Tirzepatide at 6 months: who stays on it

A claims analysis published on 10 August 2026 in Diabetes, Obesity and Metabolism followed 22,512 US adults who started tirzepatide for obesity or overweight without type-2 diabetes, and reported that 68% were still persistent on the drug six months later^[1]. Everyone counted had already filled a prescription, so this is the continuation question rather than the access question.

What the six-month numbers actually say

The study drew on the Healthcare Integrated Research Database, which combines administrative claims with electronic health records from a large US commercially insured population. Adults qualified if they had at least one tirzepatide prescription claim between November 2023 and June 2024, twelve months of continuous medical and pharmacy enrolment before that first claim, and six months after it^[1]. Mean age was 46 years and 73% of the cohort were women. Tirzepatide is a once-weekly dual GIP and GLP-1 receptor agonist, approved in the US for type-2 diabetes, weight loss and obstructive sleep apnoea^[1].

Two different measures get reported here, and they are not interchangeable. Persistence asks whether someone stayed on therapy without a gap longer than a set threshold. Adherence asks how much of the follow-up window was covered by dispensed medication, expressed as the proportion of days covered (PDC). In this cohort 68% (15,208 people) were persistent across the six months, while 55% (12,292) met the usual adherence cut-off of PDC at 80% or higher^[1]. The space between those two figures is the group who stayed nominally on treatment but missed or delayed enough doses to fall under the threshold.

The cohort was not a well one at baseline. Ninety-one percent (20,554 people) had at least one obesity-related complication recorded before their first claim, with dyslipidaemia, hypertension and anxiety the most frequent^[1].

The weight-loss figure rests on a small subgroup

Among persistent individuals with both a pre-index and a post-index weight on file, the mean body-weight reduction was 10.5%, rising to 12.0% in the subgroup who had not previously used a GLP-1 receptor agonist for obesity^[1]. The number worth holding on to is the denominator. That 10.5% comes from 808 people, roughly 5% of the 15,208 who were persistent^[1]. Weight only enters a dataset like this when a clinician happened to record it at a visit, and people who turn up for visits both before and after starting a drug are not a random sample of everyone who started it.

Three further limits apply. The design is retrospective and descriptive with no comparison arm, so it can describe what happened alongside treatment and cannot show that treatment caused it. The follow-up is six months, short next to the 72-week and 88-week horizons of the tirzepatide trial programme. And the work was funded by Eli Lilly, with authors based at Eli Lilly and at Carelon Research^[1]. None of that makes the figures wrong. It does mean they read as a description of routine care, not as an effect estimate.

What this study does not tell you about stopping

The effectiveness results are reported among persistent individuals only. The paper does not describe what became of the roughly one in three people who were not persistent, so it cannot answer the question most readers arrive with: what happens to your weight when you come off the drug^[1].

For that, the randomised evidence is SURMOUNT-4. After 36 open-label weeks on maximum tolerated tirzepatide, during which participants lost a mean 20.9% of body weight, 670 adults were randomised to continue the drug or switch to placebo for a further 52 weeks. Between week 36 and week 88 the continued arm lost another 5.5% while the placebo arm regained 14.0%. By week 88, 89.5% of those still on tirzepatide had held at least 80% of the lead-in weight loss, against 16.6% on placebo^[2].

That is why persistence is treated as an outcome rather than as a compliance footnote. Read together, the two papers say that most of the measured benefit sits with the people who keep taking tirzepatide, and that stopping hands a large share of it back.

How 68% compares, and why people stop

Six months is not twelve, and cohorts are not interchangeable, so treat what follows as direction rather than a like-for-like match. A 2024 analysis of commercial pharmacy and medical claims followed 4,066 adults with obesity and without diabetes who started a GLP-1 receptor agonist in 2021, largely semaglutide and liraglutide products, and reported persistence of 46.3% at 180 days and 32.3% at one year, with 27.2% adherent on PDC^[5]. A larger electronic-health-record cohort of 125,474 adults who began a dual-labelled GLP-1 receptor agonist between 2018 and 2023 found one-year discontinuation of 64.8% in people without type-2 diabetes, against 46.5% in people with it^[3].

Why people stop is better answered by a third paper. Researchers reviewed records for 288 adults with overweight or obesity and without type-2 diabetes who began injectable semaglutide or tirzepatide and came off within a year. Cost or insurance problems were the primary reason for 47.6%, inability to tolerate side effects for 14.6%, and inability to fill because of shortages for 11.8%. Unsatisfactory weight loss accounted for 1.7%^[4]. The large record-based cohort points the same way on tolerability: moderate or severe gastrointestinal adverse events carried a higher hazard of discontinuation (hazard ratio 1.19 in people without type-2 diabetes), while each 1% of body weight lost was associated with a 3.3% lower hazard of stopping^[3].

Practical context

The honest summary of the new paper is narrower than its headline. Among US adults with commercial insurance who started tirzepatide for weight in late 2023 and the first half of 2024, about two in three were still on it at six months, and the small subset with recorded weights had lost around a tenth of their body weight^[1]. What the study cannot say is whether the drug caused that loss, what became of the third who came off, or whether the pattern holds past six months or outside commercially insured populations.

Cost and coverage are doing a lot of the work in these numbers, which is part of why continuation rates differ so much between health systems. Our country-by-country regulatory status page tracks where tirzepatide is licensed and how it is funded. This article is educational and is not medical advice. Decisions about starting, continuing or stopping a prescription medicine belong with a clinician who knows your history.