1 Background

Neurodegenerative diseases can manifest with neuropsychiatric symptoms, which include, for example, loss of impulse control and foresight, disinterest, and aggressive behavior [1-4]. Changes in behavior may be among the first manifestations of a neurodegenerative disease and pose a great burden not only to the patients themselves but also to society and the legal system [5, 6].

Especially in behavioral variant frontotemporal dementia (bvFTD), neuropsychiatric and behavioral symptoms are common [7, 8]. It has been proposed that these symptoms can predispose an individual to criminal behavior [9]. Previous studies [10-17] and case reports [18] have described increased rates of antisocial and criminal behavior in patients with FTD. However, in these studies, the evaluation has mostly been based on caregivers' interviews or examination of the participants' patient records, and only one of them has included a control group of participants without a neurodegenerative disease [16]. Furthermore, previous knowledge on the crime rates in various other neurodegenerative diseases is limited.

The aim of this case–control study was to examine the frequency and the types of criminal acts in a large cohort of patients consisting of all common neurodegenerative diseases and compare these to matched controls without a neurodegenerative disease. We consider the study ethically justifiable as so far, criminal behavior in these patient groups has not been thoroughly studied and the lack of knowledge can weaken the legal protection of these patients.

2 Methods

3 Results

3.3 Criminal Behavior in FTD

Due to the observed differences in the crime rates, especially in the FTD group (N = 297), subanalyses were conducted. In the FTD group, 133 patients (44.8%) were women and 164 (55.2%) were men. The mean age at diagnosis was 61.3 years. The yearly differences in criminal rates between FTD patients and controls are demonstrated in Figure 1.

Next, we compared the types of crimes in the FTD group, divided as follows: (1) violence, (2) crimes against property, (3) drunk driving, and (4) traffic offenses. We calculated the frequency of all types of crimes during 10 and five calendar years preceding the year of diagnosis and during the year of diagnosis, separately in patients aged 69 years and younger and under 60 years (at diagnosis).

Overall, rates of violent crimes were low among FTD patients during the whole study period, with no statistically significant differences compared to control participants.

Crimes against property were over twice as frequent in the FTD group compared to the control group during the 10-year period preceding the year before the diagnosis (5.4% in FTD patients aged 69 years and younger vs. 2.4% in the control group (p = 0.003)). In the group of participants under 60 years, the numbers were 8.4% for FTD and 3.0% for controls (p = 0.004). During the 5 years preceding the year before the diagnosis, property crimes appeared more common in the FTD group (2.7%) compared to the control group (1.4%); however, this was not statistically significant (p = 0.077). In individuals < 60 years of age, the numbers were 5.6% for FTD and 1.8% for controls, and the difference between these groups was statistically significant (p = 0.009). During the year before the diagnosis, property crimes were more common in the FTD group (1.3%) than in the control group (0.2%) (p = 0.002). Comparing participants under 60 years, the numbers were 2.8% in the FTD group and 0.4% in the control group (p = 0.002).

During the 10-year period preceding the year before the diagnosis, drunk driving was more common in the FTD group (4.7% of the FTD patients compared to 2.7% in the control group, p = 0.048). The difference was even more significant when comparing participants under 60 years, with percentages of 7.5% for FTD and 2.9% for controls (p = 0.012).

Traffic violations were almost equally common between FTD patients and controls during the 10 and five calendar years preceding the year before the diagnosis, although during the calendar year before the diagnosis, FTD patients showed higher rates of traffic offenses, indicated by 7.7% in the FTD group and 5.0% in the control group (p = 0.049). Among individuals aged < 60 years, 11.2% of FTD patients and 6.2% of controls had committed traffic offenses (p = 0.046) during the year before the diagnosis.

Results for the prevalence of different crime subtypes in FTD patients and control participants are listed in Table 3.

TABLE 3. Prevalence of different crimes among all FTD patients during 10 and five calendar years preceding the year of diagnosis and during the calendar year of diagnosis compared to matched controls.

	10 years FTD %, n (control %, n) p	5 years FTD %, n (control %, n) p	1 year FTD %, n (control %, n) p
Violence
	2.7% n = 8 (2.2%, n = 64) p = 0.550, NS	1.0%, n = 3 (0.9%, n = 27) p = 0.864, NS	N/A
Property
	5.4%, n = 16 (2.4%, n = 72) p = 0.003	2.7%, n = 8 (1.4%, n = 41) p = 0.077, NS	1.3%, n = 4 (0.2%, n = 7) p = 0.002
Drunk driving
	4.7%, n = 14 (2.7%, n = 80) p = 0.048	2.4%, n = 7 (1.7%, n = 51) p = 0.429, NS	N/A
Traffic violations
	37.4%, n = 111 (37.0%, n = 1096) p = 0.893, NS	26.6%, n = 79 (24.0%, n = 715) p = 0.317, NS	7.7%, n = 23 (5.0%, n = 150) p = 0.049

• Note: It was not possible to calculate the rates of violent crimes and drunk driving during the 1 year before diagnosis due to low n. Statistically significant differences are bolded.
• Abbreviations: FTD , frontotemporal dementia; n, number of cases; N/A, not applicable; NS, not significant.

4 Discussion

Here, we have evaluated the prevalence of criminal behavior among a substantially large group of patients with neurodegenerative diseases. Crime rates were found to be higher, especially in FTD patients than in controls several years before the diagnosis. Interestingly, crime rates appear to substantially decrease shortly after the diagnosis, also in the FTD group. To our knowledge, this is one of the most comprehensive studies to examine criminal behavior in these disease groups, including data spanning over a decade before and 3 years after the diagnosis, complete criminal record data as well as a remarkable number of adjusted controls. We used objective data from Statistics Finland and utilized nonselective population data as a control group.

Our results considering FTD patients are in line with previous reports [12, 15, 26, 27]. We separately analyzed the 10 and five calendar years preceding the year before the diagnosis as well as the year before the diagnosis. Patients with FTD committed significantly more criminal offenses, especially during the year before the diagnosis. Some criminal acts were more common even during the period of 10 years prior to the diagnosis year.

The reasons underpinning increased rates of criminal behavior among FTD patients are most likely multifactorial. Diehl-Schmid and colleagues interviewed some of the patients with a history of criminal behavior. Most of them did not feel the need to explain the reasons behind their criminal acts, and none of the patients showed guilt, shame, or remorse [11]. Similarly, Mendez and colleagues described cases of four patients with FTD who had committed criminal acts [18]. All of them understood the wrongfulness and consequences of their actions, yet were not able to refrain themselves from committing criminal acts. Particularly, the frontal lobes are anatomical structures involved in controlling behavior, and as FTD causes progressive damage to these brain areas, it is not surprising that these patients often show inappropriate or even criminal behavior. Disturbances, especially in the prefrontal cortex, have been suggested to underlie these behaviors [18]. Also, other studies support this notion, as emotional dysfunction, deficits in decision-making, and morally improper behavior have been observed in patients with lesions in this area [28-32]. Interestingly, crime rates were lower specifically in patients with PD already several years before diagnosis. This finding seems plausible from a neurobiological perspective, considering the reduced dopaminergic activity in PD, as conversely, increased dopaminergic transmission, induced by, for example, alcohol or cocaine use, is known to increase impulsivity [33].

To our knowledge, there are only two other studies that have previously used nationwide data on police-recorded crimes as a source to assess criminal acts in FTD patients. Talaslahti and colleagues investigated the time period of 4 years before the diagnosis and reported that criminal behavior was more frequent among the group of FTD patients, with 94% of the crimes being traffic offenses or crimes against property [26], but the clinical diagnoses were purely register-based and not re-assessed and validated. Another study in a small cohort from Germany reported a history of criminal behavior in over half (54%) of bvFTD patients and in 56% of svPPA patients, but only in 12% of patients with AD [11]. Although the evaluation of criminal behavior was based on caregiver interviews, the findings are similar to ours as mainly minor (property) crimes and drunk driving were more common in patients with FTD, whereas violent crimes were not associated with FTD or other neurodegenerative diseases.

A previous study also examined crime rates after the diagnosis of a neurodegenerative disease, showing that the number of criminal offenses may be diminished even up to 50% compared to the general population [34]. The results in our present study were parallel, indicating substantially decreased crime rates after diagnosis. This might be explained by the medical surveillance with regular follow-up visits, by the use of symptomatic medication, or closer surveillance by a caregiver [27]. This underlines the importance of early and accurate neurodegenerative disease diagnosis and care. Another possible explanation could be that the patient's condition worsens to the state in which committing crimes may not be physically and/or cognitively possible. Interestingly, one study showed that FTD patients with a reported history of criminality had higher MMSE scores than the ones without, and the disease severity was classified as mild in nearly all patients with criminal acts [15]. This suggests that increased criminality in FTD is emphasized rather as an early symptom of the disease than that of the later dementia phase. Moreover, it has been debated whether there should be a system that would enable examining the possibility of a somatic reason behind the sudden criminal behavior for first-time offenders older than 55–60 years and excluding diseases damaging especially the frontal lobes [11]. Currently, there are no common medical criteria to assess criminal responsibility for people with FTD or other dementias, and the classification of “criminal responsibility” varies between countries. This study raises the question of whether specific criteria should be developed since the phenomenon of criminal behavior in prodromal neurodegenerative disease exists.

The strengths of this study include the comprehensive, well-characterized study groups, including all major neurodegenerative diseases, and the inclusion of a powerful 1:10 case-to-control ratio of nonselected control participants without a diagnosis of a neurodegenerative disease representing the general population. Furthermore, we utilized objective and complete data from Statistics Finland, which leaves only a small chance for reporting bias. All the diagnoses were made in expert-level memory outpatient clinics by experienced neurologists specialized in neurodegenerative diseases, according to the latest diagnostic criteria. All patients were validated by thorough retrospective re-evaluation, substantially increasing the specificity of the diagnoses compared to conventional register-based approaches. The time period reaching 10 years prior to the diagnosis offers insights into the very early (prodromal) phase of the disease. A limitation of this study is the small size of some of the subcohorts, which reduces the statistical power of part of the analyses. Furthermore, although the diagnoses were made in special-level clinics and ascertained with retrospective evaluation of patient records, there is a minor but possible chance for individual misdiagnoses.

In conclusion, this study suggests increased overall criminal rates, especially in patients with FTD during the last year before the diagnosis. Significant differences in drunk driving and property crimes were detectable already 10 years before the diagnosis, proposing that criminal behavior is an early symptom of FTD. Sudden criminal behavior in previously law-abiding individuals during the middle or elderly years of life can be a symptom of a neurodegenerative disease. Therefore, possible organic causes, especially brain disorders damaging the frontal areas of the brain, should be considered as potential underlying reasons. Early and accurate diagnosis may result in decreased criminal behavior, reducing both individual and societal burdens.

Author Contributions

Helmi Soppela: conceptualization, investigation, writing – original draft, validation, writing – review and editing. Mikko Aaltonen: conceptualization, investigation, methodology, formal analysis, writing – review and editing. Kalle Aho: data curation, writing – review and editing, resources. Sami Heikkinen: writing – review and editing, investigation. Ave Kivisild: data curation, resources, writing – review and editing. Adolfina Lehtonen: writing – review and editing, data curation, resources. Laura Leppänen: writing – review and editing, data curation, resources. Anna Mäki-Petäjä-Leinonen: validation, writing – review and editing. Iina Rinnankoski: data curation, resources, writing – review and editing. Laura Tervonen: writing – review and editing, data curation, resources. Jari Tiihonen: validation, writing – review and editing. Markku Lähteenvuo: validation, writing – review and editing. Annakaisa Haapasalo: validation, writing – review and editing, investigation. Anne M. Portaankorva: validation, writing – review and editing, supervision. Päivi Hartikainen: validation, writing – review and editing. Kasper Katisko: validation, writing – review and editing, methodology, conceptualization, funding acquisition, investigation, writing – original draft, supervision. Johanna Krüger: conceptualization, investigation, funding acquisition, methodology, validation, writing – review and editing, supervision. Eino Solje: conceptualization, investigation, funding acquisition, methodology, validation, writing – review and editing, project administration, supervision, resources.

Consent

Informed consent was not necessary.

Conflicts of Interest

E. Solje has served on the advisory board of Novartis, EISAI, and Roche Oy, served as a consultant for Novo Nordisk, and received honoraria for lectures from Lundbeck and Roche. J. Krüger has served on the advisory board of Novartis, Nutricia, EISAI, and Roche Oy, and received support for congress participation from Merck. J. Tiihonen has participated in a research project funded by Janssen-Cilag to his employing institution, has received lecture fees from Janssen-Cilag, Lundbeck, and Otsuka, and has served as a consultant to Healthcare Global Village, HLS Therapeutics, Janssen-Cilag, Orion, Teva, and WebMed Global.