Antipsychotics Diabetes Risk Is Highest Right After Treatment Starts

The most dangerous stretch of antipsychotic treatment for diabetes risk is also the earliest one. People starting second-generation antipsychotics for schizophrenia or bipolar disorder see the biggest changes in blood sugar and weight within the first 90 days, according to new research published in The Lancet Psychiatry in mid-October. The findings sharpen the picture of how quickly antipsychotics diabetes risk builds after the first prescription, and the researchers say the early window is exactly when doctors should be watching both blood sugar and the scale.

What the researchers found

A team led by University College London analyzed anonymous NHS GP records from nearly 97,000 adults in England, aged 18 to 59, who had severe mental illness and began taking one of four second-generation antipsychotics, olanzapine, quetiapine, risperidone or aripiprazole, between 2000 and 2023. The group tracked weight and blood test results from two years before treatment started to two years after, comparing the rate of change across the two windows. That design let the team pin down when the antipsychotics diabetes risk signal accelerates most, not just whether it appears.

That acceleration turned out to be front-loaded. During the early treatment window, HbA1c, the standard measure of average blood sugar, rose about 14 times faster than it had before treatment, while body weight rose about nine times faster. The rate of increase then settled back toward pre-treatment levels after the initial period. The pattern held across the whole group, and the antipsychotics diabetes risk signal showed up in both blood sugar and weight rather than in one alone.

The two measures did not always move together, which the authors flag as a practical problem. Some people had blood sugar changes without much weight change, and the reverse also happened. That means tracking only weight, the most common routine check, can miss metabolic shifts that blood tests would catch. The finding makes the case for paired monitoring from the start, reported by HealthDay.

The four drugs did not behave identically. Olanzapine showed the largest early increases, with blood sugar rising about 50 times faster and weight about 10 times faster in the early window than before treatment. The other three drugs showed smaller but still meaningful accelerations. The differences matter because the choice of medication is often weighed against side effects, and the data give doctors a clearer picture of antipsychotics diabetes risk per drug, showing which measure each one tends to move fastest, as detailed by MedicalXpress.

Two groups faced steeper odds. People who already had prediabetes when they started treatment developed type 2 diabetes at a much higher rate, with about 19 percent receiving a diabetes diagnosis within two years compared with 3 percent of those with normal blood sugar at the start. Women also saw larger metabolic changes than men, a difference the researchers say deserves closer attention in follow-up work.

What doctors should do now

The practical message is about timing. First author Dr. Juan Carlos Bazo-Alvarez, a senior research fellow at UCL, said through a news release that antipsychotic medicines are valuable treatments that help many people with serious mental illnesses, and that the new data point to the first months as the key stretch when weight gain and diabetes-linked changes begin. His recommendation is straightforward: clinicians should record both weight and blood sugar before treatment starts and repeat both checks during the early months, so the antipsychotics diabetes risk signal is caught before it turns into a diagnosis.

That guidance lands in a population that already carries higher cardiometabolic risk. Schizophrenia affects roughly 23 million people worldwide, while around 37 million live with bipolar disorder, and people with severe mental illness develop diabetes and heart disease at higher rates than the general population. Catching the antipsychotics diabetes risk spike early could prevent diagnoses that, once established, add another layer of illness onto people already managing a serious condition.

The study has the usual limits of records-based research. The data come from one country's health records and the design is observational, so it can show when changes happened and how fast they moved, but it cannot by itself prove the drugs caused every shift. Still, the sheer size of the cohort and the consistency of the timing across subgroups give the early-window finding real weight. For patients, the takeaway is simple: if an antipsychotic prescription is new, ask about a baseline blood test and a follow-up plan, and treat the scale as only half of the antipsychotics diabetes risk picture.