Old data, new interpretation: a re-analysis of Sir Aubrey Lewis' M.D. thesis

Melancholia defined with the precision of a machine

BACKGROUND

The status of melancholia as a categorical or dimensional condition remains unclear, and no measure of melancholia has achieved definitive status. This study aimed to use a machine learning approach to assess whether a pre-established cut-off score on the Sydney Melancholia Prototype Index (SMPI) provided clear differentiation of melancholic/non-melancholic depression, and to identify the items making the most distinct contribution.

METHODS

We analysed amalgamated data sets of 1513 clinically depressed patients assessed via the clinician-rated version of the SMPI (SMPI-CR). We also evaluated the self-report version of the SMPI (SMPI-SR) in a combined clinical/community sample of 2025 depressed patients and senior high school students. Rule ensembles were derived in which the outcome measure was the presence/absence of melancholia (defined as scoring at or above a SMPI cut-off score that had been established in previous studies) and the predictive variables were the individual SMPI items.

RESULTS

The pre-established SMPI cut-off score was confirmed as differentiating melancholic/non-melancholic with near perfect accuracy for the SMPI-CR, and with very high accuracy for the SMPI-SR. The relative importance of all SMPI items was quantified.

LIMITATIONS

It is difficult to validate SMPI-assigned diagnoses due to the lack of any similar measures.

CONCLUSIONS

The SMPI-CR was confirmed to be a highly precise instrument for differentiating melancholic and non-melancholic depression. Its use will advance clinical decision making and studies evaluating causes, mechanisms and treatments for the two depressive sub-types, as well as assist clarification as to whether melancholia is categorically or dimensionally distinct from non-melancholic depression.

The bipolar disorders: A case for their categorically distinct status based on symptom profiles

BACKGROUND

It is unclear whether the bipolar disorders (i.e. BP-I/BP-II) differ dimensionally or categorically. This study sought to clarify this issue.

METHODS

We recruited 165 patients, of which 69 and 96 had clinician-assigned diagnoses of BP-I and BP-II respectively. Their psychiatrists completed a data sheet seeking information on clinical variables about each patient, while the patients completed a different data sheet and scored a questionnaire assessing the prevalence and severity of 96 candidate manic/hypomanic symptoms.

RESULTS

We conducted a series of analyses examining a set (and two sub-sets) of fifteen symptoms that were significantly more likely to be reported by the clinically diagnosed BP-I patients. Latent class analyses favoured two-class solutions, while mixture analyses demonstrated bimodality, thus arguing for a BP-I/BP-II categorical distinction. Statistically defined BP-I class members were more likely when manic to have experienced psychotic features and over-valued ideas. They were also more likely to have been hospitalised, and to have been younger when they received their bipolar diagnosis and first experienced a depressive or manic episode.

LIMITATIONS

The lack of agreement between some patients and managing clinicians in judging the presence of psychotic features could have compromised some analyses. It is also unclear whether some symptoms (e.g. grandiosity, noting mystical events) were capturing formal psychotic features or not.

CONCLUSIONS

Findings replicate our earlier study in providing evidence to support the modelling of BP-I and BP-II as categorically discrete conditions. This should advance research into aetiological factors and determining optimal (presumably differing) treatments for the two conditions.

Subtypes of Depression: Latent Class Analysis in Spanish Old People with Depressive Symptoms

Major depressive disorder (MDD) is one of the most disabling disorders and the one that most contributes to disability. When it occurs in older people, it is an additional burden to their potential physical and cognitive deficiencies, making MDD an important public health problem that supposes a large investment in health. There is a clear lack of consistency between the subtypes of depression found in the literature, ranging from two to seven classes, with three being the most commonly found non-melancholic, melancholic and psychotic, or putative psychotics. The aim of this research is to add knowledge to the profiles of depressive symptoms in a representative sample of older Spanish people, and to study the possible relationship of these symptom profiles with variables that have traditionally been related to depression. Spanish data from the sixth wave of SHARE were used, with 612 Spanish older adults living in Spain. A routine of several LCAs with a different number of classes was performed to answer this first aim to classify Spanish adults with depression symptoms. The results pointed out the presence of three different classes among the participants in the study: psychosomatic (11.12%), melancholic (14.21%), and anhedonic (74.67%). This work represents a step forward to understand the heterogeneity of major depressive disorder, facilitating the diagnosis, and subsequent treatment of older adults.

The use of latent class analysis for identifying subtypes of depression: A systematic review

Depression is a significant public health problem but symptom remission is difficult to predict. This may be due to substantial heterogeneity underlying the disorder. Latent class analysis (LCA) is often used to elucidate clinically relevant depression subtypes but whether or not consistent subtypes emerge is unclear. We sought to critically examine the implementation and reporting of LCA in this context by performing a systematic review to identify articles detailing the use of LCA to explore subtypes of depression among samples of adults endorsing depression symptoms. PubMed, PsycINFO, CINAHL, Scopus, and Google Scholar were searched to identify eligible articles indexed prior to January 2016. Twenty-four articles reporting 28 LCA models were eligible for inclusion. Sample characteristics varied widely. The majority of articles used depression symptoms as the observed indicators of the latent depression subtypes. Details regarding model fit and selection were often lacking. No consistent set of depression subtypes was identified across studies. Differences in how models were constructed might partially explain the conflicting results. Standards for using, interpreting, and reporting LCA models could improve our understanding of the LCA results. Incorporating dimensions of depression other than symptoms, such as functioning, may be helpful in determining depression subtypes.

The Major Depressive Disorder Hierarchy: Rasch Analysis of 6 items of the Hamilton Depression Scale Covering the Continuum of Depressive Syndrome

OBJECTIVES

Melancholic features of depression (MFD) seem to be a unidimensional group of signs and symptoms. However, little importance has been given to the evaluation of what features are related to a more severe disorder. That is, what are the MFD that appear only in the most depressed patients. We aim to demonstrate how each MFD is related to the severity of the major depressive disorder.

METHODS

We evaluated both the Hamilton depression rating scale (HDRS-17) and its 6-item melancholic subscale (HAM-D6) in 291 depressed inpatients using Rasch analysis, which computes the severity of each MFD. Overall measures of model fit were mean (±SD) of items and persons residual = 0 (±1); low χ2 value; p>0.01.

RESULTS

For the HDRS-17 model fit, mean (±SD) of item residuals = 0.35 (±1.4); mean (±SD) of person residuals = -0.15 (±1.09); χ2 = 309.74; p<0.00001. For the HAM-D6 model fit, mean (±SD) of item residuals = 0.5 (±0.86); mean (±SD) of person residuals = 0.15 (±0.91); χ2 = 56.13; p = 0.196. MFD ordered by crescent severity were depressed mood, work and activities, somatic symptoms, psychic anxiety, guilt feelings, and psychomotor retardation.

CONCLUSIONS

Depressed mood is less severe, while guilt feelings and psychomotor retardation are more severe MFD in a psychiatric hospitalization. Understanding depression as a continuum of symptoms can improve the understanding of the disorder and may improve its perspective of treatment.

The properties and utility of the CORE measure of melancholia

BACKGROUND

The CORE measure was designed to assess a central feature of melancholia - signs of psychomotor disturbance (PMD) - and so provide an alternate non-symptom based measure of melancholia or of its probability. This review evaluates development and application studies undertaken over the last 25 years to consider how well it has met its original objectives.

METHODS

All studies published using the CORE measure as either the only or an adjunctive measure of melancholia were obtained and are considered in this review.

RESULTS

Findings suggest high reliability in quantifying CORE scores can be achieved and that it has construct validity as a measure of PMD. A number of application studies assessing socio-demographic factors, cognitive and motor impairment, dexamethasone suppression and thyrotropin-releasing hormone, response to psychotherapy and to electroconvulsive therapy support its validity as a measure of melancholia, while functional brain imaging studies suggest that the measure identifies regions of decreased connectivity.

LIMITATIONS

Use of the CORE benefits from rater training and for subjects to be assessed at or near nadir of their depressive episode. There have been insufficient studies evaluating genetic factors, and the treatment response of CORE-defined melancholic patients to antidepressant drugs of differing classes.

CONCLUSIONS

The CORE, either as a proxy or direct measure of melancholia, provides a strategy for assigning depressed subjects a diagnosis or melancholic or non-melancholic depression or for estimating the probability of melancholia.

Cleaving depressive diseases from depressive disorders and non-clinical states

OBJECTIVE

We sought to determine whether putative depressive diseases could be differentiated categorically from clinical depressive disorders and non-clinical mood states.

METHOD

We interviewed volunteers who reported or denied any lifetime depressive mood state and analyzed data from the former group reporting on their 'most severe' depressive episode. We employed latent class analysis (LCA) to determine whether a two-class solution was supported and the contribution of individual variables to class allocations.

RESULTS

All variables were significant predictors of class allocation. LCA-assigned Class I participants reported more depressive symptoms, had more distressing episodes and more lasting consequences, were more likely to view their depression as 'like a disease', and as being both disproportionately more severe and persistent in relation to any antecedent stressor. Validation involved comparison of LCA assignment with DSM-IV diagnosis for their most severe depressive episode. Of those assigned to Class I, 89% had a DSM diagnosis of melancholic, psychotic or bipolar depression. Class II had all those failing to meet criteria for a depressive episode and the majority of those with a non-melancholic depressive condition.

CONCLUSION

Despite not including individual depressive symptoms, study variables strongly differentiated putative depressive diseases from a composite of clinical depressive conditions and subclinical depressive states.

A case for reprising and redefining melancholia

OBJECTIVE

To identify limitations to severity-based classifications of depression, and to argue for positioning melancholia as a distinct melancholic subtype.

METHOD

An overview of relevant literature was conducted.

RESULTS

First, dimensionalizing depressive disorders effectively aggregates multiple heterogeneous depressive types and syndromes, and thus limits, if not prevents, identification of differential causes and treatments. Second, the melancholic depressive subtype can be defined with some relative precision, and that as it shows quite distinct differential treatment responsiveness, identification should be a clinical priority.

CONCLUSION

Melancholia can be positioned and classified as a distinct depressive subtype.

Does testing for bimodality clarify whether the bipolar disorders are categorically or dimensionally different to unipolar depressive disorders?

BACKGROUND

It has been held that if bipolar disorder is categorically distinct, it should differentiate from unipolar depressive disorders by showing bimodality or a 'zone of rarity' in bipolar symptom scores. Two previous studies have failed to demonstrate bimodality. We undertook a third study.

METHODS

A total of 1106 patients attending the Black Dog Institute Depression Clinic completed the Mood Disorders Questionnaire (MDQ), in addition to undergoing clinical assessment by an Institute psychiatrist.

RESULTS

The distributions of scores for the total number of hypomanic symptoms endorsed by unipolar and bipolar patients were both skewed, with the bipolar group endorsing a high number of hypomanic symptoms and the unipolar group endorsing few symptoms--and so giving the impression of an 'even' distribution generated by two quite distinctly differing sub-groups. However, formal statistical analyses involving mixed modelling provided no clear evidence that a bimodal distribution provided a better fit to the data than a unimodal one.

CONCLUSIONS

Failure to statistically demonstrate a 'point of rarity' did not marry with visual inspection of the plotted data--which clearly suggested two groups putatively capturing those with bipolar and unipolar disorders respectively. The paper considers some limitations to the emphasis on 'bimodality' in differentiating potentially differing conditions.

Defining melancholia: the primacy of psychomotor disturbance

OBJECTIVE

To argue that melancholia is a categorically distinct depressive condition, able to be differentiated from other depressive conditions by a neuropathological marker [observable psychomotor disturbance (PMD)] and having a differential response to various antidepressant treatments.

METHOD

The above statements are addressed by review of a wide body of research, which identified observable PMD as the cardinal marker of melancholia and developed the CORE measure as a strategy for assessing PMD and its severity. Properties of the CORE measure, including reliability, validity and treatment prediction, are overviewed.

RESULTS

A case is made for defining melancholia and a strategy for establishing its probability.

CONCLUSION

Melancholia is positioned as a categorical entity capable of being circumscribed by its cardinal feature of PMD.

Normal ageing, impaired cognitive functioning and senile dementia--a mixture distribution analysis

Dementia scores in the population of very old subjects (70 to > 95) do not demonstrate an obviously bimodal distribution. Heterogeneity analysis shows that the assumption of two distributions explains the data significantly better than a one-distribution model. The first distribution component can be regarded as representing the normal range of scores and the second the pathological range. An additional finding shows the normal range distribution shifting to the pathological range with increasing age.