Sensitivity of the Social Behavior Observer Checklist to Early Symptoms of Patients With Frontotemporal Dementia

Abstract

Background and objectives: Changes in social behavior are common symptoms of frontotemporal lobar degeneration (FTLD) and Alzheimer disease syndromes. For early identification of individual patients and differential diagnosis, sensitive clinical measures are required that are able to assess patterns of behaviors and detect syndromic differences in both asymptomatic and symptomatic stages. We investigated whether the examiner-based Social Behavior Observer Checklist (SBOCL) is sensitive to early behavior changes and reflects disease severity within and between neurodegenerative syndromes.

Methods: Asymptomatic individuals and patients with neurodegenerative disease were selected from the multisite ALLFTD cohort study. In a sample of participants with at least 1 time point of SBOCL data, we investigated whether the Disorganized, Reactive, and Insensitive subscales of the SBOCL change as a function of disease stage within and between these syndromes. In a longitudinal subsample with both SBOCL and neuroimaging data, we examined whether change over time on each subscale corresponds to progressive gray matter atrophy.

Results: A total of 1,082 FTLD pathogenic variant carriers and noncarriers were enrolled (282 asymptomatic, 341 behavioral variant frontotemporal dementia, 114 semantic and 95 nonfluent variant primary progressive aphasia, 137 progressive supranuclear palsy, and 113 Alzheimer disease syndrome). The Disorganized score increased between asymptomatic to very mild (p = 0.016, estimate = -1.10, 95% CI = -1.99 to -0.22), very mild to mild (p = 0.013, estimate = -1.17, 95% CI = -2.08 to -0.26), and mild to moderate/severe (p < 0.001, estimate = -2.00, 95% CI = -2.55 to -1.45) disease stages in behavioral variant frontotemporal dementia regardless of pathogenic variant status. Asymptomatic GRN pathogenic gene variant carriers showed more reactive behaviors (preoccupation with time: p = 0.001, estimate = 1.11, 95% CI = 1.06 to 1.16; self-consciousness: p = 0.003, estimate = 1.77, 95% CI = 1.52 to 2.01) than asymptomatic noncarriers (estimate = 1.01, 95% CI = 0.98 to 1.03; estimate = 1.31, 95% CI = 1.20 to 1.41). The Insensitive score increased to a clinically abnormal level in advanced stages of behavioral variant frontotemporal dementia (p = 0.003, estimate = -0.73, 95% CI = -1.18 to -0.29). Higher scores on each subscale corresponded with higher caregiver burden (p < 0.001). Greater change over time corresponded to greater fronto-subcortical atrophy in the semantic-appraisal and fronto-parietal intrinsically connected networks.

Discussion: The SBOCL is sensitive to early symptoms and reflects disease severity, with some evidence for progression across asymptomatic and symptomatic stages of FTLD syndromes; thus, it may hold promise for early measurement and monitoring of behavioral symptoms in clinical practice and treatment trials.

Classification of evidence: This study provides Class II evidence that the SBOCL is sensitive to early behavioral changes in FTLD pathogenic variants and early symptomatic individuals in a highly educated patient cohort.

Figure 1. Reactive Subscale Scores in Asymptomatic Individuals

Asymptomatic GRN pathogenic variants are more preoccupied with time and more self-conscious than asymptomatic noncarriers. Linear models showed that asymptomatic GRN pathogenic variant carriers had significantly higher score than asymptomatic noncarriers on 2 checklist items that belong to the Reactive subscale: “Preoccupied with time” (p = 0.001; GRN: 1.11 ± 0.03, 95% CI = 1.06–1.16; noncarriers: 1.01 ± 0.01, 95% CI = 0.98–1.03) and overly self-conscious (p = 0.003; GRN: 1.77 ± 0.13, 95% CI = 1.52–2.01; noncarriers: 1.31 ± 0.05, 95% CI = 1.20–1.41). No significant differences were found between asymptomatic C9orf72 carriers and asymptomatic noncarriers or between asymptomatic MAPT carriers and asymptomatic noncarriers. C9orf72 = chromosome 9 open reading frame 72; GRN = progranulin; MAPT = microtubule-associated protein tau; SBOCL = Social Behavior Observer Checklist.

Figure 2. BvFTD Patients With a Reference Group of Asymptomatic Individuals

Of the 3 SBOCL subscales, the disorganized subscale best reflects disease severity in a sample of asymptomatic individuals and patients with bvFTD, regardless of whether they are pathogenic variant carriers or noncarriers. (A) The main effects model showed that CDR plus NACC FTLD was a significant predictor of the Disorganized score, showing that the score increased at each stage of disease severity from asymptomatic to very mild (p = 0.016, estimate = −1.10, 95% CI = −1.99 to −0.22), from very mild to mild (p = 0.013, estimate = −1.17, 95% CI = −2.08 to −0.26), and from mild to moderate/severe (p < 0.001, estimate = −2.00, 95% CI = −2.55 to −1.45) disease stage across pathogenic variant carriers and noncarriers. The interaction model incorporating pathogenic variant status showed that the score increased from asymptomatic to very mildly symptomatic stage in noncarriers (p = 0.024, estimate = −1.78, 95% CI = −3.28 to −0.28). (B) The Reactive score was a predictor of CDR plus NACC FTLD in the main effects analysis, showing that the score increased only in the very earliest disease stage, during the conversion from asymptomatic to very early symptomatic stage (p = 0.013, estimate = −0.72, 95% CI = −1.28 to −0.16). The interaction between CDR plus NACC FTLD and pathogenic variant status was not significant. (C) The Insensitive score was a predictor of CDR plus NACC FTLD in both the main effects and interaction models, showing that the score increased from asymptomatic to very mild disease (p = 0.049, estimate = −0.70, 95% CI = −1.39 to 0.00) and from mild to moderate/severe (p = 0.001, estimate = −0.74, 95% CI = −1.17 to −0.31) disease across pathogenic variant carriers and noncarriers. The interaction analysis showed that the significant increase from mild to moderate/severe disease stage occurred primarily in the noncarrier group. The gray line represents the threshold for clinically significant symptoms found in our cross-sectional validation study. The lower line in part C shows an updated, suggested threshold derived on the basis of these findings across different disease stages within syndromes, which differs from the threshold initially suggested in the earlier, cross-sectional validation paper. bvFTD = behavioral variant frontotemporal dementia; SBOCL = Social Behavior Observer Checklist.

Figure 3. Subscale Scores Across Disease Stages Within Syndromes

Disorganized and insensitive subscales are sensitive to disease progression in patients with bvFTD. (A) CDR plus NACC FTLD was a significant predictor of the Disorganized score, but the interaction of CDR plus NACC FTLD by diagnostic group was not significant. The results show that the subscale can detect early behavior changes occurring between very mild and mild disease stage (p = 0.019, estimate = 1.15, 95% CI = 0.22–2.07) in patients with bvFTD. In addition, the Disorganized score increased from mild to moderate/severe disease stage in patients with bvFTD (p < 0.001, estimate = −2.04, 95% CI = −2.59 to −1.47), nfvPPA (p = 0.002, estimate = −2.46, 95% CI = −3.85 to −1.06), and AD (p = 0.003, estimate = −1.83, 95% CI = −2.90 to −0.76). (B) The diagnostic group was a significant predictor (p < 0.001) of the Reactive score, showing that patients with AD and svPPA had significantly higher scores than other diagnostic groups at multiple disease stages. (C) Both CDR plus NACC FTLD (p = 0.007) and diagnostic group (p < 0.001) was a significant predictor of the Insensitive score. The only significant within-group effect of diagnosis was the increase from early to moderate/severe disease stage in patients with bvFTD (p = 0.002, estimate = −0.74, 95% CI = −1.13 to −0.35). The upper gray line represents the threshold for clinically significant symptoms found in our cross-sectional validation study. The lower line in part C shows an updated, suggested threshold derived on the basis of these findings across different disease stages within syndromes, which differs from the threshold initially suggested in the earlier, cross-sectional validation paper. AD = Alzheimer disease; bvFTD = behavioral variant frontotemporal dementia; nfvPPA = nonfluent variant primary progressive aphasia; PSP = progressive supranuclear palsy; svPPA = semantic variant primary progressive aphasia.