Individuals with schizophrenia are at increased risk of admission as soon as 10 days after missing a prescription for an antipsychotic, say US investigators (J Clin Psychiatry 2008;69:47‐53).

Their analysis of pharmacy claims for 1191 patients suggests that, for those who do not have a prescription dispensed, the risk of admission to a mental health care institution is 54 per cent higher within the first 10 days than for people who have medication to hand. Admission for a schizophrenia‐related event is increased by 77 per cent.

Overall, 79 per cent of patients in this group had at least one gap in treatment, with medication not available for 27 per cent of follow‐up days. A medication‐free interval of over 30 days was associated with a 50‐60 per cent increased risk of admission compared with patients who had their medication dispensed.

An analysis of the CAFE study (Comparison of Atypicals in First Episode) involving 400 patients found that 29 per cent discontinued treatment with an atypical antipsychotic (olanzapine, quetiapine or risperidone) against medical advice, 30 per cent completed one year of treatment and the remainder stopped treatment due to adverse events, lack of efficacy, administrative or other reasons (J Clin Psychiatry 2008;69:106‐13).

Among those who discontinued against medical advice, a poor treatment response and low medication adherence were significant independent predictors. In turn, poor adherence was associated with current substance abuse, depression and treatment failure –but also with higher cognitive performance at baseline and reaching remission. The authors suggest these factors should help to identify patients at risk so that preventative strategies can be targeted appropriately.

Adherence to antipsychotic medication can be improved through a patient‐centred strategy that identifies and overcomes barriers to adherence, a third study has shown (J Clin Psychiatry 2008;69:74‐80).

Six centres implemented medication management guidelines but in three the strategy was reinforced by a nurse co‐ordinator who encouraged clinicians to prescribe within the guidelines and interviewed patients to discover possible objections to medication, then developed specific strategies to tackle them.

After six months' follow‐up, adherence had increased from 46 to 61 per cent at sites using the guidelines alone and from 43 to 65 per cent at sites with nurse intervention. However, the improvement was substantially greater at two of the three enhanced sites (25 and 30 per cent increase) than at the third (16 per cent increase).

UK psychiatrists have identified another factor potentially associated with low adherence among people with schizophrenia –insight (Psychiatric Bull 2008;32:53‐6). They interviewed 52 patients with stable schizophrenia who were living in the community and taking an oral or a depot antipsychotic (excluding clozapine).

They described the overall level of insight as ‘fairly high’ but found that those using oral medication had significantly higher scores for insight and attitude to medication. The mean dose of depot medication (expressed as chlorpromazine equivalents) was over three times higher than for oral antipsychotics.

The authors conclude that decisions to prescribe depot medication were being made rationally but urge that interventions to improve insight should be offered and antipsychotic doses should be reviewed regularly.

Adding the anticonvulsant valproate semisodium to established treatment with an antipsychotic improves psychosis and functioning in older people with schizophrenia, a US study has shown (Int J Ger Psychiatry 2008;23:142‐7).

While there is concern in the UK about prescribing anticonvulsants for elderly patients, psychiatrists in Ohio say that up to two‐thirds of patients with schizophrenia and related disorders are treated with anticonvulsant medication in geropsychiatric settings. In their non‐blinded trial, 20 patients (mean age 61 years) were treated with a mean dose of 587mg modified‐release valproate semisodium (dose adjusted to achieve serum levels of 50‐100μg per ml); concurrent antipsychotic medication was unchanged.

Over 12 weeks, there were significant improvements in symptom scores for psychosis and depression; there was no significant change in cognitive function or physical wellbeing. Sedation was reported by six patients and five discontinued treatment (one due to an adverse event).

By coincidence, three trials of omega‐3 fatty acid supplementation have been published in close proximity. Researchers in the UK and the Middle East report a small double‐blind trial of eicosapentaenoic acid (EPA) and fluoxetine, separately and in combination, in the treatment of depression (Austr NZ J Psychiatry 2008;42:192‐8).

Sixty outpatients were randomised to EPA 1000mg daily or fluoxetine 20mg daily or both. After eight weeks, combined treatment reduced the Hamilton Depression Rating Scale (HDRS) score significantly more than either of the monotherapies, which were not different. Anxiety and decreased appetite were more common among patients taking fluoxetine alone than with the combination. The effect of blinding may have been reduced by a fishy taste reported by some participants taking EPA.

Monotherapy with fish oil (containing omega‐3 fatty acids EPA and docosahexaenoic acid (DHA), and omega‐6 fatty acids) is not effective in the treatment of perinatal depression, Australian researchers say (Austr NZ J Psychiatry 2008;42:199205). Twenty‐six women, between the third trimester of pregnancy to six months post‐partum, were randomised to fish oil 6g daily (with the taste masked by peppermint oil) or placebo. After six weeks, there were no differences in postnatal depression scores. The authors acknowledged that recruitment had been difficult, reducing the power of the study.

In the third trial, Swedish psychiatrists evaluated supplementation with DHA 1.7g daily and EPA 600mg daily in the treatment of Alzheimer's disease (Int J Geriatr Psychiatry 2008;23:161‐9). They randomised 204 patients treated with a cholinesterase inhibitor and with an MMSE score >15 to omega‐3 supplements or placebo for six months, then assigned all patients to supplementation for a further six months.

Fifteen per cent of patients did not complete the trial. Among the remainder, omega‐3 supplementation had little effect on neuropsychiatric symptoms or activities of daily scores. There was no difference in response between APOEϵ4 carriers and non‐carriers with the exception of reducing agitation in carriers and depression scores in non‐carriers.

Over one in four older patients with depression say they are not adherent to antidepressant therapy (Int J Ger Psychiatry 2008;23:129‐34).

This naturalistic prospective study of 241 inpatients and outpatients (mean age 69 years) found that 28 per cent said they did not take their medication as prescribed. Poor adherence at baseline was associated with worse depression score after 12 months. Other factors associated with a worse depression score included less social support, worse depression at baseline and limited daily functioning.

Antidepressants may be associated with hyperglycaemia or hypoglycaemia, an observational study from The Netherlands suggests (Eur J Clin Pharmacol 2008;doi 10.1007/ s00228‐007‐0441‐y).

The case control study compared the incidence of reports of hyperglycaemia or hypoglycaemia associated with antidepressants, benzodiazepines or antipsychotics. After adjustment for risk factors (including use of medication affecting blood glucose), they found that antidepressants were associated with a 52 per cent increased risk of hyperglycaemia and an 84 per cent increased risk of hypoglycaemia compared with benzodiazepines. Anti‐psychotics were associated with a more than six‐fold increase in risk of hyperglycaemia, but had no significant effect on hypoglycaemia risk.

SSRIs, tricyclics, and maprotiline, mianserin and mirtazapine were all associated with hyperglycaemia; this was more pronounced after one year of treatment. Antidepressants significantly associated with hypoglycaemia included SSRIs and most tricyclics (but not nortriptyline).

An analysis of the UK General Practice Research Database suggests that antihypertensives may have different neuroprotective effects (Neurology 2008; doi:10.1212/ 01.wnl.0000303818.38960.44).

The use of antihypertensive drugs by 3637 patients aged over 40 years and first diagnosed with Parkinson's disease between 1994 and 2005 was compared with that in matched controls. Compared with no use of the respective drug classes, neither ACE inhibitors nor angiotensin II‐receptor antagonists were associated with an increase or decrease in the risk of Parkinson's disease.

By contrast, current and past use of beta‐blockers was associated with an increased risk. However, there was no dose‐response effect and the risk was increased only in patients who had received fewer than 10 prescriptions. This suggests confounding by indication due to the use of beta‐blockers to treat early tremor.

Current use of calcium‐channel blockers was associated with a 22 per cent lower risk of Parkinson's disease (even in patients with an indication other than hypertension); again, there was no relationship with dose.

Websites offering brief cognitive behavioural therapy (CBT) or information about depression have been shown to have an immediate effect on symptom scores. Australian investigators have now reported longer term outcomes (Br J Psychiatry 2008;192:130‐4).

They randomised 525 people with elevated scores of depression to use of the CBT or depression information websites, or (as control) a lifestyle website. Mean depression scores steadily improved among controls but, immediately after use and at 12 months, depression scores were significantly lower among those with access to the CBT or depression information sites.

Confining the analysis to participants with clinically significant depression at baseline yielded similar trends though statistical significance compared with controls was not consistently maintained.

High levels of homocysteine are associated with cognitive impairment and an increased risk of Alzheimer's disease. Homocysteine levels can be reduced by folic acid so, psychiatrists in Perth have asked, could folic acid supplements have a role in the management of Alzheimer's disease (Int J Ger Psychiatry 2008;23:155‐60)?

They randomised 57 of 148 consecutive patients treated with a cholinesterase inhibitor for probable Alzheimer's disease (according to 2001 NICE guidance) to placebo or folic acid 1mg daily. Forty‐one patients completed six months' treatment (mean MMSE 23.5).

Basing their analysis on observed cases, ie not intention to treat, those treated with folic acid had an increase in combined Instrumental Activities of Daily Living and Social Behaviour scores compared with baseline (+1.50 points) whereas placebo was associated with a decreased score (‐2.29 points) after six months.

There was no difference between placebo and folic acid on change in MMSE score. The proportion classified as responders according NICE criteria (improvement or no decline in MMSE score) was 70 per cent with folic acid and 39 per cent with placebo. Folic acid did not significantly reduce homocysteine when baseline levels were taken into account. Nevertheless, the authors say their pilot study indicates that folic acid could improve the cost effectiveness of treatment with a cholinesterase inhibitor.

Combining olanzapine with carbamazepine in the treatment of mixed and manic bipolar episodes does not improve effectiveness but does increase adverse effects, say US psychiatrists (Br J Psychiatry 2008; 192:135‐43).

Their trial involving 118 patients found initially similar withdrawal rates from carbamazepine mono‐therapy and carbamazepine/ olanzapine due to adverse reactions (8‐9 per cent) during the six‐week double‐blind phase. During the following 20‐week open‐label phase with combined treatment, a further 12 per cent withdrew for the same reason.

There was no significant difference in efficacy measures between the two groups. However, combined treatment was associated with increased serum cholesterol and triglyceride levels and a higher incidence of clinically significant weight gain (25 vs 3 per cent with monotherapy). In those originally assigned to combined therapy (who took olanzapine for a total of 26 weeks), mean total weight gain was 5.5kg.

Naltrexone is an effective treatment for alcohol dependence for some people but not others. Why? asked US investigators (Arch Gen Psychiatry 2008;65:135‐44). Their pharmacogenetic analysis compared the response to treatment with naltrexone (with or without behavioural therapy) among recently abstinent volunteers categorised by a functional polymorphism of the μ‐opioid receptor allele OPRM1 Asp40.

They found that carriers of the Asp40 allele responded better to naltrexone alone (without behavioural therapy) than placebo whereas there was no difference in response among non‐carriers. There were no gene effects in those also treated with behavioural therapy. The authors suggest that genotyping may help treatment selection for some patients.

Placebo and nocebo responses are associated with opposite changes in dopamine and opioid neurotransmission, a positron emission tomography (PET) study suggests (Arch Gen Psychiatry 2008;65:220‐31).

In 20 healthy subjects exposed to a pain challenge after taking a placebo with expected analgesic activity, a placebo response was associated with increased regional dopaminergic and opioid activity and reduced continuous pain ratings. By contrast, dopamine deactivation and reduced opioid activity were associated with a nocebo response (hyperalgesia). The authors note that the brain areas associated with these changes are implicated in reward responses and motivated behaviour.

Another PET study has demonstrated that the rate of synthesis of serotonin in many regions of the brain during a migraine attack is similar to that among healthy controls (Neurology 2008;70:431‐9). However, serotonin synthesis is slightly lower overall than in controls between attacks (‐14 per cent) and much lower (around ‐30 per cent) in specific cortical areas of the brain. Treatment with sumatriptan significantly reduced the rate of serotonin synthesis in the majority of brain regions during an attack, confirming a central site of action and a role for increased synthesis of serotonin in causing migraine attacks.

Patients with frequently relapsing bipolar disorder (FRBD) had a significant delay in the time to an initial relapse compared with placebo (p=0.004) when risperidone long‐acting injection (RLAI) was combined with standard treatment; and the relative risk of relapse was 2.4 times higher with placebo. The relapse rates were 47.8 per cent with placebo and 22.2 per cent with RLAI.

The study was presented at the 14th Biennial Winter Workshop on Schizophrenia and Bipolar Disorders in Montreux, Switzerland. Janssen‐Cilag says this one‐year, phase III, trial involving 139 patients is the first placebo‐controlled study of the use of a long‐acting injection combined with standard maintenance treatment in FRBD. FRBD, defined as four or more manic or depressive episodes in the previous year that require a doctor's care, may affect 20 per cent of the 27 million people with bipolar disorder worldwide, according to Janssen‐Cilag. Copyright © 2008 Wiley Interface Ltd