Digest

Abstract

AbstractDepressionAntidepressant RCTs meaningful?The Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial of sequenced anti‐depressant therapy recruited a broad population of patients representative of outpatients with depression seeking treatment. By contrast, most randomised controlled trials (RCTs) have more restrictive eligibility criteria. STAR*D investigators have now used their data to assess the impact of this aspect of trial design (Am J Psychiatry 2009;166:599–67).They divided 2855 STAR*D participants into a group meeting standard eligibility criteria based on HAM‐D score >19, lack of comorbidity and duration of current episode less than 24 months (22 per cent) and a second group who did not meet these criteria.Those meeting standard criteria were more likely to be seen in specialist centres, were younger, more educated, more likely to be white, wealthier, married and employed, with a shorter mean duration of illness, less substance abuse, fewer suicide attempts and fewer anxiety or atypical symptoms.The response rate to citalopram in the eligible group was 52 per cent compared with 39 per cent in the non‐eligible group; corresponding remission rates were 34 and 25 per cent (number needed to treat (NNT) 10). Participants meeting RCT criteria also experienced fewer moderate or severe adverse events and fewer admissions to hospital.The authors conclude that the usual RCT criteria select an unrepresentative population that favours better outcomes of treatment.Antidepressants increase diabetes riskLong‐term use of higher doses of antidepressants is associated with an increased risk of diabetes, an analysis of UK primary care data has shown (Am J Psychiatry 2009; 166:591–8).The risk of weight gain during treatment with an antidepressant prompted German psychiatrists to explore a possible link with diabetes. Using the UK General Practice Research Database, they identified 165 958 patients without diabetes who received at least one new prescription for an antidepressant between 1990 and 2005. Exposure was categorised as short (<12 months), medium (12–24 months) or long‐term (>24 months), and low‐dose (median dose or lower) or moderate/high‐dose (above median).A total of 2243 new diagnoses of diabetes were confirmed. Compared with non‐use of anti‐depressants within the previous two years and after adjustment for risk factors and other medication, long‐term use of moderate/high doses was associated with a significantly increased risk of diabetes (rate ratio 1.84, CI 95% 1.35–2.52). Lower doses or shorter use were not associated with increased risk and adjusting for depression severity did not alter the findings. For individual antidepressants, significant increases in risk were associated with amitriptyline, fluvoxamine, paroxetine and venlafaxine but numbers in these sub‐groups were low.More on SSRIs in pregnancySSRI use during pregnancy appears to increase the risk of preterm delivery, a relatively small US study suggests (Am J Psychiatry 2009;166:557–66).Neonatal outcomes were compared in 238 women: 131 did not have depression or use an SSRI; 71 with depression used an SSRI during some or all of their pregnancy; and 36 had depression not treated with an SSRI. The latter group consumed more alcohol but did not smoke more than the others.Neither depression nor SSRI use was associated with an increased risk of minor malformations, maternal weight gain, or changes in infant weight or length. After adjustment for age and ethnicity, continuous SSRI use was associated with a five‐ to six‐fold statistically significant increased risk of preterm delivery (<37 weeks); continuous depression was also associated with an increased risk of marginal statistical significance. There were no differences in delivery complications or neonatal development. Copyright © 2009 Wiley Interface Ltd