Affective theory of mind in patients with Parkinson's disease


Correspondence: Ubaldo Bonuccelli, MD, Dipartimento di Neuroscienze, Università di Pisa, Via Savi 10, 56100 Pisa, Italy. Email:



The aim of this study was to assess the hypothesis that patients with Parkinson's disease (PD) may have difficulties in tasks of affective theory of mind (ToM; the inference on others’ feelings) especially in moderate/advanced PD stages. Difficulties of cognitive ToM have already been described in several previous studies.


Affective ToM was assessed with the Reading the Mind in the Eyes task in 35 PD patients and 35 healthy controls. Depression, global cognitive status and executive functioning were also evaluated. Patients were distinguished in early PD and moderate PD according to their scores in the Hoehn and Yahr Staging Scale.


PD patients had more difficulties with affective ToM than healthy controls, also controlling for other variables that resulted in association with this ability. Early PD patients outperformed moderate PD patients, but this difference did not reach statistical significance when controlling for other variables.


These findings confirmed that affective ToM may be impaired in PD, but any conclusion can be made on the effect of disease progression on this ability of social cognition. Therefore, longitudinal studies are needed to investigate this potential effect.

EMPIRICAL INTEREST IN theory of mind (ToM; the ability to infer other people's mental states)[1] has recently involved neurodegenerative diseases.[2] ToM difficulties may occur in Parkinson's disease (PD)[3] from the early clinical stages and clearly involve cognitive ToM (inference about others’ beliefs),[4] while findings on affective ToM (inference about others’ feelings) are more controversial.[2, 3] The spatiotemporal progression of striatal dopamine depletion in PD patients[5] supports the hypothesis that this component could be impaired in more advanced stages,[6] when the ventromedial prefrontal cortex (involved in affective ToM), is hypo-stimulated by the orbital frontostriatal loop. Therefore moderate/advanced PD patients should have more difficulties of affective ToM than early PD patients, as investigated in the current study.


Thirty-five PD patients and 35 age-matched healthy controls (HC) were enrolled and gave informed written consent, in compliance with research standards for human research and in accordance with the Helsinki Declaration. All patients fulfilled diagnostic criteria for idiopathic PD.[7] HC had negative neurological and psychiatric history.

In PD patients, we recorded sex, age, Levodopa Equivalent Daily Dose (LEDD) and PD stage according to the Modified Hoehn and Yahr Staging Scale (H&Y).[8] According to the MDS Task Force Guidelines on the H&Y[9] we defined clinical stages of PD as early (H&Y Stages 1, 1.5 and 2), moderate (H&Y Stage 2.5 and 3) and advanced (H&Y Stages ≥ 4). PD patients and HC performed the 36-item full version of the Reading the Mind in the Eyes (RME),[10] a test of affective ToM consisting in the presentation of photographs of the eye region of human faces; participants are required to choose which word, among four options, best describes what the individual in the photograph is thinking or feeling. Explanations of ‘emotion’ labels, according to definitions provided by the Italian validation of the task,[11] were given to subjects when asked by the subjects themselves. The RME performance is indicated by the total number of correct choices. In addition to the RME, we administered the RME sex control task (identification of the sex of the subject in the photograph) to exclude the presence of visuospatial impairment. The Mini Mental State Examination (MMSE)[12] and the Frontal Assessment Battery (FAB)[13] were also administered to all subjects for a screening of global cognitive functioning and executive functioning; considering that affective status may influence ToM performance,[14] depression was assessed with the Beck Depression Inventory (BDI).[15]

Statistical analysis

Statistical explorations of the data by normality tests indicated a normal distribution only for RME performance; therefore the non-parametric Mann–Whitney U-test was used to compare PD and HC groups. Linear correlation and rank correlation were used to establish possible relations between RME performance and other variables. Although many variables did not have a normal distribution, considering that anova is robust against normality assumption violation in samples with adequate sizes, we performed two one-way ancova to compare RME performances of the subgroups controlling for variables that resulted significantly associated with the RME performance in the whole sample of subjects. Statistical analyses were performed with spss 18 (spss, Chicago, IL, USA).


There were no significant differences in demographic characteristics between patients and HC (Table 1). HC reported higher mean MMSE and FAB scores and a lower mean BDI score than patients (Table 1). The mean H&Y score of PD patients was 2.08 ± 0.72 (range 1–3; median 2.0); the H&Y score classified 20 patients as early PD and 15 patients as moderate PD.

Table 1. Sociodemographic, clinical and cognitive characteristics of PD patients and HC
 HC subjects (n = 35) Mean (SD)PD patients (n = 35) Mean (SD)Early PD patients (n = 20) Mean (SD)Moderate PD patients (n = 15) Mean (SD)
  1. *≤0.05 and **≤0.01 in the comparison between PD patients and HC. °≤0.05 and °°≤0.01 in the comparison between early PD patients and moderate PD patients.
  2. BDI, Beck Depression Inventory; FAB, Frontal Assessment Battery; H&Y, Hoehn & Yahr Stage of Parkinson's disease; HC, healthy controls; LEDD, Levodopa Equivalent Daily Dose; MMSE, Mini Mental State Examination; PD, Parkinson's disease; RME, Reading the Mind in the Eyes.
Sex: M/F15/2022/1314/68/7
Age66.97 (10.03)68.45 (6.69)66.83 (6.69)71.90 (4.88)°
Education10.83 (3.81)9.10 (3.90)9.68 (3.74)8.00 (4.16)
PD duration/6.31 (4.02)5.47 (4.19)7.90 (3.31)
H&Y Stage/2.08 (0.72)1.68 (0.41)2.85 (0.52)°°
LEDD/585.78 (382.27)542.89 (313.67)931.70 (385.16)°°
BDI1.89 (1.27)9.65 (4.49)**10.05 (4.36)8.80 (4.89)
MMSE29.37 (1.06)28.52 (1.59)**29.16 (1.06)27.30 (1.67)°°
FAB16.74 (1.50)14.21 (3.16)**15.53 (2.67)11.70 (2.49)°°
RME22.14 (4.20)16.86 (4.81)**18.47 (4.40)13.80 (4.18)
RME sex control task34.03 (1.33)33.31 (2.56)33.74 (2.05)32.50 (3.30)

Moderate PD patients were older, with a higher mean LEDD and reported lower mean MMSE and FAB scores than early PD patients (Table 1). HC outperformed PD patients in the RME (P < 0.01). In distinguishing patients on the basis of their PD stage (early vs moderate), the HC outperformed both PD groups (both P < 0.01) and early PD patients outperformed moderate PD patients (P ≤ 0.01).

In the whole sample, the RME performance correlated with age (r = −402; P = 0.001), education (r = 0.274; P ≤ 0.05), MMSE (r = 0.422; P ≤ 0.001), FAB (r = 0.552; P = 0.001) and BDI (r = −0.342; P = 0.005). In HC, the RME performance did not correlate with any characteristic. Within PD, the RME performance correlated with age (r = −0.633; P ≤ 0.01), MMSE (r = 0.563; P ≤ 0.01) and FAB (r = 0.522; P ≤ 0.01). Considering these correlations between RME performance and other demographic and characteristics, we had to verify if RME differences between HC and PD and between moderate PD patients and early PD patients were significant, also controlling for potential covariates. We performed two ancova with the RME as dependent variables and other variables (age, education, MMSE, FAB, BDI) as covariates. Also controlling for these variables, differences between RME performances of HC and PD were statistically significant (F(1,60) = 4.549; P = 0.037), while the difference between moderate PD patients and early PD patients was not (F(1,25) = 0.986; P = 0.330).


In the current study we investigated affective ToM with the RME in PD patients. Two main findings emerged. The first finding is that PD patients at early and moderate PD stages (with at least 6.31 ± 4.02 years of disease duration; H&Y range 1–3) were impaired in comparison to HC. This is in line with previous findings,[16, 17] although preserved RME performances have been reported in early and moderate PD stages.[18-21] Different RME versions adopted in previous studies could have produced this heterogeneous pattern of empirical findings; in this study we adopted the full 36-item version, while most previous studies adopted shorter RME versions, maybe hampering the detection of mild or sub-threshold impairments. The impairment of PD patients in this task cannot be due to sociodemographic (age and education) and affective characteristics (depression) or primary cognitive impairments (as detected by MMSE and FAB) considering that the difference in comparison with HC was significant when also controlling for these variables. The impairment also cannot be attributed to dopaminergic treatment, whose dose was not correlated with RME performance, as previously reported.[2, 6, 21]

Second, moderate PD patients had poorer RME performances in comparison to early PD patients, but this difference did not reach statistical significance when variables correlated with RME were controlled for. Indeed, moderate PD patients were older and had worse MMSE and FAB performances. Therefore, on the basis of these data we cannot exactly establish the role of PD progression on RME performance: further studies assessing PD patients with at least the same ages but different PD stages are needed to better understand this issue.


In conclusion this study found that affective ToM, as assessed by the RME, may be impaired from the early PD stages, with at least 5 years of mean disease duration. This finding needs further empirical confirmation, and longitudinal studies are needed to investigate the impact of PD progression on affective ToM ability in this clinical population.


There are no conflicts of interest. Full financial disclosure: U.B. is on the advisory boards of GSK, Lundbeck, Novartis, and UCB; U.B. received honoraria for speeches at meetings of Boehringer Ingelheim, GSK and Novartis.