Empathy in frontotemporal dementia and Alzheimer's disease

Abstract

Using naturalistic stimuli, we assessed the ability to infer what other people are feeling in three groups of participants: healthy elderly adults, patients suffering from the behavioral variant of frontotemporal dementia (FTD-b), and patients suffering from Alzheimer's disease (AD). After watching videotaped interviews of everyday people (nonactors) discussing an emotionally relevant event in their lives, participants answered questions regarding the interviewee's feelings. Both patient groups inferred emotions as accurately as the healthy elderly, provided the emotions were displayed unambiguously and consistently across the interview. However, when the displayed emotions became more variable and ambiguous, performance in both patient groups became impaired relative to healthy elderly participants. The similar profile across the two clinical groups despite their differences in social skills suggests that nonsocial cognitive processes affected in dementia may be an important factor in drawing inferences about other people's feelings.

Keywords:

• Empathy
• Dementia
• Emotion
• Theory of mind
• Naturalistic stimulus

We would like to thank the patients and their families for their time and effort. Vicki del Rio, Allison Todd, and Ryan Corser helped with data coding. This research was supported in part by a grant from the Center for Consciousness Studies of the University of Arizona to Diego Fernandez-Duque and an operating grant from the Canadian Institute for Health Research to Sandra Black (Grant No. 13129). Diego Fernandez-Duque was also supported by a fellowship from the Heart and Stroke Foundation of Ontario (Grant No. F4866).

Notes

¹Not only did coders agree with our decision of which items matched the interviewee's statements, but also the intercoder agreement was very high, ranging from 82% agreement between Coders 1 and 2 to 88% agreement between Coders 2 and 3. This agreement remained very high independent of which interviewee's feelings were being assessed. There was an intercoder agreement of 88.1% for emotions attributed to the positive mother, 83.9% for emotions attributed to the ambivalent mother, and 82.7% for emotions attributed to the senior citizen. A chi-square analysis revealed no statistically significant differences, χ²(2) = 1.19, p > .50.

²The IRI also has a “fantasy” subscale. We did not include this scale because it was less relevant to the concept of empathy as discussed in this study.

³In an initial analysis, we included interviewee type (positive, ambivalent, negative) as a third factor. This analysis revealed a significant three-way interaction, F(4, 48) = 3.45, p = .01, which we explored by running 3 × 2 analyses at each level of interviewee type, as described in the main text.