Weight and Blood Sugar Rose Fastest in the First 90 Days After Starting Antipsychotics, Study Finds

This is an observational study, not a trial. The researchers used electronic health records from England's Clinical Practice Research Datalink Aurum database, covering January 2000 to June 2023. They included 96,849 people aged 18 to 59 with severe mental illness. All had at least a second prescription for an oral second-generation antipsychotic within 90 days of their first. People with dementia or type 2 diabetes were excluded. So were those prescribed antipsychotics before age 18.

The team used an interrupted time-series design. It compares each person's trend before treatment with their trend afterward. The study measured HbA1c, a blood test that reflects average blood sugar over recent weeks to months, and body weight. Before treatment, HbA1c rose by 0.01 mmol/mol per week and weight by 0.02 kg per week. In the first 90 days after starting, HbA1c rose by 0.12 mmol/mol per week and weight by 0.21 kg per week. From day 91 to two years, HbA1c returned to its pre-treatment rate of 0.01 mmol/mol per week. Weight continued to rise, but more slowly, at 0.03 kg per week.

The drugs differed. Increases were largest for olanzapine: HbA1c rose 0.18 mmol/mol per week and weight 0.28 kg per week. Quetiapine was linked to weight gain of 0.19 kg per week but little change in HbA1c (0.02 mmol/mol per week, with a confidence interval that included zero). Rises were generally higher in women, at 0.14 mmol/mol and 0.22 kg per week. Weight and blood sugar did not always move together.

Starting blood sugar mattered too. Among 918 people with prediabetes at baseline, 126 (13.7%) developed type 2 diabetes. The estimated two-year cumulative incidence was 19.0%. Among 9,171 people with normal baseline HbA1c, 152 (1.7%) did, an estimated 3.0% over two years.

There are limits. Only 24.9% of participants had at least one recorded HbA1c measurement, and 76.6% had at least one weight measurement. Because the data come from routine care, the study shows association and timing. It cannot prove that the drugs alone caused each change.

The authors conclude that the imperfect match between HbA1c and weight supports checking both at baseline and at 12 weeks, followed by monitoring based on each person's risk. In coverage by Medical Xpress, study author Magnus Boman of Karolinska Institutet said the greatest changes occur early after treatment begins, so both measures need watching from the start. The work was funded by the UK's National Institute for Health and Care Research and other UK bodies.

Whether monitoring schedules like the one the authors propose are adopted, and whether they change outcomes.

Alive. Awake. Aware.