Self-help interventions for depression

Gellatly et al. (pp. 1217–1228) report findings from a meta-analysis and meta-regression of possible moderators of the effectiveness of self-help interventions for depression. From 34 studies, the authors found that a number of factors were associated with greater effectiveness, including aspects of study design (e.g. unclear allocation concealment), type of population (e.g. non-clinic samples) and type of intervention (e.g. guided self-help). In their meta-regression, only guided self-help remained associated with greater effectiveness.

Depression

This issue contains six papers on aspects of depression. In the first, Elgamal et al. (pp. 1229–1238) report findings from a study of the potential for a 10-week programme of cognitive remediation therapy (CRT) to improve cognitive function in unipolar depression. In a comparison of 12 subjects with major depressive disorder (MDD) receiving CRT, 12 matched subjects with MDD and 22 healthy controls, the authors found that the group receiving CRT showed greater improvements across a range of cognitive domains, including verbal learning and memory. These effects were independent of levels of depression.

Weich et al. (pp. 1239–1248) examined attitudes towards depression and its treatment in a sample of 866 subjects with an episode of MDD in the preceding 12 months. Subjects were drawn from a larger study of GP attenders and attitudes were assessed using a 19-item self-report questionnaire. A factor analysis revealed a three-factor solution: (1) depression as a disabling, permanent state; (2) depression as a medical condition responsive to support; and (3) antidepressants as addictive and ineffective. Those who complied with antidepressant medication and those who disclosed their problems to friends had lower scores on factors 1 and 3, but higher scores on factor 2.

Lau et al. (pp. 1249–1259) examined the theory that inhibitory dysfunction influences the degree to which activated self-schemas result in depressive cognition. In a sample of 43 subjects with MDD, 32 subjects with an anxiety disorder, and 32 controls, the authors used the Prose Distraction Task (PDT) and the Stop-Signal Task (SST), with modifications to present emotionally valenced semantic stimuli. They found no differences between subjects on the SST. However, on PDT, MDD subjects performed worse than others, particularly when negative stimuli were presented. One interpretation of this disruption in prose reading is that it indicates a cognitive inhibitory deficit in subjects with MDD.

Kaptein et al. (pp. 1261–1271) investigated whether recently proposed random mood models of depression, based on heterogeneous population-based cohorts, hold for more homogenous groups defined by known determinants of MDD, and whether the model's parameters (susceptibility – Z; relaxation time – T) could be used to differentiate sub-cohorts on the basis of course of illness. Data from the Netherlands Mental Health Survey and Incidence Study was used. The authors found that the Z-T classification was more sensitive at predicting survival curves than prior statistical methods and, particularly for sex (men v. women) and co-morbid anxiety, revealed differences that have previously been unnoticed.

Cooper et al. (pp. 1273–1280) examined the question of whether postnatal depression (PND) and non-postnatal depression (NPND) are distinct by conducting a series of within- and between-group comparisons of clinical presentations in subjects with PND (n=50) and subjects with NPND (n=132). Few differences were found between postnatal and non-postnatal episodes of depression and these, the authors propose, are likely to be explicable within the context of having a new baby (e.g. reduced symptoms of poor appetite and early morning waking in PND episodes).

Chan et al. (pp. 1281–1291) investigated aspects of emotional processing in 33 never-depressed subjects who scored high on neuroticism (N) and 32 who scored low on N. They found that high N subjects showed increased processing of negative or decreased processing of positive information in a series of tests assessing emotional categorization and memory, facial expression recognition and emotion-potentiated startle. The authors conclude that some negative processing biases precede rather than follow the onset of depression and, consequently, may mediate the vulnerability of high N subjects to depression.

Physical health/disability

Four further papers examine links between physical health/disability and mental health. Raznahan et al. (pp. 1293–1304) investigated the neurobiological correlates of intellectual disability in 48 subjects aged between 2 and 25 years with tuberous sclerosis. The factors found to be associated with intellectual disability were: a history of infantile spasm, total cortical tuber (CT) count, and the presence of CTs in the frontal and temporal lobes. Only the latter two remained independently associated with intellectual disability in subsequent multivariable analyses.

Rasul et al. (pp. 1305–1313), in a prospective study of 1864 middle-aged men, examined whether the association between psychological distress (PD) and myocardial infarction (MI) can be explained by the effect of pre-existing illness on both PD and MI. They found that PD was independently associated with risk of MI. However, when the analyses were repeated excluding those with baseline coronary heart disease, no association was observed between PD and MI. The authors conclude that PD may be a moderator for risk of MI in those with pre-existing illness.

Verdoux et al. (pp. 1315–1322) investigated the relationship between prenatal exposure to diethylstilboestrol (DES) and risk of a serious psychiatric outcome (suicide, hospital admission, other psychiatric or psychological consultation) in offspring in a cohort study of 1352 mothers (giving a total of 1680 exposed children and 1447 unexposed siblings), with a mean follow-up period of 36 years. After adjusting for a range of potential confounders, including education and parental history of psychiatric hospitalization, the authors found no association between prenatal exposure to DES and risk of a serious psychiatric outcome in offspring.

Nomura et al. (pp. 1323–1334) examined the relationship between exposure to perinatal problems (e.g. low birth weight, preterm birth, low Apgar scores) and risk of co-morbid psychiatric (depression, suicidal ideation) and physical (hypertension, asthma) problems in a sample of mothers and their children (n=1525) followed from birth to adulthood. The authors found that perinatal problems, particularly preterm low birth weight, were associated with risk of depression, suicidal ideation and hypertension, and co-morbid depression and hypertension, after adjusting for potential confounders.

Other topics

This issue concludes with three papers examining a variety of topics. Van der Elst et al. (pp. 1335–1344) present findings from a study designed to establish normative data for the Paper & Pencil Memory Scanning Test (P&P MST), a test developed from the longer and less user-friendly MST. The P&P MST was administered to a large sample (n=1839) of healthy and cognitively intact adults aged 24–81 years. The authors found that age and level of education affected all components of information processing in working memory, and present a user-friendly table of normative scores for the P&P MST by age and education for use in clinical settings.

Lynskey & Agrawal (pp. 1345–1355) used data collected for the National Epidemiological Survey on Alcohol and Related Conditions (n=43093) to examine a number of issues related to the psychometric properties of DSM-IV criteria for drug abuse and dependence. Results of a factor analysis supported a uni-dimensional construct for abuse/dependence for all eight drug types considered. The authors further found that, for each drug type, between 29% and 51% of those who had at least one symptom did not meet criteria for abuse or dependence.

In the final paper, Morley et al. (pp. 1357–1367) used multivariate biometrical modelling to investigate the genetic and environmental contributions to smoking age-at-onset, cigarette use and smoking persistence in a sample of twins and their siblings (n=14472). The authors found that all these traits were moderately heritable in males and females, but that the extent to which additive genetic traits was shared across traits varied by gender. Twin specific environmental factors accounted for a substantial proportion of the variance in smoking age-at-onset. Unique environmental factors had a moderate influence on smoking age-at-onset and a stronger influence on the other traits.