Do early changes in the HAM-D-17 anxiety/somatization factor items affect treatment outcome among depressed outpatients? Comparison of two controlled trials of St John’s Wort (Hypericum Perforatum) versus an SSRI

In the pooled SSRI arms (sertraline and fluoxetine combined, n=145), early improvement in the HAM-D-17 anxiety-psychic item from baseline to week 1 was significantly associated with remission at study

Effect: improvement; p<0.001, r2=0.184

In the pooled hypericum arms (n=147), early improvement in somatic-gastrointestinal (p<0.001, r2=0.262) and hypochondriasis (p=0.004, r2=0.218) items from baseline to week 1 significantly predicted re

Effect: improvement; somatic-GI: p<0.001, r2=0.262; hypochondriasis: p=0.004, r2=0.218

In the pooled placebo arms (n=153), early improvement in anxiety-psychic (p=0.003, r2=0.110) and somatic symptoms-general (p=0.001, r2=0.118) items from baseline to week 1 significantly predicted remi

Effect: improvement; anxiety-psychic: p=0.003, r2=0.110; somatic-general: p=0.001, r2=0.118

When all 445 participants were collapsed regardless of treatment, early improvement in anxiety-psychic (p<0.001, r2=0.143), somatic-gastrointestinal (p<0.001, r2=0.106), and somatic-general (p<0.001,

Effect: improvement; anxiety-psychic: p<0.001, r2=0.143; somatic-GI: p<0.001, r2=0.106; somatic-general: p<0.001, r2=0.104

In the MGH study (N=125), no HAM-D-17 anxiety/somatization item reached significance after Bonferroni correction in any treatment arm (fluoxetine, hypericum, or placebo), despite nominal significance

Effect: null; best: anxiety-psychic in fluoxetine arm p=0.026; somatic-GI in hypericum arm p=0.042