Abstract
Keywords
INTRODUCTION
Creativity is a complex and multidimensional phenomenon that is difficult to place in the world of research. A broadly accepted definition of creativity is the ability to generate ideas that are both novel and useful in a particular social setting [1–4]. This definition places creativity within a social and cultural context, distinguishing it from strangeness. An approach to the assessment of creativity is the measurement of divergent thinking, which includes various aspects, such as fluency, flexibility, originality and elaboration [5].
Some of the research papers on creative cognition have focused their attention on illness associated with altered creative output. In the last decade, several studies have reported the emergence or the modification of creative products in patients with dementia [6, 7], subarachnoid hemorrhage [8], brain injury [9] and epilepsy [10]. A further opportunity to examine this phenomenon has emerged from studies on the effects of psychotropic substances on creative processes [11]. In fact, creative drive, which is required to generate creative ideas and is influenced by mesolimbic dopamine, may be altered in drug users or during medication intake [12].
Artistic-like production may emerge or change in patients with Parkinson’s disease (PD) receiving dopaminergic therapy (DT) [13–18]. Although an increase in artistic-like production has been associated with compulsive and repetitive behaviours [19], recent data suggest that it could be due to the emergence of artistic-like skills triggered by DT [20]. Indeed, PD patients on dopaminergic treatment seem to show increased artistic-like production, although they do not differ from healthy controls in terms of creative thinking. However, artistic-like production and creative thinking have never been described in PD patients who are professional artists or in patients who have artistic hobbies. Since creative thinking is supposed to be more pronounced in artists than in the general population [21], the principal aim of this study was to characterize creativity features in PD patients and healthy controls who are also professional artists, in order to evaluate whether or not artistic production and creative thinking is influenced by DT and whether it is associated with compulsive behaviors and/or linked to the emergence of artistic-like skills triggered by DT.
MATERIALA AND METHODS
Ethics
The study was performed in agreement with the principles of the Declaration of Helsinki and the protocol was approved by the local Ethics Committee. We obtained written informed consent from all participants.
Participants
Three groups of PD out-patients consecutively attending our Institute were included: professional artists before the onset of PD (PD-A; N = 12), patients who were not professional artists before the onset of the disease and developed artistic output after the onset of PD and the introduction of DT (PD-C; N = 12; mean time [±SD] to onset of artistic skills after the introduction of dopaminergic therapy, 20±6 months) and patients who never showed artistic-like production (PD-NC; N = 12). All patient groups were matched (1:1:1) for gender, age (±1 year) and disease duration (±1 year). Then, we included two groups of healthy controls matched to patients for gender and age at assessment (±1 year): professional artists(HC-A; N = 12) and subjects who never presented creative output (HC-NC; N = 12). Professional artists (both PD or HC) were designers, architects, stylists, painters or photographers. HC were selected from the general population and none were relatives or spouses of patients.
Inclusion criteria for patients were: idiopathic PD according to UK Brain Bank criteria [22]; Mini Mental State Examination (MMSE) >24 [23, 24]; stable DT (including the use of a dopamine agonist) for at least 8 weeks before inclusion in the study. Patients treated with either atypical (i.e. clozapine and quetiapine) or typical antipsychotics, lithium and mood stabilizer drugs or treated with advanced-stage therapies (deep brain stimulation, continuous levodopa duodenal infusion and apomorphine subcutaneous infusion) were excluded. Selective serotonin reuptake inhibitors (SSRI) users were eligible of inclusion.
We considered the age at onset of PD as the age of occurrence of the first cardinal symptom. PD motor laterality was defined as having predominant right or left motor symptoms by means of the Unified Parkinson’s Disease Rating Scale-motor score (UPDRS-III) [25]. Side of onset of motor symptoms was included in PD data collection as well as handedness in PD and HC [26]. Tremor (TD) and non-tremor dominant (NTD) subtypes were also considered and we used the Lewis subtype definition [27].
Assessment
Neurological assessment was performed in the morning and in the best “on” conditions by means of UPDRS-III and the Hoehn and Yahr (H&Y) staging system [28]. LEDD was also calculated using the formula provided by Tomlinson et al. [29].
Creativity
Creative thinking was evaluated by means of the Abbreviated Torrance Test for Adults (ATTA) [30]. Testing was conducted in the best “on” phase in PD groups and during the morning hours in all the study groups. The test is used to investigate four factors of creative thinking: Flexibility (a kind of thinking that allows an exchange of ideas and strategies, and the ability to pass from one scheme to another, from one category to another), Fluency (the ability to conceive many ideas and hypotheses without focusing on their quality, which is potentially useful in problem solving), Originality (the capacity to find unusual and rare answers) and Elaboration (the ability to further develop a concept by adding new elements) [31, 32].
Neuropsychological evaluation
It was performed in “on” phase by two trained neuropsychologists with expertise in movement disorders and blinded to sub-group status. It included: global cognitive evaluation: MMSE (inclusion criteria), verbal phonemic and semantic fluency test [33], Frontal Assessment Battery (FAB)[34]. mood assessment: Geriatric Depression Scale (GDS) [35], Hamilton Anxiety Scale (HAM-A) [36], behavioral evaluation: modified Minnesota Impulsive Disorders Interview (mMIDI), focusing on punding (a complex, purposeless and stereotyped behaviour) subscore [37]. Besides, patients were assessed for the presence of any addictive disorder, including pathological gambling, binge eating, hyper-sexuality, compulsive shopping, internet addiction and dopamine dysregulation syndrome [38, 39]. In order to confirm the diagnosis of impulse control disorders (ICDs) according to established diagnostic criteria [40–42] the results from extensive neuropsychological testing were combined with a semi-structured interview by experienced neuropsychologists. Considering that patients with ICDs (ICD+) may underestimate or deny any behavioural disturbance, also caregivers were assessed by means of mMIDI. The overall assessment lasted 45 minutes in PDgroups.
Statistical analysis
Data were reported as mean and standard deviation. Between-group comparisons of categorical and continuous variables were performed by Chi-Square test and ANOVA (one-way) analysis, respectively. Post-hoc analyses were also considered and multiple-group comparisons were performed considering the Bonferroni correction. Finally, correlations with ATTA scores (in the whole study population or in PD patients as specified) were evaluated by means of Spearman’s Rho test. All the analyses were performed with the software STATISTIX 7.0 (Analytical Software, Tallahassee, FL, USA). A p-value <0.05 was considered statistically significant.
RESULTS
Demographic and clinical features of study groups are presented in Table 1.
Handedness (p = 1.00) and education (p = 0.39) were similar in patients and controls. Disease severity was similar in PD groups (UPDR-PartIII, p = 0.73; H&Y, p = 0.18). However, prevalence of TD phenotype was higher in PD-A and PD-C than PD-NC (p < 0.001). Although three patients in both PD-A and PD-NC groups were treated with dopamine agonist mono-therapy, no significant differences in total LEDD and LEDD from dopamine agonist treatment were found among PD groups (p = 0.21 and p = 0.43, respectively). Table 2 shows results of tests for neuropsychological and mood profiles, which were similar in all PD and HC groups, with the exception of a higher and lower HAM-A anxiety score in PD-NC and HC-NC, respectively.
Moreover, only PD patients had a positive mMIDI test, while HC did not. Five patients in each PD group showed behavioral compulsions (single or combined) while combined ICDs were found in two PD-A patients, in four PD-C patients and in two PD-NC patients. mMIDI test performed by caregivers confirmed the results obtained in PD groups.
Punding sub-score measured by means of mMIDI was negative in all study groups.
ATTA total score was significantly higher in HC-A and PD-A as compared to the other study groups (p < 0.001). Moreover, although PD-NC showed the lowest ATTA total score, this was not different as compared to HC-NC. However, when ATTA sub-scores were considered, we found that those for elaboration and originality in PD-A, PD-C and HC-A were higher (p < 0.001), while fluency and flexibility showed a trend towards lower values in NC groups(Table 3).
Finally, in PD patient population ATTA scores were not significantly correlated with LEDDs (p > 0.40 for all groups) or with duration of DT (p > 0.35 for all). Moreover, ATTA scores were not correlated with mMIDI results in PD subjects (P > 0.50 for all) or in the whole study population (P > 0.55 for all groups). Furthermore, we evaluated ATTA total scores by ICD (ICD+/ICD–) in all PD groups. Our results suggested that ICDs are not associated with creative thinking (mean [SD], total ATTA score: ICD+, 73.2 [21.1]; ICD-, 67.5 [23.6]; p = 0.46). An example of artistic-like production is presented in Figs. 1 and 2.
DISCUSSION
Enhanced artistic drive is a distinctive everyday feature of PD-A patients and the occurrence is not spontaneously reported as a novelty as in PD-C.
No univocal interpretation of results is possible. In our study, we found that creative thinking, as assessed by means of the ATTA, does not appear to be related to DT. This observation is primarily supported by similar ATTA scores in PD-A and HC-A groups, although only PD population was on DT. ATTA scores in artists (both PD and HC) resulted to be significantly higher than in the other groups. Indeed, PD-C revealed a de-novo artistic production during dopaminergic exposure, but ATTA scores were significantly lower than in artist groups. Our data suggest that DT may trigger artist-like drive in non-professional artists, but also that it is irrelevant in divergent thinking measured by means of ATTA. These results support also the concept that creative thinking and artist-like production are different processes. Besides, ATTA sub-scores for elaboration and originality appear to be useful to distinguish professional from non-professional artists in terms of creative thinking, independently of PD. Although several components should be taken into account in order to investigate the creative process, the evaluation of thinking subcomponents, such as fluency, flexibility, originality and elaboration, is currently considered a valid process to assess creativity [43]. Finally, ATTA score in PD-C was higher than in non-creative groups (PD and HC). The improvement in ATTA score in patients reporting de novo artistic production may be a consequence of repetitive artistic activities - a phenomenon that is observed with cognitive stimulation [44]. On the other hand, the data suggest that PD-C are “intermediate” creativity level patients and that DT influences creative thinking to some extent in a subgroup of PD patients. With this hypothesis, we acknowledge that assessment of patients through ATTA test during the best “on” medication probably is a limitation. It would be interesting to test patients during “on” and “off” DT in order to evaluate whether creative thinking is substantially influenced by DT in a subset of PD patients (PD-C). To partly cope with this limitation, we included patients on stable treatment. However, we should recognize that also changes in DT may contribute with higher creativity occurring during dose increase. We have observations pointing out on the misuse of DT to enhance creative output in PD patients who were creative before the onset of the disease [45] but there are also case-report describing no significant changes in artistic production after suspension of DT [46].
Contrary to our previous study [20] we used ATTA in place of Torrance Test of Creative Thinking since the former requires a shorter time of administration, thus reducing the risk of testing patients in different conditions of medication efficacy.
The present study also confirms our previous results, which showed that creative cognition is independent of ICDs [20]. In fact, in our sample we observed that impulsive disorders occurred in all PD sub-groups, but not in all PD patients, i.e. the % of patients without ICDs was 58 in PD-C and PD-NC and 66 in PD-A. Indeed, although previous studies have reported heterogeneous estimates (0.3% –37%) [19], the prevalence of ICDs in our PD population was quite high. This might be due to some risk factors for ICDs, such as male gender, young age, young age at onset of PD, observed in our study population. Likewise, preserved cognitive status might be an inclusion bias [47]. For example, Siri et al. showed that PD with pathological gambling (PG) performed better in some cognitive skills as compared to PD with no PG. On the other hand, the prevalence of ICDs was similar in the different subgroups and it is unlikely that this potential inclusion bias has weakened the value of our findings. However, we cannot fully exclude that the enhanced drive to produce artistic work may mimic compulsive behaviour in some PD patients on DT [15, 16]. Furthermore, the enhanced artistic drive in PD-C does not seem to be related to punding-like activities as suggested by the mMIDI punding subscore. Punding is reported as a side effect of DT, but none of our patients reported dopamine dysregulation syndrome [48] or punding-related activities, such as cleaning, tidying, gardening, collecting, sorting [49].
GDS score was in the normal range in all groups and only a few patients were on antidepressant theraphy. This result is consistent with the observation that severe depression and manic state are too debilitating to facilitate the production of new and useful ideas, and that creative work is done during periods of increased energy, possibly in-between episodes of mood deflexion [12, 50].
Finally, it was interesting to note a significantly higher prevalence of TD phenotype in PD-A and PD-C (both 50%) than in PD-NC (8%). This phenotype is usually linked to better cognitive performance and less likely to develop dementia as compared to the other phenotypes [51, 52].
Although our results suggest that DT may be related to increased artistic productivity and not to enhanced creative thinking, other studies seem to highlights a role of DT in creativity. Faust-Socher et al. [13] have recently reported enhanced verbal and visual creativity in PD patients as compared to HC, suggesting that DT may act through the reduction of latent inhibition. On the other hand, no relationship was found with ICDs. Altogether, these findings highlight that further research on this issue is needed.
Some limitations of our study should be borne in mind. The first refers to the methodology used, as we assessed creativity through the measurement of divergent thinking by means of ATTA. In addition, artistic-like skills might be part of a personality trait and might be influenced by the socio-cultural environment. Although subjects live in an urban contest and most of them completed high school, we recognize that personality tests would provide useful information on the role of internal factors. The cross-sectional study design limits the inference of a cause-effect relationship and prospective studies are required in this area. This would enable identifying also independent predictors of emerging creativity (e.g. age, etc.). We are also aware that the small sample size of our study is a limitation, partly due to the paucity of PD-A. In respect with this, despite the use of several matching rules and restrictive inclusion criteria, particularly the use of anti-dopaminergic treatment, lithium or other mood stabilizers [11], selection bias are probably inevitable.
CONCLUSION
Creative thinking in PD, as assessed by means of ATTA, not only appears to be unrelated to dopaminergic treatment, unlike creative output, but it is already at the highest values in professional artists, independently of PD. Since all PD groups (professional artists and patients with/without artistic-like production) had comparable dopaminergic medications, our results do not support a direct relationship between antiparkinsonian drug dose and increased creativity in the “on” phase. As DT may increase the creative drive, further studies in “on” and “off” medication as well as in drug naïve PD patients are needed to clarify this issue. We believe that, in order to clarify the role of DT, future studies should investigate creative thinking in drug naïve PD patients.
CONFLICTS OF INTEREST
There are no conflicts of interest.