Essential data dimensions for prospective international data collection in older age bipolar disorder (OABD): Recommendations from the GAGE-BD group

BACKGROUND

By 2030, over 50% of individuals living with bipolar disorder (BD) are expected to be aged ≥50 years. However, older age bipolar disorder (OABD) remains understudied. There are limited large-scale prospectively collected data organized in key dimensions capable of addressing several fundamental questions about BD affecting this subgroup of patients.

METHODS

We developed initial recommendations for the essential dimensions for OABD data collection, based on (1) a systematic review of measures used in OABD studies, (2) a Delphi consensus of international OABD experts, (3) experience with harmonizing OABD data in the Global Aging & Geriatric Experiments in Bipolar Disorder Database (GAGE-BD, n ≥ 4500 participants), and (4) critical feedback from 34 global experts in geriatric mental health.

RESULTS

We identified 15 key dimensions and variables within each that are relevant for the investigation of OABD: (1) demographics, (2) core symptoms of depression and (3) mania, (4) cognition screening and subjective cognitive function, (5) elements for BD diagnosis, (6) descriptors of course of illness, (7) treatment, (8) suicidality, (9) current medication, (10) psychiatric comorbidity, (11) psychotic symptoms, (12) general medical comorbidities, (13) functioning, (14) family history, and (15) other. We also recommend particular instruments for capturing some of the dimensions and variables.

CONCLUSION

The essential data dimensions we present should be of use to guide future international data collection in OABD and clinical practice. In the longer term, we aim to establish a prospective consortium using this core set of dimensions and associated variables to answer research questions relevant to OABD.

Clinical correlates of late-onset versus early-onset bipolar disorder in a global sample of older adults

OBJECTIVES

Late-onset bipolar disorder (LOBD) represents a significant subgroup of bipolar disorder (BD). However, knowledge for this group is mostly extrapolated from small studies in subjects with early/mixed age of illness onset. In this global sample of older adults with BD (OABD: ≥50 years old) we aim to characterize the sociodemographic and clinical presentation of LOBD (≥40 years at BD onset) compared to early-onset BD (EOBD: <40 years at BD onset).

METHODS

The Global Aging and Geriatric Experiments in Bipolar Disorder consortium provided international data on 437 older age bipolar disorder participants. We compared LOBD versus EOBD on depression, mania, functionality, and physical comorbidities. Exploratory analyses were performed on participants with BD onset ≥50 years old.

RESULTS

LOBD (n = 105) did not differ from EOBD (n = 332) on depression, mania, global functioning, nor employment status (p > 0.05). Late-onset bipolar disorder was associated with higher endocrine comorbidities (odds ratio = 1.48, [95%CI = 1.0,12.1], p = 0.03). This difference did not remain significant when subjects with BD onset ≥50 years old were analyzed.

LIMITATIONS

This study is limited by the retrospective nature of the variable age of onset and the differences in evaluation methods across studies (partially overcame by harmonization processes).

CONCLUSION

The present analysis is in favor of the hypothesis that LOBD might represent a similar clinical phenotype as classic EOBD with respect to core BD symptomatology, functionality, and comorbid physical conditions. Large-scale global collaboration to improve our understanding of BD across the lifespan is needed.

Illness progression in older-age bipolar disorder: Exploring the applicability, dispersion, concordance, and associated clinical markers of two staging models for bipolar disorder in an older population

OBJECTIVES

The validity and applicability of two existing staging models reflecting illness progression have been studied in bipolar disorder (BD) in adults, but not in older adult populations. Staging model A is primarily defined by the number and recurrence of mood episodes, model B is defined by the level of inter-episodic functioning. This study aimed to explore the applicability, dispersion, and concordance of, and associations with clinical markers in these two staging models in older-age bipolar disorder (OABD).

METHODS

Using cross-sectional data from the Dutch Older Bipolars study, OABD outpatients (N = 126, ≥50 years) were staged using models A and B. Dispersion over the stages and concordance between the models were assessed. Associations were explored between model stages and clinical markers (familial loading, childhood abuse, illness duration, episode density, treatment resistance, Mini-Mental State Examination, and composite cognitive score).

RESULTS

Ninety subjects could be assigned to model A, 111 to model B, 80 cases to both. The majority (61%) had multiple relapses (model A, stage 3C) but were living independently (model B, stage I-III). Concordance between models was low. For model A, the markers childhood abuse, illness duration, and episode density significantly increased over subsequent stages. Model B was not associated with a significant change in any marker.

CONCLUSIONS

Assigning stages to OABD subjects was possible for both models, with age-related adjustments for model B. Model B as currently operationalized may be less suitable for OABD or may measure different aspects of illness progression, reflected by its low correspondence with model A and lack of associated clinical markers.

Episodic density, subsyndromic symptoms, and mood instability in late-life bipolar disorders: A 5-year follow-up study

OBJECTIVES

Characterization of clinical course in old age bipolar disorder (OABD) is scarce and based solely on episode density (ED). The aim of this study was to explore mood instability (MI) and subsyndromal symptomatology (SS) in a prospective cohort of OABD. Further, we contrasted these measures with a cohort of young age bipolar disorder (YABD).

METHODS

Life charts from weekly mood ratings were used to compute the number of weeks spent with subsyndromal symptoms (SD), the ED, and the MI during follow-up for a cohort of OABD (N = 38) that excluded late onset BD. Linear and logistic regression models were fitted to compare the clinical course of OABD with a cohort of YABD (N = 52) and to explore the relationship between these measures and functional outcomes.

RESULTS

Median follow-up was 5 years (IQR: 3.6-7.9). OABD (61.6 years, SD: 8.3) spent 15%, 6%, and 3% of their follow-up with depressive, manic, and mixed symptoms, respectively, and suffered 4.2 mood changes per year (SD: 2.6). No significant differences between OABD and YABD regarding ED or MI emerged in multivariate analysis, while a higher subsyndromal manic symptom burden was observed in OABD (β coefficient: 3.79, 95%CI: 0.4-7.2). Both SS and MI were associated with functional outcomes in OABD.

CONCLUSIONS

The course of illness throughout OABD was similar to the one observed in YABD except for a higher subsyndromal manic burden. This study extended the association of MI and SD with global functioning to the late-life BD.