The prevalence of NPSs found in our study was very high, tending to increase in the more advanced stages of dementia that present worse functionality and presenting changes in the distribution of different NPSs based on disease severity.
According to their caregivers, almost 100% of the patients presented some neuropsychiatric symptoms, and approximately 85% had at least one clinically significant symptom, that is, an NPI ≥ 4. The prevalence results obtained in our study in noninstitutionalized patients are similar to those reported by other studies carried out in specialized clinics, with prevalence rates exceeding 90% (8–11), but higher than those reported for community-dwelling patients, with an overall NPS prevalence of 50–85% (5, 6, 12–14) and significant symptom prevalence of 40–67% (4, 12, 18, 41). The higher prevalence of NPSs found can be explained because patients who had a previous diagnosis of dementia were included, unlike other studies that recruited patients by screening for dementia in the general population, possibly leading to a higher percentage of cases of mild dementia (5, 12).
Our patients presented an average of five neuropsychiatric symptoms. Studies that have used the same 12-item version of the NPI have reported similar findings (6, 9). Apathy was the most frequent NPS, appearing in 70% of patients and being clinically relevant in almost 40%. These data are consistent with those of other studies, with prevalence rates between 74 and 76% (9, 42, 43). Apathy stands out as the most common symptom in most publications (8–10, 14, 15), but depression (7, 44–46), sleep (12) and appetite disturbances (13) have also been described as the most prevalent. These differences may be related to the baseline characteristics of the study population. Thus, depression predominated when there was a high percentage of mild (7, 45, 46) or moderate (44) dementia in patients. Although apathy was the most frequent symptom, agitation, sleep disorders and hallucinations were the most intense and had the greatest weight in the overall NPI score due to the impact they had on the patient and their caregivers.
The average total NPI score was 25, without differences based on sex or age group, as was the case in other publications (12, 46). We also did not find differences based on level of education or level of dependency. Regarding coexistence, the NPI score was higher for patients who lived with a professional caregiver than for those who lived alone or with family (partner or other relatives). Although this association was not statistically significant, it reflects the reality of having to hire outside caregivers for individuals with a greater intensity of symptoms. The average NPI score was very similar to the scores reported in studies by Aalten et al. 2007 (2) (mean NPI 23) and García-Alberca et al. 2008 (9) (mean NPI 27.9); in contrast, the mean score obtained herein was higher than the 12–15 points obtained by other authors, a difference that could be explained by different characteristics of the study populations, with a predominance of mild dementia, or the use of a 10-symptom version of the NPI instead of the NPI-12 (27, 44, 46, 47).
The NPI score (NPS intensity) tended to increase as the disease progressed, a relationship already described in other studies (15, 18, 48, 49). In our study, an increase of 7.6 points in the NPI was demonstrated between mild and advanced dementia after adjusting for sex, age and clinical factors that could influence this score, such as the duration of dementia and symptomatic treatment of NPS with neuroleptics or antidepressants. Regarding the distribution of symptoms based on the different stages, apathy was the most frequent in all stages. In the early stages (GDS 3), depression, anxiety, irritability and sleep disorders predominated, as observed in other studies in which symptoms were analysed in patients with mild cognitive decline or mild dementia (10, 14, 15, 45). Agitation became more frequent in stages GDS 4 and 5, which correspond to mild-moderate dementia, and psychotic symptoms (delusions and hallucinations) were more common in advanced stages (GDS 6 and 7), as observed in other studies (15, 42).
Changes in the frequency and intensity of NPSs as the disease progresses can be better observed by grouping symptoms into subsyndromes (2). The frequency of NPSs remained stable for the hyperactivity and apathy subsyndromes, decreased for the affective subsyndrome and increased for the psychotic subsyndrome. In contrast, intensity behaved differently, increasing both for the psychotic subsyndrome and for apathy throughout disease progression, and was more homogeneous for the affective and hyperactivity subsyndromes.
Numerous studies have grouped NPSs into subsyndromes (13, 15, 21, 27, 45, 47, 50, 51); however, few studies have examined their distribution based on GDS stage (52), especially in community-dwelling patients. This grouping by subsyndrome is of great interest from the clinical point of view, especially in a scenario of high frequency consultations, as happens in PC, because it allows simplification of the detection of NPSs and provides a more suitable approach with pharmacological or nonpharmacological measures based on the predominating symptom at any given time (53–56).
Among the limitations, it should be noted that although the size of our sample was small compared with population-based studies (7, 11, 15, 18, 57), it is consistent with those in other studies performed in clinical practice (8–10, 14, 27, 41). One strength is that our entire eligible population was included as study subjects. We believe that our work with noninstitutionalized patients in PC follow-up allows a better approach for community-based patients than those approaches performed in neurology and geriatric consultations (2, 7–11, 15, 16, 18, 41) with more restrictive selection criteria or in institutionalized patients (49, 58) who have more advanced stages of dementia.
In our study, the cognitive level and stage of progression were determined using the GDS by questioning the caregiver. Interviewing or administering cognitive tests to the patient directly was not considered, following the recommendations of the Clinical Research Ethics Committee (CEIm), which suggests limiting patient distress as much as possible if interviewing the caregiver can answer the research question. When available, the Mini-mental State Examination (MMSE) was also used, but it was not included in the final analysis because these data were not available for all patients, and it was not essential for staging.
Unlike other studies that only included individuals with mild-moderate stages of dementia (9, 10, 14) or that limited the age of inclusion (6, 7, 12, 13), in our study, we included patients at all stages of dementia of any age and with a diagnosis of dementia of any aetiology. We believe that this approach provides greater external validity to our results, although it may limit comparisons with other studies that only study Alzheimer's disease (10, 11, 15, 41) and/or the most frequent dementias (7, 18). For example, the wide dispersion of symptoms that we observed in advanced stages may be related to some patients who presented rapid-progression non-Alzheimer dementia.
Although patients in any stage of dementia were included, there was a low representation of the initial and final stages. The latter could have occurred because a high percentage of these individuals are treated in nursing homes, which usually occurs with more advanced dementias and/or with more intense symptoms. The lower representativeness of the initial stages could be explained by the usual delay in confirming the diagnosis from the onset of the first symptoms of dementia (59). Patients with memory deterioration or very mild cognitive decline could have early dementia but were not included, as only patients with confirmed dementia were included in the study.
This study is of interest because of the importance that these symptoms have in the management of the disease and the few studies on NPSs performed at this level of care (6, 27). NPSs condition the quality of life of the patient and caregiver and are one of the main reasons for PC consults by caregivers of patients with dementia. Knowing the frequency and intensity of the most significant symptoms and their course in the disease is crucial for patient management. However, in clinical practice, objective measurement instruments are not usually used to identify symptoms or to monitor treatment regimens. They are also not usually the object of research in PC, unlike the scales that measure cognitive symptoms (such as the MMSE) or function (such as the Barthel or Lawton activities of daily living scales), commonly used both in clinical practice and research. We believe that to improve the approach to these symptoms and the adequate care of patients and caregivers, it is essential to have a perspective from the reality of PC, reinforcing research in this field.