What long-term risks are still unknown for tirzepatide after 3–5 years?

The short version: no retrieved study follows tirzepatide users past 88 weeks, so the long-term risk profile after 3–5 years is simply not in this evidence.

The longest trial here is the SURMOUNT-4 extension at 88 weeks, which measured weight regain after stopping, not safety [1]. The sleep apnea trial ran 52 weeks and explicitly notes longer-term safety data was not provided [2]. The 2025 review flags that long-term safety evaluation is still needed for pancreatic, biliary, retinopathy, and thyroid concerns [4]. These are open questions, not answered ones.

What the evidence does show is that adverse events are common. The network meta-analysis found tirzepatide users had about 2.8 times the risk of adverse events versus placebo during the trial period [3]. Gastrointestinal side effects are the known trade-off [4]. But none of these records track what happens at year three, four, or five.

The missing pieces are the ones that matter most: whether years of GLP-1/GIP stimulation changes pancreatic or thyroid tissue, whether the weight loss is maintained or rebounds, and what happens to the cardiovascular system over time. The evidence here measures efficacy in months and calls it a day. No retrieved study compares long-term users to a genuinely untreated group.

My call: the evidence cannot support any claim about 3–5 year risks, because no such data appears in these records. Confidence: not clear — the absence is real, not a search artifact.