Weight change over two years in people prescribed olanzapine, quetiapine and risperidone in UK primary care: Cohort study in THIN, a UK primary care database

Abstract

Olanzapine is more commonly prescribed to people with lower weight. However, after accounting for baseline weight, age, sex and diagnosis, olanzapine is still associated with greater weight gain over two years than risperidone or quetiapine. Baseline weight does not ameliorate the risks of weight gain associated with antipsychotic medication. Weight gain should be assertively discussed and managed for people prescribed antipsychotics, especially olanzapine.

Effects of long-term antipsychotics treatment on body weight: A population-based cohort study

“…In the long-term, contrary to one previous finding (Haddad, 2005), we found that weight gain did not stabilize during four years of follow up. However, our finding of long-term effect of weight gain is consistent with previous studies by Bushe et al (2012) and Osborn et al (2018), but we are able to quantify the effect more accurately. Previous research has suggested women’s weight is more affected by AP exposure (Seeman, 2008); however, we found that only olanzapine (in the long-term) and quetiapine (in the long and short-term) induced more weight gain in women.…”

“…A major advantage of our approach is that it includes pre-treatment weight change information, so patients act as their own controls in the analysis and any additional weight change after baseline is attributable to the AP treatment. The approach utilizes all individual weight records at their time of measurement, therefore avoiding the loss of information seen in previous studies which categorize outcomes or use period means or incidence rates as summary measures (Bak et al, 2014; Osborn et al, 2018). Our longitudinal model-based approach also accounts for missing weight records – assuming weight recording is missing at random within strata, conditional on observed weight measurements (Haneuse et al, 2016) – while incorporating informative pre-baseline weight data.…”

“…Our longitudinal model-based approach also accounts for missing weight records – assuming weight recording is missing at random within strata, conditional on observed weight measurements (Haneuse et al, 2016) – while incorporating informative pre-baseline weight data. From this method we expect unbiased estimates if data were missing at random (Molenberghs et al, 2014), a property that is not ensured by complete case analyses applied elsewhere (Bushe et al, 2012; Osborn et al, 2018). Following standard recommendations (Bernal et al, 2017), we guaranteed good statistical power by having equal periods of observation before and after baseline, and a large sample size.…”

“…However, we know that drugs prescribed for ameliorating weight gain would only reduce the estimate of the real weight gain of the target population, thus we are not overestimating the weight gain effect. We did not assess weight gain associated with other antipsychotic medications as there were not enough data on them, but the three drugs included in this study are the most commonly prescribed antipsychotic medications in the UK (Marston et al, 2014) and have previously been associated with weight gain (Osborn et al, 2018). The weight gain trajectories we described are averages, thus they should be interpreted as typical patient trajectories.…”

“…Second-generation antipsychotics (AP) are a known cause of weight gain (Bak et al, 2014; Osborn et al, 2018). Some evidence suggests that women gain more weight than men during AP treatment (Seeman, 2008).…”