Last general practitioner consultation during the final month of life: a national medical record review of suicides in Sweden.

OBJECTIVES: Individuals who die by suicide often consult their general practitioners (GPs) in their final weeks of life. The aim of this study was to gain a deeper knowledge of the clinical characteristics and GP assessments and treatments among individuals who consulted their GPs during the month preceding suicide. Further, we compared these features in those with and without contact with psychiatric services (PC and NPC, respectively) during the two years that preceded the suicide. DESIGN: A nationwide retrospective explorative study investigating medical records. SETTING: Primary care in Sweden. PARTICIPANTS: Individuals who died by suicide in Sweden in 2015 with a GP visit within 30 days of death. RESULTS: The study cohort corresponds to one fifth (n = 238) of all suicides that occurred in Sweden in 2015 (n = 1179), representing all those with available primary care records showing contact with a GP during the final 30 days of life (NPC: n = 125; PC: n = 113). The mean age was 58 years ± 19. Patients in the NPC group were older (NPC: 63 years ± 19 vs. PC: 53 years ± 18, p < 0.0001) and presented psychiatric symptoms less often (NPC: 50% vs. PC: 67%, p < 0.006). Somatic symptoms were as common as psychiatric symptoms for the whole sample, being present in more than half of individuals. Suicide risk was noted in only 6% of all individuals. Referral to psychiatric services occurred in 14%, less commonly for the NPC group (NPC: 6% vs. PC: 22%, p < 0.001). Cardiovascular or respiratory symptoms were noted in 19%, more often in the NPC group (NPC: 30% vs. PC: 6%, p < 0.001), as were diagnoses involving the circulatory system (all 10%, NPC:14% vs. PC: 5%, p < 0.020). CONCLUSION: A high level of somatic symptoms was observed in primary care patients both with and without psychiatric contact, and this might have influenced GPs' management decisions. Psychiatric symptoms were noted in two thirds of those with psychiatric contact but only half of those without. While GPs noted psychiatric symptoms in over half of all individuals included in the study, they seldom noted suicide risk. These findings highlight the need for increased attention to psychiatric symptoms and suicide risk assessment, particularly among middle-aged and older individuals presenting with somatic symptoms. STRENGTHS AND LIMITATIONS OF THIS STUDY: The National Cause of Death Register has excellent coverage of suicide deaths and access to medical records was very good. The medical record review provided detailed information regarding primary care utilization before death by suicide. Because of the lack of statistical power, due to the limited number of persons with GP contact during the last month of life, we chose not to correct for multiple comparisons. Our study approach did not capture the reasons behind GPs' documentation of elevated suicide risk. No systematic inter-rater reliability (IRR) testing was made, however, reviewers received training and continuous support from the research group.

Anxiety symptoms preceding suicide: A Swedish nationwide record review.

BACKGROUND: The literature on the relationship between anxiety and suicidal behaviors is limited and findings are mixed. This study sought to determine whether physicians noted anxiety symptoms and suicidality in their patients in the weeks and months before suicide. METHODS: Data were derived from a nationwide medical record review of confirmed suicides in Sweden in 2015. Individuals with at least one documented physician consultation in any health care setting during 12 months before suicide (N = 956) were included. Clinical characteristics were compared between decedents with and without a notation of anxiety symptoms. Odds ratios were calculated to estimate associations between anxiety symptoms and suicidality in relation to suicide proximity. RESULTS: Anxiety symptoms were noted in half of individuals 1 week before suicide. Patients with anxiety were characterized by high rates of depressive symptoms, ongoing substance use issues, sleeping difficulties, and fatigue. After adjustment for mood disorders, the odds of having a notation of elevated suicide risk 1 week before death were doubled in persons with anxiety symptoms. Associations were similar across time periods (12 months - 1 week). Two-thirds had been prescribed antidepressants at time of death. LIMITATIONS: Data were based on physicians' notations which likely resulted in underreporting of anxiety depending on medical specialty. Records were not available for all decedents. CONCLUSIONS: Anxiety symptoms were common in the final week before suicide and were accompanied by increases in documented elevated suicide risk. Our findings can inform psychiatrists, non-psychiatric specialists, and GPs who meet and assess persons with anxiety symptoms.

The Meaning of Mental Imagery in Acute Suicidal Episodes: A Qualitative Exploration of Lived Experiences.

Clinical assessment of suicidal ideation focuses on cognitions in the form of verbal thoughts. However, cognitions also take the shape of mental imagery. The aim of this qualitative study was to explore the meaning of mental imagery in acute suicidal episodes (ASEs). Eight persons with severe previous ASEs participated in repeated in-depth interviews and in the semi-structured Suicidal Cognitions Interview. Textual data from both sources underwent content analysis. All participants experienced suicide-related imagery during ASEs. Analysis resulted in two themes. (1) Suicide-approaching imagery: intrusive looming images that contributed to loss of control, flashforwards that clarified the suicidal solution, or desirable but unattainable images. (2) Suicide preventive imagery: death-alienating, life-affirming, or potentially helpful images. The meaning of mental imagery in ASEs is suggested to be understood in relation to the context of the individual ASE. A narrative approach is encouraged, as is an increased clinical focus on mental imagery in general.

Utilization of psychiatric services prior to suicide- a retrospective comparison of users with and without previous suicide attempts.

INTRODUCTION: The aim was to investigate psychiatric health care utilization two years before death by suicide among individuals with previous suicide attempts (PSA) compared with those without (NSA). METHOD: A retrospective population-based cohort study was conducted including 484 individuals who died by suicide in Sweden in 2015 and were in contact with psychiatric services within the two years preceding death, identified through the Cause of Death register. Data on psychiatric health care two years before death, including suicide attempts according to notes in the medical record was used. Associations between having at least one PSA vs. NSA and health care utilization were estimated as odds ratios (OR) with 95% confidence intervals (CI) by logistic regression analyses. RESULTS: Of the 484 individuals included, 51% had PSA. Those with PSA were more likely than NSA to have received a psychiatric diagnosis [OR 1.96 (CI 95% 1.17-3.30)], to have ongoing psychotropic medication [OR 1.96 (CI 95% 1.15-3.36)] and to have been absent from appointments during the last three months [1.97 (1.25-3.13)]. In addition, elevated suicide risk was more often noted in the psychiatric case records of those with a PSA than those without [OR 2.17 (CI 95% 1.24-3.79)]. CONCLUSION: The results underline the importance of improved suicide risk assessment as well as thorough diagnostic assessment and when indicated, psychiatric treatment as suicide preventive interventions regardless of PSA. Furthermore, the larger proportion of absence from appointments in individuals with PSA may indicate a need of improved alliance between psychiatric care providers and individuals with PSA.HIGHLIGHTSBeing assessed with elevated suicide risk was more common among those with previous attempt/s (PSA).One-fifth of all with no previous attempt (NSA) had no psychiatric diagnosis, compared to one in ten in those with PSA.Receiving psychotropic medication was more common among those with PSA.

Health care utilisation two years prior to suicide in Sweden: a retrospective explorative study based on medical records.

OBJECTIVE: Previous literature has suggested that identifying putative differences in health care seeking patterns before death by suicide depending on age and gender may facilitate more targeted suicide preventive approaches. The aim of this study is to map health care utilisation among individuals in the two years prior to suicide in Sweden in 2015 and to examine possible age and gender differences. METHODS: Design: A retrospective explorative study with a medical record review covering the two years preceding suicide. SETTING: All health care units located in 20 of Sweden's 21 regions. PARTICIPANTS: All individuals residing in participating regions who died by suicide during 2015 (n = 949). RESULTS: Almost 74% were in contact with a health care provider during the 3 months prior to suicide, and 60% within 4 weeks. Overall health care utilisation during the last month of life did not differ between age groups. However, a higher proportion of younger individuals (< 65 years) were in contact with psychiatric services, and a higher proportion of older individuals (≥ 65 years) were in contact with primary and specialised somatic health care. The proportion of women with any type of health care contact during the observation period was larger than the corresponding proportion of men, although no gender difference was found among primary and specialised somatic health care users within four weeks and three months respectively prior to suicide. CONCLUSION: Care utilisation before suicide varied by gender and age. Female suicide decedents seem to utilise health care to a larger extent than male decedents in the two years preceding death, except for the non-psychiatric services in closer proximity to death. Older adults seem to predominantly use non-psychiatric services, while younger individuals seek psychiatric services to a larger extent.

Application of an accident approach to the study of acute suicidal episodes through repeated in-depth interviews.

We modified an accident analysis model for the study of acute suicidal episodes (ASEs). Our aim was to use this model (SCREAM, Suicidal Cognitions' Reliability and Error Analysis Method) as a lens to understand the worst-ever ASEs of nine patients who took part in repeated in-depth interviews. Guided by the theory of SCREAM including four predetermined categories, nine themes were identified. ASEs were triggered by interactions with the interpersonal and physical environment and spiraled into a state of lost control. Timing and the availability of promoters and barriers in the environment were salient features. Findings may aid person-centered safety planning.

Rejection sensitivity and pain in bipolar versus unipolar depression.

OBJECTIVES: Recent neuroimaging studies support the contention that depression, pain distress, and rejection distress share the same neurobiological circuits. In two recently published studies we confirmed the hypothesis that the perception of increased pain during both treatment-refractory depression (predominantly unipolar) and difficult-to-treat bipolar depression was related to increased state rejection sensitivity (i.e., rejection sensitivity when depressed). In the present study, we aimed to compare the correlates of pain and rejection sensitivity in individuals with bipolar versus unipolar depression and test the hypothesis that bipolar disorder may be distinguished from unipolar depression both by an increased perception of pain and heightened rejection sensitivity during depression. METHODS: We analyzed data from 113 bipolar and 146 unipolar depressed patients presenting to the Black Dog Institute, Sydney, Australia. The patients all met DSM-IV criteria for bipolar disorder or unipolar depression (major depressive disorder). RESULTS: Bipolar disorder predicted a major increase in state rejection sensitivity when depressed (p = 0.001), whereas trait rejection sensitivity (i.e., a long-standing pattern of rejection sensitivity) was not predicted by polarity. A major increase in the experience of headaches (p = 0.007), chest pain (p < 0.001), and body aches and pains (p = 0.02) during depression was predicted by a major increase in state rejection sensitivity for both bipolar and unipolar depression. CONCLUSIONS: State, but not trait, rejection sensitivity is significantly predicted by bipolar depression, suggesting that this might be considered as a state marker for bipolar depression and taken into account in the clinical differentiation of bipolar and unipolar depression.

An exploratory study on the feasibility and appropriateness of family psychoeducation for postpartum women with psychosis in Uganda.

BACKGROUND: We explored how family psychoeducation could be made culturally sensitive for postpartum mothers with psychotic illness in a Ugandan setting. METHODS: A qualitative multi-method approach using an already existing family psychoeducation Tool Kit was adapted to incorporate lay perceptions related to psychotic illness in the postpartum period in this Ugandan setting. The participants consisted of postpartum women with psychotic illness, caregivers/family members, psychiatric nurses and psychologists. A modified version of a family psychoeducation programme for postpartum women with psychosis was formulated and pilot-tested. RESULTS: Modifications in the standard family psychoeducation programme were both in the process and content of family psychoeducation. Under process, effective communication, cultural background, appropriate dress, involving only one family member, low literacy, and flexibility in timekeeping were raised. The theme of content yielded the incorporation of lay perceptions of mental illness, family planning, income generating, and an emphasis of premorbid and morbid personalities of the patients. CONCLUSION: The basic principles and assumptions underlying psychoeducation remained the same. Changes made in the process and content of family psychoeducation reflected the social, cultural and gender reality of the population.

Malaria with neurological involvement in Ugandan children: effect on cognitive ability, academic achievement and behaviour.

BACKGROUND: Malaria is a leading cause of ill health and neuro-disability in children in sub-Saharan Africa. Impaired cognition is a common outcome of malaria with neurological involvement. There is also a possibility that academic achievement may be affected by malaria with neurological involvement given the association between cognitive ability and academic achievement. This study investigated the effect of malaria with neurological involvement on cognitive ability, behaviour and academic achievement. METHODS: This prospective case-control study was carried out in Kampala City, Uganda between February 2008 and October 2010. Sixty-two children with a history of malaria with neurological involvement were followed up and given assessments for cognitive ability (working memory, reasoning, learning, visual spatial skills and attention), behaviour (internalizing and externalizing problems) and academic achievement (arithmetic, spelling and reading) three months after the illness. Sixty-one community controls recruited from the homes or neighbouring families of the cases were also given the same assessments. Tests scores of the two groups were compared using analysis of covariance with age, sex, level of education, nutritional status and quality of the home environment as covariates. This study was approved by the relevant ethical bodies and informed consent sought from the caregivers. RESULTS: Children in the malaria group had more behavioural problems than the community controls for internalizing problems (estimated mean difference = -3.71, 95% confidence interval (CI), = -6.34 to -1.08, p = 0.007). There was marginal evidence of lower attention scores (0.40, CI = -0.05 to 0.86, p = 0.09). However, excluding one child from the analyses who was unable to perform the tests affected the attention scores to borderline significance (0.32, CI, = 0.01 to 0.62, p = 0.05). No significant differences were observed in other cognitive abilities or in academic achievement scores. CONCLUSION: Malaria with neurological involvement affects behaviour, with a minimal effect on attention but no detectable effect on academic achievement at three months post discharge. This study provides evidence that development of cognitive deficits after malaria with neurological involvement could be gradual with less effect observed in the short term compared to the long term.

Cognition, behaviour and academic skills after cognitive rehabilitation in Ugandan children surviving severe malaria: a randomised trial.

BACKGROUND: Infection with severe malaria in African children is associated with not only a high mortality but also a high risk of cognitive deficits. There is evidence that interventions done a few years after the illness are effective but nothing is known about those done immediately after the illness. We designed a study in which children who had suffered from severe malaria three months earlier were enrolled into a cognitive intervention program and assessed for the immediate benefit in cognitive, academic and behavioral outcomes. METHODS: This parallel group randomised study was carried out in Kampala City, Uganda between February 2008 and October 2010. Sixty-one Ugandan children aged 5 to 12 years with severe malaria were assessed for cognition (using the Kaufman Assessment Battery for Children, second edition and the Test of Variables of Attention), academic skills (Wide Range Achievement Test, third edition) and psychopathologic behaviour (Child Behaviour Checklist) three months after an episode of severe malaria. Twenty-eight were randomised to sixteen sessions of computerised cognitive rehabilitation training lasting eight weeks and 33 to a non-treatment group. Post-intervention assessments were done a month after conclusion of the intervention. Analysis of covariance was used to detect any differences between the two groups after post-intervention assessment, adjusting for age, sex, weight for age z score, quality of the home environment, time between admission and post-intervention testing and pre-intervention score. The primary outcome was improvement in attention scores for the intervention group. This trial is registered with Current Controlled Trials, number ISRCTN53183087. RESULTS: Significant intervention effects were observed in the intervention group for learning mean score (SE), [93.89 (4.00) vs 106.38 (4.32), P = 0.04] but for working memory the intervention group performed poorly [27.42 (0.66) vs 25.34 (0.73), P = 0.04]. No effect was observed in the other cognitive outcomes or in any of the academic or behavioural measures. CONCLUSIONS: In this pilot study, our computerised cognitive training program three months after severe malaria had an immediate effect on cognitive outcomes but did not affect academic skills or behaviour. Larger trials with follow-up after a few years are needed to investigate whether the observed benefits are sustained. TRIAL REGISTRATION: ISRCTN: ISRCTN53183087.

Pain and rejection sensitivity in bipolar depression.

OBJECTIVES: Little is known regarding the correlates of pain in bipolar disorder. Recent neuroimaging studies support the contention that depression, as well as pain distress and rejection distress, share the same neurobiological circuits. In a recently published study, we confirmed the hypothesis that perception of increased pain during treatment-refractory depression, predominantly unipolar, was related to increased rejection sensitivity. In the present study, we aimed to test this same hypothesis for bipolar depression. METHODS: The present study analysed data from 67 patients presenting to the Black Dog Institute Bipolar Disorders Clinic in Sydney, Australia. The patients all met DSM-IV criteria for bipolar disorder and had completed a self-report questionnaire regarding perceived pain and rejection sensitivity during depression. RESULTS: A significant increase in the experience of headaches (p=0.003) as well as chest pain (p=0.004) during bipolar depression was predicted by a major increase in rejection sensitivity when depressed, i.e., state rejection sensitivity. Being rejection sensitive in general, i.e., trait rejection sensitivity, did not predict pain during depression. CONCLUSIONS: The experience of increased headaches and chest pain during bipolar depression is related to increased rejection sensitivity during depression. Research to further elucidate this relationship is required.

Arterial baroreflex dysfunction in major depressive disorder.

OBJECTIVES: Patients treated for major depressive disorder are at increased risk for sudden cardiac death. Impaired arterial baroreflex function has been associated with ventricular arrhythmias. Our hypothesis was that arterial baroreflex dysfunction prevails in major depressive disorder and that electroconvulsive therapy in conjunction to medical therapy would improve both depressive symptoms and baroreflex function. METHODS: Thirty-three patients with major depressive disorder who were treated in hospital were studied before and after electroconvulsive treatment. Eighteen patients underwent follow-up investigations 6 months after discharge. ECG and beat-to-beat blood pressures were recorded continuously. Arterial baroreflex sensitivity (BRS) and effectiveness index were calculated. Twenty healthy subjects were examined for comparison. RESULTS: Heart rate and systolic blood pressures were elevated (P < 0.01 for all) in depressive patients before treatment when compared with healthy subjects, whereas arterial BRS and baroreflex effectiveness were reduced (10 +/- 7 vs. 15 +/- 5 ms/mmHg and 0.35 +/- 0.20 vs. 0.48 +/- 0.14, P < 0.01 for both). Whereas depressive symptoms decreased after treatment (P < 0.05), blood pressures, heart rate, arterial BRS, and effectiveness remained unchanged. At follow-up, 6 months after discharge all variables were unchanged when compared with values obtained at discharge. CONCLUSION: Both the sensitivity and the number of times the arterial baroreflex is being active are reduced in major depressive disorder and this baroreflex dysfunction may prevail long-term when depressive symptoms have improved.

Identifying predictors for good lithium response - a retrospective analysis of 100 patients with bipolar disorder using a life-charting method.

PURPOSE: Our aim was to investigate bipolar patients in order to test the validity of various outcome measures and to identify prognostic predictors for pharmacological treatment. MATERIAL AND METHOD: One hundred patients were interviewed using a computerized life-charting program in a descriptive, retrospective analysis. The concept "Burden of illness" was defined as a combination of severity and duration of episodes. Response to treatment was defined as the difference in burden before and after treatment, a low burden during treatment, and freedom of episodes for at least 3 years after insertion of treatment. RESULTS: The absence of mixed episodes and a high initial burden predicted a good response measured as the difference in burden. If remission for 3 years or a low burden during lithium treatment was used, the absence of rapid cycling and of mixed episodes were the most important predictors. The severity of illness before treatment had no impact. DISCUSSION AND CONCLUSION: We suggest the use of absolute measures of severity during treatment as the most appropriate measure of the outcome. Furthermore, our data provide corroboration that treatment with lithium ameliorates the prognosis of the illness, but that mixed episodes and rapid cycling predict a poorer response to lithium.

Adaptation of the multidimensional scale of perceived social support in a Ugandan setting.

BACKGROUND: The Multidimensional Scale of Perceived Social Support (MSPSS) was developed in the USA. The adequacy of its use in Uganda to guarantee its reliability and validity has not been ascertained. AIM: Thus the aim of the present study was to adapt the MSPSS scale by testing the validity and reliability of the scale in a Ugandan setting. METHODS: A cross-sectional design was employed and 240 respondents were consecutively recruited from postnatal clinics in Mulago hospital. Analysis of reliability and validity of the adapted MSPSS was done. Cronbach alpha and principal component analyses were respectively generated. RESULTS: Three subscales of the MSPSS that had been identified in other populations were evident in the Ugandan population. Using the Cronbach's alpha, the MSPSS demonstrated good internal consistency at .83. A dendrogram indicated that all sub items of the MSPSS were inter-linked. Exploratory Factor analysis derived three components. Principal Component analysis using rotated varimax generated high loadings on all subscales. CONCLUSION: The adapted MSPSS can reliably be used in Uganda.

HPA axis activation determined by the CRH challenge test in patients with few versus multiple episodes of treatment-refractory depression.

OBJECTIVE: In clinical guidelines, risk factors for a malignant illness course include 3 or more lifetime episodes of depression. Our aim was to investigate the activation of the hypothalamic-pituitary-adrenal hormonal axis in treatment-refractory affective disorder in pauciepisodic (one or two episodes) versus multiepisodic (three or more episodes) patients. METHODS: We evaluated the HPA axis in 37 patients with treatment-refractory affective disorder and in 27 healthy volunteers by measuring adrenocorticotropin hormone (ACTH) and cortisol responses following administration of corticotropin-releasing hormone (CRH). In retrospective life charts was recorded every previous illness episode for each patient. RESULTS: Seven of the patients were pauciepisodic and 30 were multiepisodic. The pauciepisodic patients had significantly larger peak and total ACTH responses to CRH compared to the multiepisodic patients as well as to the control group. Multiepisodic patients showed no difference compared to controls in ACTH secretion pre- and post-CRH. Cortisol secretion was the same in all three groups. CONCLUSIONS: The pituitary adrenocortical responses were stronger in pauciepisodic patients than in multiepisodic patients and in volunteers. This cross-sectional study suggests that the HPA axis, in refractory multiepisodic affective disorders, might weaken its original activity as the illness recurs with more episodes.

Decreased cerebrospinal fluid neuropeptide Y (NPY) in patients with treatment refractory unipolar major depression: preliminary evidence for association with preproNPY gene polymorphism.

Extensive animal studies suggest neuropeptide Y (NPY) to be involved in coping with a wide range of stressors, and that impaired central NPY signalling could be involved in the pathophysiology of anxiety and depression. Human studies of central NPY levels in depression have, however, been inconclusive. Here, we examined levels of NPY-like immunoreactivity (NPY-LI) in the cerebrospinal fluid (CSF) of medication-free subjects with treatment refractory unipolar depression. Patients were admitted to a research inpatient unit, examined under standardized conditions, and compared with a sample of volunteers in whom psychiatric morbidity was excluded. A robust suppression of NPY levels in patient CSF was found, while other putative CSF markers (monoamine metabolites, somatostatin) did not differ between the groups. We then explored whether this finding might be related to a recently described T1128C coding polymorphism which results in a Leu7-> Pro7 substitution of the signal peptide, and a previously not described T -399C polymorphism in the promoter region of the preproNPY gene. Preliminary evidence was found for an association of both markers with a diagnosis of depression, indicating the possibility of an underlying haplotype influencing the vulnerability for developing depressive illness. Our present findings are in line with an extensive animal literature, and further support the notion that impaired NPY function could contribute to depressive illness.

Lifetime burden of mood swings and activation of brain norepinephrine turnover in patients with treatment-refractory depressive illness.

BACKGROUND: We tested if duration and intensity of episodes in treatment-resistant affectively ill patients were related to cerebrospinal fluid (CSF) concentrations of monoamine metabolites. METHOD: In retrospective life charts were recorded every previous episode of 37 patients with severe treatment-refractory affective disorders. 'Accumulated burden of mood swings' (ABMS, sum of each episode length x episode depth) was used to estimate the accumulated illness burden. Homovanillic acid (HVA), 3-methoxy-4-hydroxyphenylglycol (MHPG), and 5-hydroxyindoleacetic acid (5-HIAA) were analyzed in CSF of all patients and compared with 27 healthy controls. Data were analyzed using multiple regression analysis. RESULTS: CSF MHPG contributed strongly significant and positively to the ABMS. LIMITATIONS: The retrospective nature of the study. CONCLUSION: CSF concentrations of MHPG is positively related to ABMS over life. Thus, a specific involvement of norepinephrine in the long-term burden of affective illness is a likely reality.

Patterns of sensitisation in the course of affective illness. A life-charting study of treatment-refractory depressed patients.

BACKGROUND: A 'sensitisation' process over time has been suggested by Post and collaborators--the affective illness course shows a tendency towards more frequent, deeper, and less stress-related episodes over time. The main aim of the present study was to test the sensitisation hypothesis using a Swedish Life Charting program. METHODS: Thirty patients with treatment-refractory affective disorder, of whom four had bipolar I disorder, were first interviewed using a semistructural interview manual covering episodes, treatment and stress. All previous psychiatric records were then recruited. Information from the records and from the interview was coalesced into individual, retrospective life charts. RESULTS: Twenty patients showed a sensitisation course and 10 patients showed a non-sensitisation course. In both groups, almost 90% of illness episodes had undergone treatment. Time spent in illness since onset of the affective disorder was about 33% for the sensitisation group and more than 50% for the non-sensitisation group. LIMITATIONS: The retrospective nature of the study is a limitation. The results apply to patients with severe treatment-refractory affective disorder and may not be generalisable to general patients with less-severe mood disorders. CONCLUSIONS: Our results partially validate the Post hypothesis of affective sensitisation in demonstrating this phenomenon in more than half of our affectively recurring patients. However, a substantial minority were clearly non-sensitisers showing a stable but more malignant illness course. Future studies need to elucidate whether these two groups benefit from different kinds of treatment.