The TZD story is one of the most dramatic in modern pharmacology. A drug class that arrived with genuine scientific promise — the first agents to actually address insulin resistance at the cellular level rather than just forcing the pancreas to work harder — became the center of one of the largest drug safety controversies and corporate fraud settlements in FDA history. One drug was pulled. One survived but carries a trail of warnings. Both contain lessons that still matter for anyone navigating the T2D medication landscape.
Thiazolidinediones are PPAR-γ (peroxisome proliferator-activated receptor gamma) agonists. PPAR-γ is a nuclear receptor — a protein inside the cell nucleus that regulates the expression of genes involved in fat cell differentiation, glucose uptake, and insulin signaling. By activating PPAR-γ, TZDs make fat cells, muscle cells, and liver cells more responsive to insulin. They are insulin sensitizers, not insulin secretagogues — they do not push the pancreas to produce more insulin. They make existing insulin more effective.
This mechanism is genuinely different from sulfonylureas (which force more insulin output) and metformin (which reduces hepatic glucose production). TZDs work at the gene expression level in peripheral tissues, and they take 4–12 weeks to reach full effect — slower than any other oral diabetes drug.
Rezulin (Warner-Lambert) was the first TZD approved anywhere in the world — January 1997 in the US. Early trials showed meaningful HbA1c reductions and genuine insulin-sensitizing effects. It was a genuinely novel mechanism and received significant prescribing uptake. Then the hepatotoxicity reports began. Liver failure cases — some fatal — accumulated in post-market surveillance. The FDA required progressively stricter liver monitoring protocols. By 1999–2000, with 90 confirmed deaths and over 150 cases of acute liver failure reported, the calculus became undeniable. Rezulin was voluntarily withdrawn from the US market in March 2000. It had been on the market for three years. The Rezulin saga permanently shaped how the FDA and clinicians viewed the entire TZD class — and contributed to the heightened caution that surrounded its successors.
Avandia (GlaxoSmithKline) was FDA-approved in May 1999 and became one of the best-selling drugs in the world — peaking at $3.2 billion in annual sales in 2006. Then came May 2007: Dr. Steven Nissen, a Cleveland Clinic cardiologist, published a meta-analysis in the New England Journal of Medicine analyzing 42 clinical trials involving Avandia. The conclusion: rosiglitazone was associated with a 43% increase in the risk of myocardial infarction and a 64% increase in the risk of cardiovascular death compared to placebo or other diabetes drugs.
The fallout was immediate. FDA advisory committees convened. Avandia sales collapsed. GlaxoSmithKline eventually paid a $3 billion settlement in 2012 — the largest healthcare fraud settlement in US history at the time — after it was found to have promoted Avandia for uses not approved by the FDA and to have downplayed cardiovascular risks. The European Medicines Agency suspended Avandia across Europe in September 2010. The FDA imposed a Risk Evaluation and Mitigation Strategy (REMS) in September 2010 restricting Avandia to patients already using it or unable to use other treatments.
The story has a complicated epilogue. The RECORD trial — GlaxoSmithKline’s own cardiovascular outcomes trial — was reanalyzed with full patient-level data in 2013. An independent adjudication committee found the CV risk was not confirmed at the level Nissen’s 2007 meta-analysis suggested. The FDA lifted most of the REMS restrictions in November 2013. Avandia is technically available today — but prescribers have largely moved on, and its reputation has never recovered.
Actos (Takeda) was also FDA-approved in July 1999 — the same year as Avandia — and is the only TZD still widely used today. The PROactive trial (2005) examined pioglitazone in T2D patients with macrovascular disease. The primary composite endpoint did not reach statistical significance, but a pre-specified secondary endpoint — death, myocardial infarction, and stroke — showed a significant 16% reduction. This positive CV signal distinguished pioglitazone from rosiglitazone in clinicians’ minds and helped preserve its prescribing base even as Avandia collapsed.
Actos survived the TZD storm but picked up its own warning. A 2011 FDA safety communication flagged a potential increased risk of bladder cancer with long-term use (more than 1 year). The signal came from a 10-year Kaiser Permanente epidemiological study. France and Germany suspended Actos temporarily in 2011. The FDA added a bladder cancer warning to the label. Subsequent analyses have shown a modest signal — roughly 40% increased relative risk after 2+ years — though absolute risk remains low. Pioglitazone is contraindicated in patients with active bladder cancer or a history of bladder cancer.
An unexpected second act for pioglitazone: it is now one of the most studied agents for non-alcoholic steatohepatitis (NASH) / metabolic dysfunction-associated steatohepatitis (MASH) — fatty liver disease closely tied to T2D and insulin resistance. Multiple randomized trials show pioglitazone improves liver histology in NASH, reducing steatosis, inflammation, and fibrosis. As NASH/MASH has emerged as a major complication in T2D patients, pioglitazone has found renewed clinical relevance. It is also available as a generic — among the cheapest diabetes medications available.
Weight gain — Significant and class-defining. PPAR-γ activation promotes fat cell differentiation and fat storage. Average weight gain: 2–4 kg in trials, but some patients gain considerably more. This is fat redistribution as well as simple caloric storage — pioglitazone tends to shift fat from visceral (abdominal, more metabolically harmful) to subcutaneous stores, which may partially offset the metabolic harm of the absolute weight gain.
Fluid retention and edema — TZDs cause sodium and water retention via PPAR-γ receptors in the kidney. Peripheral edema (swelling in the legs and ankles) occurs in 4–8% of patients. This is the class-wide mechanism behind the heart failure risk.
Heart failure exacerbation — TZDs are contraindicated in patients with NYHA Class III or IV heart failure. Fluid retention increases cardiac preload and can precipitate or worsen heart failure in susceptible patients. This is a class effect and applies to pioglitazone as well as rosiglitazone. Even Class I and II heart failure patients warrant close monitoring.
Bone fractures — PPAR-γ activation in bone affects osteoblast-osteoclast balance, reducing bone formation and increasing fracture risk. The effect is more pronounced in women. Long-term pioglitazone use is associated with a 2–3× increased fracture risk in postmenopausal women. Baseline and periodic bone density assessment is warranted.
Macular edema — Fluid accumulation in the macula of the retina has been reported with TZDs. Blurred vision or any visual disturbance should prompt ophthalmological evaluation.
Hepatotoxicity — The Rezulin legacy. Pioglitazone and rosiglitazone have significantly less hepatotoxic potential than troglitazone, and liver failure has not emerged as a clinical signal in large trials. However, liver function monitoring is recommended at initiation and periodically thereafter. Do not use in patients with active liver disease or ALT >2.5× the upper limit of normal.
Bladder cancer (pioglitazone-specific) — FDA warning added 2011. Risk appears greatest with cumulative exposure exceeding one year. Absolute risk increase is small but real. Contraindicated in patients with bladder cancer history.
TZDs are used far less than they were in their 2003–2007 peak. The combination of weight gain, edema, heart failure risk, fracture risk, and the Avandia scandal pushed prescribers toward newer classes with cleaner safety profiles. In the current landscape, SGLT2s and GLP-1s have taken the second-line position that TZDs once occupied.
Pioglitazone retains a niche: it is exceptionally cheap (generic pioglitazone 15mg costs under $10/month), it addresses insulin resistance directly (a mechanism SGLT2s and DPP-4s do not), and its NASH/MASH evidence is genuine in a disease with few pharmacological options. For patients with severe insulin resistance, NAFLD/NASH, and cost constraints who do not have heart failure or bladder cancer history, pioglitazone remains a reasonable choice.
Please understand that I’m not a doctor, and everything on this site reflects my own research and personal experience managing Type 2 diabetes. It is provided for informational and educational purposes only — not as medical advice, diagnosis, or treatment.
Before starting any supplement, changing your diet, or adjusting your medications, consult your physician or a qualified healthcare provider. Individual results vary. Never delay or disregard professional medical advice because of something you read here.
Product photos and brand names appear for illustrative purposes only. Denyabetic does not endorse, sponsor, or have any commercial relationship with any brand shown unless explicitly stated.
These statements have not been evaluated by the Food and Drug Administration. No product or protocol discussed on this site is intended to diagnose, treat, cure, or prevent any disease.