Digest

More on NSAIDs and Alzheimer's disease Two large observational studies have confirmed that NSAID use is associated with a lower risk of Alzheimer's disease (AD) but, paradoxically, this property does not seem to be related to inhibition of amyloid‐beta (Aβ). The latest study was a pooled analysis of six prospective studies involving a total of 13 499 partici‐pants initially free of dementia (Neurology 2008;70:2291‐8) . After follow‐up of 5‐15 years, 820 new cases of AD were diagnosed. Overall, NSAIDs were associated with a 23 per cent reduction in the risk of AD compared with non‐use. However, NSAIDs that selectively inhibited Aβ deposition were not associated with a lower risk than non‐selective NSAIDs. These findings confirmed the results of a large case‐control study in the US (Neurology 2008;70:1672‐7) . In 49 349 cases of incident AD and 196 850 controls, NSAIDs were associated with a 24 per cent lower risk of developing AD after more than five years' use. For ibuprofen (which inhibits Aβ amyloid deposition), the relative risk reduction was 44 per cent after more than five years but risk reductions were not statistically significant for other NSAIDs. Again, there was no difference between inhibitors of Aβ deposition and non‐inhibitors. Galantamine: no improvement on antidepressants Augmentation with galantamine does not enhance the response to antidepressants, US investigators have found (Int J Geriatr Psychiatry 2008;23:625‐31) . They randomised 38 older patients (over 50 years) with major depression but not dementia to receive placebo or treatment with galantamine 16mg daily, after switching from their current anti‐depressant therapy to venlafaxine or citalopram. Over the next six months, depression scores (and cognitive function) improved equally in both groups. Post hoc analysis revealed that patients assigned to galantamine had lower depression scores and received a lower mean dose of venlafaxine (129 versus 182mg daily). A smaller proportion taking galantamine reported adverse effects than placebo (58 versus 94 per cent) but more discontinued treatment (63 versus 32 per cent) — though for no clear reason. Paroxetine during early pregnancy Following the publication of observational studies showing an association between paroxetine use during pregnancy and a 1.5‐2.0 per cent incidence of congenital cardio‐vascular defects, the US regulatory authorities advised women to avoid paroxetine during pregnancy. Prompted by research showing that many pregnancies are unplanned and that women who discontinue their treatment during pregnancy are more than twice as likely to relapse, an international group of investigators have pooled data from teratology centres and databases worldwide to provide a more authoritative estimate of the risk (Am J Psychiatry 2008;165: 749‐52) . They identified 1174 unpublished cases of first trimester exposure to paroxetine in teratology centres and matched them with controls who had contacted the centres to enquire about medication considered to be associated with lower risks during pregnancy, eg paracetamol. The rates of cardiovascular defects were 0.7 per cent in both case and control groups. A further 2061 cases were found in published database studies, in which the incidence was 1.5 per cent. Pooling all cases, the overall incidence was 1.2 per cent. The authors say this estimate is well within the background incidence of cardiovascular defects in the general population and they conclude that paroxetine does not appear to be associated with increased risk following use in early pregnancy. Depression one risk factor for prescription drug misuse In the US, misuse of prescription drugs by adolescents is exceeded only by use of alcohol, tobacco or marijuana. Analysis of a 2005 national survey on drug use and health, involving 18 678 adolescents, revealed that the prevalence of medication misuse in the previous year was 8.2 per cent. Of these, over one‐third reported symptoms consistent with prescription drug misuse, of which almost two‐thirds involved opioids (J Am Acad Child Adolesc Psychiatry 2008;47:doi 10.197/CHI.0b013e318172ef0d) . Multivariate regression analysis showed that factors associated with medication misuse were aged at least 15 years, poor academic performance, an episode of major depression within the past year (odds ratio 1.81) and treatment for mental health illness. Factors associated with DSM‐IV symptoms of substance use disorder were recent depression (odds ratio 1.53), use of cocaine, or frequent misuse. Novel adjunct for Parkinson's disease A pilot study of istradefylline, an adenosine A 2A ‐receptor antagonist, showed that it enhanced the response to levodopa without a substantial increase in dyskinesia in patients with Parkinson's disease (PD). Investigators in the US and Canada now report a larger dose‐ranging study (Neurology 2008;70:2233‐40) . A total of 395 patients with PD (mean duration eight to nine years, mean age 63‐65 years) were randomised to receive placebo or istradefylline 20mg or 60mg daily. All were taking levodopa (four doses per day) and 90 per cent were taking a second dopaminergic agent; they averaged five to six hours of ‘off’ time per day. After 12 weeks, the mean reduction in ‘off’ time while awake was 0.6 hours (10 per cent reduction) with placebo, 1.24 hours (22 per cent) for istradefylline 20mg daily and 1.37 hours (24 per cent) for the 60mg daily dose. The change occurred in all groups during the first two weeks and was sustained for the remainder of the trial. ‘On’ time without dyskinesia was increased by 0.25 and 0.46 hours above placebo by istradefylline 20mg and 60mg daily respectively. The commonest adverse events associated with istradefylline were dyskinesia, nausea, dizziness and hallucinations. More patients taking the 60mg dose discontinued treatment due to adverse events (10 per cent, compared with 4 per cent with 20mg dose and 6 per cent with placebo). Relaxation training for anxiety Relaxation training has ‘consistent and significant efficacy’ in reducing anxiety, say Italian psychologists (BMC Psychiatry 2008;8:41. doi:10.1186/1471‐244X‐8‐41). Their meta‐analysis of 27 randomised trials and observational studies of teaching relaxation techniques for people with anxiety covered the last 10 years of publications. The techniques included autogenic training, Jacobson's progressive relaxation, meditation and Benson's technique, applied relaxation and combinations of these. The endpoint was the effect size, calculated from anxiety scores before and after intervention. In 19 randomised controlled trials (half from the USA) involving a total of 1005 patients, a between‐group analysis showed that relaxation training had medium to high efficacy (effect size 0.51). However, there was significant heterogeneity so, the authors caution, this finding should be applied carefully. Meditation and (on the basis of a single trial) applied relaxation were the most effective techniques. Factors associated with greater efficacy included younger patients, longer duration of treatment and use of homework. A within‐group analysis of 25 studies involving 748 patients (44 per cent from the USA) again found a medium to high effect size (0.57) with significant heterogeneity. There were no differences between the main types of training but all were superior to combined approaches. Efficacy was lowest among patients with medical problems. Quetiapine for bipolar depression Quetiapine is, according to the BipOLar DEpRession (BOLDER) I and II studies, effective in the treatment of depression associated with bipolar disorder. A post‐hoc analysis of data pooled from these nearly identical trials has now been published, revealing a more detailed picture of its effects (J Clin Psychiatry 2008;69:769‐82) . The BOLDER trials were randomised placebo‐controlled trials of fixed doses of quetiapine (300mg or 600mg daily); this analysis included a total of 694 patients with bipolar I disorder and a current episode of depression from the two studies. The primary endpoint was the change in depression score (Montgomery‐Asberg Depression Rating Scale, MADRS). Discontinuation rates were high (36 and 45 per cent with quetiapine 300mg or 600mg daily and 40 per cent with placebo); the commonest reason for this was adverse effects with quetiapine and lack of effect with placebo. There was no difference between quetiapine doses in efficacy after eight weeks (mean reduction in MADRS score 19.4 and 19.6 with 300mg and 600mg respectively versus 12.6 with placebo) and treatment was equally effective in rapid)cycling and non‐rapid cycling disease. The pooled analysis showed that all core MADRS symptoms improved (though reduced appetite less significantly so), some after the first week of treatment. Response rates at week eight (≥50 per cent reduction in MADRS score ) were 61 per cent with each dose of quetiapine and 39 per cent with placebo; the median time to response was 22 days with quetiapine and 36 days with placebo. Remission (MADRS score ≥12) was achieved in 54‐55 per cent of patients treated with quetiapine and 32 per cent with placebo. Quetiapine significantly improved overall illness scores, quality of life and quality of sleep but not overall functioning. The most commonly reported adverse events included sedation, somnolence, dry mouth and dizziness; however, the presence of somnolence or sedation did not affect response rates. The incidence of extrapyramidal effects was 11‐12 per cent with quetiapine and 4 per cent with placebo. Quetiapine was also associated with dose‐dependent weight gain (mean 0.9 and 1.8kg with 300mg and 600mg respectively versus 0.2kg with placebo). Predicting interferon beta response in multiple sclerosis MRI activity and the presence of neutralising antibodies during the first six months reliably predict the long‐term response to treatment with interferon beta in patients with relapsing‐remitting multiple sclerosis (MS), Italian neurologists have confirmed (J Neurol Neurosurg Psychiatry 2008;79:646‐51) . Their prospective study found that 43 per cent of 147 patients had a relapse or confirmed disease progression over two years of treatment. The occurrence of lesions by MRI scan within the first six months predicted a good outcome in 80 per cent of patients but correctly identified a poor outcome in only 52 per cent of patients. The presence of neutralising antibodies within the first six months identified 66 per cent of responders and 71 per cent of non‐responders. Combining these endpoints, the authors found that 94 per cent of patients in whom neither marker was found had a good outcome though only half of those with both markers had a poor outcome. Trial design key to determining SSRI efficacy Many longer‐term clinical trials of antidepressants are of the discontinuation type: patients undergo a preliminary non‐blinded treatment phase after which responders are randomised to placebo, ie discontinuation, or treatment. While this design minimises exposure to placebo, its findings only apply to people who respond to treatment and excludes those with spontaneous improvement. The second design, more often used for shorter trials, randomises patients to placebo or treatment from the outset. Endpoints therefore reflect spontaneous improvement and are arguably more relevant to clinical practice. Canadian researchers have now reported a meta‐analysis of six trials of this design involving a total of 1299 patients treated with sertraline, citalopram or paroxetine (Can Med Assoc J 2008;178:1293‐301) . The trials lasted between six and eight months. The mean dropout rate was 48 per cent, with no difference between SSRIs and placebo. SSRIs were associated with better response rates (at least 50 per cent improvement in depression score) (odds ratio, OR) but this was statistically significant only for patients with no comorbidities (OR 2.13). Similarly, SSRIs were associated with remission (HAMD score 7 or less) only in patients with no comorbidities (OR 2.06). The authors say their study supports the current recommendation that treatment should continue for up to nine months after recovery from an acute episode but more studies of this design, with longer treatment duration, are needed. Genes and response to SSRIs Two studies have explored the significance of serotonin transporter genotype for the response to treatment with an SSRI ‐ with conflicting results. The first was an analysis of the response to citalopram among participants in the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial (PLoS ONE 2008;3: e1872. doi:10.1371/ journal.pone.0001872). Clinical response and tolerability were evaluated in 831 participants with 15 polymorphisms of five pharmacokinetic enzyme genotypes (CYP2D6, ABCB1, CYP2C19, CYP3A4, and CYP3A5). Seven polymorphisms — six occurring only in African Americans — were shown to be associated with poor tolerance of treatment or response. However, when these relationships were tested in a second group of 1046 participants, none were found to be statistically significant. The second study investigated the effect of serotonin transporter promoter (5‐HTTLPR) polymorphism on the response to treatment with paroxetine in 110 older patients (J Psychiatry Neurosci 2008;33:123‐30) . People with an s allele in 5‐HTTLPR express the serotonin transporter at only half the level of those who are homozygous for the l allele, and have lower maximal transport of serotonin. First, the authors found a relationship between the plasma concentration of paroxetine at two weeks and the clinical response. Second, they concluded that individuals with the s allele (low expression of the serotonin transporter) and with plasma levels of paroxetine well below 60ng per ml after two weeks' treatment would probably benefit from a dose increase or their adherence to their medicine addressed. Third, those homozygous for the l allele or with levels above 60ng per ml may not benefit from a dose increase. The authors suggest that, subject to further study, these tests could help decide the best therapeutic option at an early stage of treatment. Metabolic effects of atypicals in adolescents Young people treated with an atypical antipsychotic should be monitored carefully for adverse metabolic effects, warn Spanish investigators (J Clin Psychiatry 2008;June 3. e1‐e10. pii: ej07m03445). They followed up 66 children and adolescents (mean age 15 years) treated with olanzapine, risperidone or quetiapine. After six months, BMI was significantly increased in those taking olanzapine or risperidone; total cholesterol was significantly increased in those taking olanzapine and quetiapine; and systolic BP was significantly increased in those taking olanzapine. At baseline, 11 patients (17 per cent) were considered at risk due to these metabolic parameters; after six months, the number of patients increased to 25 (38 per cent) and the increase was significant only among those treated with olanzapine. College calls for end to mental health inequalities The Royal College of Psychiatrists College has drawn up its Fair Deal manifesto to highlight the problems that people with both mental health problems and learning disabilities encounter daily. The report highlights that: One‐in‐four older people living in the community have symptoms of depression severe enough to warrant help, but only half are diagnosed and treated. A third of substance misuse patients with mental health needs do not receive any interventions. In‐patients may remain in hospital for months after their need for hospitalisation has ended while they await transfer to local authority accommodation. Mental health research receives 6.5 per cent of total research funding, compared with 25 per cent for research into cancer and 15 per cent for neurological diseases.Skunk risk highlighted People who smoke skunk, the extra strong cannabis grown in hothouse conditions, are 18 times more likely to develop psychosis than those who use the milder forms, according to data from a new study presented at the Royal College of Psychiatrists' annual meeting in London in July. Psychiatrist Dr Marta Di Forti, as part of the GAP (Genetics and Psychosis) study led by Professor Robin Murray, from the Institute of Psychiatry, compared cannabis use between 112 patients with first‐episode psychosis and 75 healthy ‘controls’ screened for psychosis. People who had a psychotic episode were twice as likely to have used cannabis for longer, three times more likely to have used it every day and 18 times more likely to use skunk. Copyright © 2008 Wiley Interface Ltd