Neurodegenerative pathologies associated with behavioral and psychological symptoms of dementia in a community-based autopsy cohort

Abstract

In addition to the memory disorders and global cognitive impairment that accompany neurodegenerative diseases, behavioral and psychological symptoms of dementia (BPSD) commonly impair quality of life and complicate clinical management. To investigate clinical-pathological correlations of BPSD, we analyzed data from autopsied participants from the community-based University of Kentucky Alzheimer's Disease Research Center longitudinal cohort (n = 368 research volunteers met inclusion criteria, average age at death 85.4 years). Data assessing BPSD were obtained approximately annually, including parameters for agitation, anxiety, apathy, appetite problems, delusions, depression, disinhibition, hallucinations, motor disturbance, and irritability. Each BPSD was scored on a severity scale (0-3) via the Neuropsychiatric Inventory Questionnaire (NPI-Q). Further, Clinical Dementia Rating (CDR)-Global and -Language evaluations (also scored on 0-3 scales) were used to indicate the degree of global cognitive and language impairment. The NPI-Q and CDR ratings were correlated with neuropathology findings at autopsy: Alzheimer's disease neuropathological changes (ADNC), neocortical and amygdala-only Lewy bodies (LBs), limbic predominant age-related TDP-43 encephalopathy neuropathologic changes (LATE-NC), primary age-related tauopathy (PART), hippocampal sclerosis, and cerebrovascular pathologies. Combinations of pathologies included the quadruple misfolding proteinopathy (QMP) phenotype with co-occurring ADNC, neocortical LBs, and LATE-NC. Statistical models were used to estimate the associations between BPSD subtypes and pathologic patterns. Individuals with severe ADNC (particularly those with Braak NFT stage VI) had more BPSD, and the QMP phenotype was associated with the highest mean number of BPSD symptoms: > 8 different BPSD subtypes per individual. Disinhibition and language problems were common in persons with severe ADNC but were not specific to any pathology. "Pure" LATE-NC was associated with global cognitive impairment, apathy, and motor disturbance, but again, these were not specific associations. In summary, Braak NFT stage VI ADNC was strongly associated with BPSD, but no tested BPSD subtype was a robust indicator of any particular "pure" or mixed pathological combination.

Keywords: Aphasia; FTD; FTLD; HS-aging; Psychiatric; Psychoses; Psychosis; Synuclein; Tau; Tauopathy.

Fig. 1

A schematic representation of selected data analyzed, questions addressed, and statistical analyses in the present study

Fig. 2

Radar chart depicts the percent of cases with moderate or severe BPSD subtypes, stratified by Braak NFT stages (0-VI). Cases included for this chart were selected among participants (n = 212) that lacked neocortical LBs and also lacked LATE-NC stage > 1. The UDS parameters utilized in this chart were the maximum values experienced at any point in the research volunteers’ longitudinal course on study. In addition to those BPSD subtypes, final CDR assessments were used for language dysfunction and global cognitive impairment. The severity of multiple BPSD subtypes trended to be worse in more advanced Braak NFT stages. Asterisks indicate statistical significance: *(p < 0.05), **(p < 0.01), ***(p < 0.001): these are nominal p values, using Chi-square test with 5 degrees of freedom

Fig. 3

Radar chart depicts the percent of cases with moderate or severe BPSD subtypes, stratified by presence or absence of neocortical LBs and LATE-NC Stage > 1, among cases lacking severe ADNC (i.e., Braak NFT stages < V). Cases included for this chart were selected among participants (n = 202) that lacked severe ADNC. A number of the BPSD subtypes were more severe on average in cases with both LATE-NC and neocortical LBs. Asterisks indicate statistical significance: *(p < 0.05), **(p < 0.01), ***(p < 0.001): these are nominal p values, using Chi-square test with 3 degrees of freedom

Fig. 4

Radar chart depicts the percent of cases with moderate or severe BPSD subtypes, stratified by presence or absence of neocortical LBs and LATE-NC Stage > 1, among cases with severe ADNC (i.e., Braak NFT stages V or VI). Cases included for this chart were selected among participants with severe ADNC (n = 166). As was true in cases lacking severe ADNC, some of the BPSD subtypes were more severe on average in subjects with both LATE-NC stage > 1 and neocortical LBs. Asterisks indicate statistical significance: *(p < 0.05), **(p < 0.01), ***(p < 0.001): these are nominal p values, using Chi-square test with 3 degrees of freedom

Fig. 5

Non-proportional Venn diagram depicts the numbers of included cases in the present study according to various pathologic combinations, along with a summary list of the main BPSD subtypes associated with those pathology-defined categories