The main goals of this study were to assess the nature and prevalence of psychological and behavioral symptoms in a Belgian sample of people with and without dementia and to compare prevalence of behavioral and psychological symptoms between both populations. Within the dementia group, depression, anxiety, and delusions constituted the most frequent psychological symptoms (from 20% to 30%) (). Elation was less common (<12%). Regarding behavioral symptoms, agitation/aggression (34.1%), apathy (35%), and irritability (30%) were reported among a third of people with dementia. Prevalence rates for other NPI symptoms varied from 10% to 23%. Within the group without dementia, depression was the most common psychological symptom (19.5%) (). Prevalence rates of anxiety, delusions, and elation were between 6% and 9%. Hallucinations were infrequent (<5%). Regarding behavioral symptoms, irritability and agitation/aggression were the most common symptoms (15%). Prevalence rates for disinhibition and apathy were approximately 10%. Aberrant motor behavior and sleeping and eating disorders were rare (close to 5%). Comparative analysis showed that prevalence rates of depression, elation, irritability, disinhibition, and sleeping disorders did not differ significantly between groups. Delusions, hallucinations, anxiety, apathy, agitation/aggression, aberrant motor behavior, and eating disorders were significantly more frequent in the dementia group. In the dementia group, only the frequencies of apathy, irritability, and sleeping disorders were higher. Regarding severity criteria, the scores were higher in the dementia group for apathy and irritability.
Within the dementia group, data distribution is similar to that in previous studies
15–
18 (). However, we observed quantitative differences between our data and those from Hollingworth et al
17 and Peters et al,
18 whose prevalence rates were much higher than ours (). The sample included in the study by Hollingworth et al
17 included 1120 individuals “diagnosed with late-onset probable AD [Alzheimer’s disease].” Their results are higher than those from other NPI studies.
15–
18 The group of persons with dementia in Peters et al
18 included 576 subjects whose average MMSE score was clearly higher, and whose average age was lower, than our group of patients with dementia. This elicits a question of possible evaluator-related bias. Indeed, it is possible that some psychological symptoms are more easily detectable among people presenting with minor cognitive disorders than among those presenting with severe cognitive disorders. This hypothesis has not been investigated. Our high percentage of institutionalized dementia subjects could also explain these differences. Using the Behave-AD,
20 the MFS,
21 and the CMAI,
7 Engelborghs et al
19 noted the importance of aggressiveness, activity disturbance, and disinhibition in dementia semiology. Their prevalence rates were also higher than those presented here. The variability between this study and our current investigation is likely related to methodological differences, as well as variation in life conditions, of the subjects (eg, home, institution, hospital).
Regarding the group without dementia, data distribution is similar to that in previous studies
12,
15,
18 (). However, Peters et al
18 reported higher prevalence rates than those shown in previous studies, except for elation and delusions (). Their sample contained 193 CIND subjects taken from “a Canadian cohort study of cognitive impairment and related dementias.”
18 According to these authors, an explanation for these high rates may have emerged within their definition of CIND. In another study, Lyketsos et al
12 studied 673 elderly subjects without dementia who were living at home or in an institution. Their observed prevalence rates were much lower than those in our study (). This variability is likely related to various factors, such as demographic and clinical characteristics and size, as well as life conditions of the evaluated populations. Moreover, the changing nature of the examined symptoms is likely to explain part of this variability.
Regarding comparative analysis, Lyketsos et al
12 showed that all NPI symptoms were significantly more frequent in their dementia group than in their non-dementia group, except for sleeping and eating disorders, which were not assessed. The heterogeneity of reported data may be due to the use of distinct procedures for subject selection and to the nature of the assessed population. Despite the fact that our results are roughly similar to those of Lyketsos et al,
15 this study also found that all NPI symptoms are significantly more frequent in their dementia group than in their MCI group, except for elation. This may be explained by the difference between the size of our group without dementia and their MCI group. Peters et al
18 found similar results to ours except for anxiety (no significant difference) and disinhibition.
This comparative analysis suggests a categorization of NPI symptoms. Some disorder prevalence rates did not significantly differ between groups (ie, depression, elation, irritability, disinhibition, and sleeping disorders). This report suggests that these symptoms are nonspecific to dementia. Prevalence rates of others NPI symptoms (ie, delusions, hallucinations, anxiety, agitation/aggression, apathy, aberrant motor behavior, and eating disorders) are significantly higher in the dementia group. Among these symptoms, some are rarely reported in the group without dementia (ie, hallucinations, delusions, aberrant motor behavior, and sleeping and eating disorders), with prevalence rates close to 5%. As a result, these symptoms can be considered specific to dementia, with prevalence rates ranging between 15% and 25% of the subjects with dementia. The other symptoms are more frequent in the dementia group, but have been regularly reported in the group without dementia with prevalence rates close to 10%. These symptoms are not specific to the dementia group. Nevertheless, it is important to note that some studies have reported prevalence rates of hallucinations () and aberrant motor behavior () as being lower than 5% among persons without dementia. Data concerning eating disorders are more heterogeneous.
Comparative analysis allows us to suggest a distinction between three categories of symptoms: dementia-specific symptoms, symptoms more frequent in dementia, and symptoms nonspecific to dementia. This distinction may be useful when examining etiological factors. Indeed, it may be hypothesized that dementia-specific symptoms are very likely associated with neurobiological factors. However, symptoms that are more frequent in dementia and symptoms that are nonspecific to dementia may be related to a variety of environmental, psychological, neurobiological, and somatic factors, the respective weights of which may be variable from one disorder to another and from one subject to another. The multifactorial etiology of some psychological and behavioral symptoms should have a direct impact on the nature of the interventions on patients with dementia and their families.
Thus, based on these hypotheses, a neurobiological approach would be more appropriate when examining the etiological basis of delusions, hallucinations, aberrant motor behavior, and sleeping and eating disorders. Conversely, a biopsychosocial approach would be more advisable to study depression, irritability, elation, sleeping disorders, anxiety, disinhibition, agitation/aggression, and apathy. The statement that these symptoms are often observed among subjects with other severe somatic pathology
30–
33 is consistent with the multifactor causality hypotheses. However, since these hypotheses are inconsistent with those of previous studies,
34–
36 they must be verified by further research. More precise comparison of the symptomatology of disorders between people with and without dementia should give some indication of whether our hypothesis is correct.
Limitations of the study
The small size of the non-dementia group constitutes a primary limitation of our study. Second, inclusion criteria in both groups were not sufficiently precise. Third, we did not differentiate dementia by type in D group. Fourth, neither group was homogeneous regarding age and sex; additional analyses have been conducted to control for the influence of these factors on the presence/absence of NPI symptoms; the results supported our hypothesis. Fifth, subjects’ pharmacological treatment was not controlled: some substances could have induced psychological and behavioral disorders; other substances can reduce, alleviate, or eliminate some symptoms, and the influence of pharmacotherapy on psychological and behavioral symptoms should be the subject of further studies. Sixth, information concerning the durations of dementia had not been collected.
The NPI assessment tools also had some limitations: information obtained from an informant may have been influenced by his/her expectations and by his/her capacity to deal with critical situations. Patient–proxy assessment is the subject of debate in the literature.
37,
38 Thus, the results must be interpreted according to the definition of symptoms induced by the NPI. Indeed, some NPI domains include either specific symptoms such as hallucinations or highly heterogeneous syndromes such as aberrant motor behavior. Aberrant motor behavior includes various symptoms, such as intrusive behavior,
39 vocally disruptive behavior,
40 dyskinesia,
41 and ritualized behavior,
42 all of which involve different etiological factors.