ratio (95% CI) of NAFLD/NASH adjusted for age, sex, race, and type 2 diabetes was 1.65 (1.12 to 2.43) in cholecystectomy patients and 2.33 (1.63 to 3.33) in non-surgery patients. Risk did not differ between cholecystectomy patients and non-surgery patients (adjusted risk ratio 0.89, 95% CI 0.70 to 1.12).

CONCLUSIONS : NAFLD/NASH in obese patients is less often diagnosed after bariatric surgery than after cholecystectomy or no surgery. This may indicate that weight loss associated with bariatric surgery may halt or at least partly reverse the underlying pathophysiology leading to NAFLD/NASH. Nonalcoholic fatty liver disease frequently goes undiagnosed, and the rate of diagnosis will be affected by the use of imaging techniques and diagnostic biopsies in each group.