XENDOS publishes: orlistat cuts diabetes progression over four years

Results from XENDOS, a four-year randomized trial of the fat-blocking drug orlistat, were published in Diabetes Care in 2004. In more than 3,000 adults with obesity, orlistat plus lifestyle counseling lowered the share of people who developed type 2 diabetes from 9.0% to 6.2% compared with placebo plus the same counseling, and produced roughly twice as much weight loss ^[2].

What the trial tested

XENDOS enrolled 3,305 patients aged 30 to 60 with a body mass index of 30 or higher, average weight 242 pounds (110 kg) and average BMI 37 ^[2]. People with existing diabetes, cardiovascular disease or gastrointestinal disease were excluded, and 21% entered with impaired glucose tolerance ^[2]. One group of 1,640 took orlistat 120 mg three times a day with weight-loss counseling; 1,637 took placebo with the same counseling ^[2]. (Those group totals add to 3,277, slightly fewer than the 3,305 enrolled figure listed in the same source ^[2].) Everyone was prescribed a diet with an 800-calorie daily deficit, and an oral glucose tolerance test was done every six months ^[2].

The co-primary outcomes were time to onset of type 2 diabetes and weight change at four years ^[2].

The numbers

After four years, 6.2% of orlistat patients had developed type 2 diabetes versus 9.0% on placebo, a hazard ratio of 0.63 (95% CI 0.46–0.86, p=0.0032) ^[2]. Weight loss averaged 12.8 pounds (about 5.8 kg) with orlistat versus 6.6 pounds (about 3.0 kg) with placebo (p<0.001) ^[2]. LDL cholesterol fell 12.8% from baseline on orlistat versus 5.1% on placebo (p<0.01) ^[2].

Two caveats came with the headline result. First, the diabetes benefit was statistically significant only in the subgroup with impaired glucose tolerance at baseline; among people with normal blood sugar, diabetes was too rare to show a difference ^[2]. Weight loss, by contrast, was similar in both subgroups ^[2]. Second, dropout was heavy: 48% of the orlistat group and 66% of the placebo group left the study before four years (p<0.0001) ^[2].

Side effects tracked orlistat's mechanism. The drug blocks pancreatic and gastric lipase, cutting intestinal fat absorption by as much as 30% at the 120 mg dose ^[2]. Gastrointestinal side effects were reported by 91% of orlistat patients in year one versus 65% on placebo, falling to 36% versus 23% by year four ^[2]. In separate data, first-year rates included oily rectal spotting (26.6%), flatus with discharge (23.9%) and fecal urgency (22.1%), most of which dropped sharply in year two ^[2]. Orlistat patients also had significant declines in vitamins A, D, E and K versus placebo, though average levels stayed within the normal reference range throughout ^[2].

Why it matters for patients

XENDOS is one of the few obesity drug trials that followed patients for four years with a hard clinical endpoint rather than pounds alone. It established a pattern that later weight medications would be measured against: modest average weight loss paired with a measurable drop in new diabetes cases, concentrated in people whose blood sugar was already drifting upward ^[2].

It also set expectations about tolerability and staying power. Nearly half the treated group did not finish four years ^[2], and the GI effects were common early on ^[2], which is relevant context for anyone weighing how long a weight medication can realistically be taken.

For comparison, a separate two-year trial of 729 adults found 7.6% weight loss on orlistat 120 mg three times daily versus 4.5% on placebo, with dropout rates of 35% and 44% ^[2]. Orlistat is sold as prescription Xenical (120 mg) and over-the-counter Alli (60 mg), approved for weight loss or maintenance in adults and adolescents 12 and older with a BMI of 30 or more, or 27 or more with risk factors such as hypertension, diabetes or dyslipidemia ^[2].

What is not known from these sources

The materials available here do not compare XENDOS results with GLP-1 medicines such as semaglutide (Ozempic, Wegovy, Rybelsus) or tirzepatide (Mounjaro, Zepbound), and they do not report whether the diabetes reduction persisted after orlistat was stopped ^[2]. The PubMed listing for the trial was not accessible at the time of writing ^[1].