This article critically reviews the design and methodology of studies aimed at evaluating relapse and recurrence prevention in major depression. A literature search in MEDLINE was performed with the medical subject headings ‘depression’, ‘recurrence’, ‘relapse’, ‘prevention’ and ‘study’. This search covered the period from January 1990 to July 1999. Only long-term placebo-controlled studies that included patients with non-chronic major depression were selected. Two types of design could be distinguished: randomised withdrawal studies in responders/remitters (N = 11) and extension studies in responders without re-randomisation (N = 3).

Randomised withdrawal studies are suitable for demonstrating long-term efficacy for the duration of the study period. However, this design does not permit a clear differentiation between relapse or recurrence and, therefore, is not suitable to demonstrate unequivocally relapse prevention or recurrence prevention. Extension studies in short-term responders without randomisation are not even suitable to demonstrate long-term efficacy.

A novel design is proposed that overcomes the weaknesses of designs employed thus far. In essence, this design calls for a long-term randomised placebo-controlled study in patients who are free of medication for a substantial period of time and who fulfil the criteria of major depression (recurrent) in sustained remission (e.g., HDRS < 7) as a possible option.

Studies examining OC phenomena in schizophrenic and schizoaffective disorders have shown a prevalence of such phenomena in 1 to 60% of schizophrenic or schizoaffective patients. In this prospective study, about 10% of 150 male patients suffering from acute psychotic disorders (fulfilling DSM-IV criteria for Schizophrenia or Schizoaffective Disorder) were found to have OC symptoms. These symptoms showed no correlation to the type and severity of psychosis. As only 19% of the patients with obsessions and compulsions during acute psychosis showed an obsessive-compulsive personality disorder prior to their psychotic episodes, it may be concluded that there is no clear linkage between intrapsychotic OC phenomena and premorbid anancastic personality traits.

The aim of this study was to ascertain whether the symptom profile distinguishes between schizophrenic patients born in the winter and early spring and those born in other seasons. The sample consisted of 204 patients with a DSM-III-R diagnosis of schizophrenia who had been hospitalized for acute psychotic decompensation. Symptom ratings were based on the Positive and Negative Syndrome Scale (PANSS). The use of demographic and anamnestic data as dependent variables did not detect any season-of-birth effect. In contrast, clear gender-specific differences emerged from the comparison focusing on symptom dimensions and clinical subtype. Female patients born in the winter and early spring had higher scores on the PANSS negative scale and anergia factor whereas male patients born in other seasons had higher scores on the PANSS anergia factor. In addition, we found a gender-specific association between season of birth and clinical subtype. Most paranoid female patients were born in the non-winter months whereas, among men, a slightly higher percentage of paranoid patients were born in winter months. These results suggest that gender plays a role in modulating the effect of the season of birth on symptoms of schizophrenia.

Since reports have underscored that panic attacks (PA) may be an identifiable state occurring in schizophrenia, we studied the symptomatology of PA in a group of schizophrenic patients. Of 40 patients (21 males and 19 females) attending a clinic for maintenance therapy of schizophrenia, 19 (36.8%) had a lifetime history of PA. Seven among those 19 patients (36.8%) had or had had spontaneous panic attacks, not related to phobic fears or delusional fears, and for the 12 remaining patients, the PA were related to paranoid ideas. Moreover, the paranoid subtype of schizophrenia tends to be more often associated with a history of panic attack than other subtypes of schizophrenia (52.6% vs 23.8%; χ2 = 3.5, P = .06). It seems that there are at least two types of PA in schizophrenic patients. The first one could be independent from the psychotic feature, with no psychopathological link. The second kind of PA could be directly related to a schizophrenic disorder, and found in patients with the paranoid subtype.

There is widespread evidence that schizophrenic symptomatology is best represented by three syndromes (positive, negative, disorganized). Both the disorganized and negative syndrome have been found to correlate with several neurocognitive dysfunctions. However, previous studies investigated samples predominantly treated with typical neuroleptics, which frequently induce parkinsonian symptoms that are hard to disentangle from primary negative symptoms and may have inflated correlations with neurocognition. A newly developed psychopathological instrument called the Positive and Negative and Disorganized Symptoms Scale (PANADSS) was evaluated in 60 schizophrenic patients. Forty-seven participants treated with atypical neuroleptics performed several neurocognitive tasks.

A three-factor solution of schizophrenic symptomatology emerged. Negative symptomatology was associated with diminished creative verbal fluency and digit span backward, whereas disorganization was significantly correlated with impaired Stroop, WCST and Trail-Making Test B performance.

Data suggest that disorganization is associated with tasks that demand executive functioning. Previous findings reporting correlations between negative symptomatology and neurocognition may have been confounded by the adverse consequences of typical neuroleptics.