Brazilian caregiver version of the Apathy Scale

No Brazilian version of a specific scale for evaluating apathy in dementia is available. Objectives To introduce a translated version of the Apathy Scale (AS) for use with caregivers. Methods The instrument was formally translated and then administered to the caregivers of a small sample of dementia patients, in order to assess scale comprehensibility and make final adjustments. The scale was subsequently administered to the caregivers of a second, independent sample of Alzheimer’s disease (AD) patients. The content validity of the scale was tested by correlating the AS scores with the Neuropsychiatric Inventory (NPI) - apathy sub-score and Disability Assessment in Dementia (DAD) total scores. Results The first sample consisted of eleven subjects with dementia, most of whom had AD. The second sample comprised twenty patients with probable or possible AD (10 with mild dementia), a mean age of 84.1±5.8 years, and 2.2±1.6 years of schooling. The AS scores correlated with both NPI-apathy sub-score (r=0.756, p=0.001) and DAD total scores (r=–0.793, p=0.0005). Conclusions The final version had good comprehensibility and correlated strongly with standardized apathy and functional activities of daily living measures.

hort of AD patients was significantly associated with worse cognitive and functional performance in follow-up. 9 Another important feature regarding apathy evaluation pertains to its distinction from depression. 10 Since Marin's initiative, 11 several other authors have proposed specific instruments to evaluate and quantify this syndrome in dementia, [12][13][14][15][16] helping to unravel this issue. To date, the most widely used instrument in the literature is the Neuropsychiatric Inventory (NPI). 17 However, this tool has several limitations: it is not specifically dedicated to evaluate apathy; it identifies non-relevant clinical symptoms when apathy scores fall below four; and lastly, since the NPI relies on a screening question, and apathy seems to be a heterogeneous disorder, some definite apathetic patients can be missed by the screening if the examiner is inexperienced.
Most of the research on apathy associated with AD can be ascribed to Starkstein's and his coworker's. 2,8,9 Using the Apathy Scale (AS) 12 -an instrument with 14 questions adapted from Marin's original 18-item Apathy Evaluation Scale (AES) -their group showed that apathy has major prognostic implications in AD. The scores on the AS range from zero to 42 points, with higher scores indicating greater severity of symptoms.
Based on the auspicious work of Starkstein et al. we believe it is time to improve our diagnostic capabilities and to better characterize apathy phenomenology. Additionally, any trial investigating interventions aimed at improving apathetic symptoms must have primary efficacy measures analyzed by instruments specifically dedicated to evaluating apathy in dementia. An ideal apathy quantifying scale should have a wide range of possible scores and should also be brief and easy to administer. We believe that the AS meets most of these requirements.
Although we already have a version of the NPI in Brazil, 18 we are unaware of a specific scale for evaluating apathy in our country. The primary aim of this study was to introduce a Portuguese version of the AS, suitable for caregiver interview, and to describe some of its basic and preliminary psychometric properties.

Methods
The study was conducted in three phases. In the first phase, the original version of the AS was translated independently by two of the researchers (HCG and PC). A consensus was reached to define the final translated version, which was then back translated by a linguistic expert. The back translated version was compared with the original for final adjustments. Minor adaptations to the final version were necessary in order to make it suitable for caregiver interview. Briefly, we only changed the questions from first to third person.
In the second phase, the final translated version was used to interview the caregivers of a small sample of 11 patients with AD or frontotemporal dementia (FTD). This pilot stage was designed to assess the translated scale's comprehensibility, and make any necessary adjustments to the instrument. The subjects were recruited from the Behavioral and Cognitive Neurology Outpatient Unit at the Hospital das Clínicas from the Federal University of Minas Gerais in Belo Horizonte (MG), Brazil. Caregivers were defined as those who spent most time with the patient, usually on a daily basis, and at least 12 hours a week. The study was approved by the local ethics committee and all participants gave their written informed consent.
In the third phase, another independent sample consisting of 20 patients fulfilling the diagnosis of either possible or probable AD were randomly selected from among the demented subjects identified in a large population-based epidemiological survey, The PIETÀ study, 19 conducted in Caeté, Minas Gerais state, southeast Brazil. This study also has local ethics committee approval and all the participants gave written informed consent. To form this sample, 10 patients were selected with mild stage dementia and ten with moderate or moderate advanced stages, according to the Functional Assessment Staging in Alzheimer's disease. 20 All the subjects were evaluated with the Mini-Mental State Examination. 21 Caregivers were defined as outlined above and were submitted to an interview consisting of the AS, NPI 17,18 and Disability Assessment for Dementia (DAD). 22,23 Dementia diagnosis was established according to DSM-IV criteria (APA). 24 AD and FTD were diagnosed according to standard published criteria. 25,26 AS was always administered by the same examiner (HCG); the other evaluations were administered by experienced neuropsychologists (ELS, PPAF and VAC). For statistical analysis, Spearman's rank correlation tests were performed between AS, NPI -apathy and DAD scores. The significance level adopted was 0.05.

Results
The original 12 and the final version of the translated scale are shown in the Appendix at the end of the manuscript.
In the pilot phase, the first sample consisted of 11 patients, whose caregivers were interviewed with the translated scale. Eight of the patients had AD (four women), a mean age of 73.8±4.7 years and mean educational level of 5.8±4.2 years, all presenting mild stage of dementia (FAST 4). The remaining three patients had FTD (two women), were aged 55.0±8.7 years and had 10.0±6.6 years of schooling. Based on clinical judgment it seemed that two of the FTD subjects were at mild stages of dementia (subjects 10 and 11), since there is no standard method for staging this kind of patient.
In the second AD patient sample, the AS scores correlated strongly with the NPI-apathy sub-score (r=0.756, p=0.001; Figure 1). Additionally, we found a robust inverse correlation between AS and performance on activities of daily living assessed with the DAD (r=-0.793, p=0.0005; Figure 2). We also found a moderate correlation between AS scores and FAST categories from the 28 AD patients from both first and second samples (r=0.401, p=0.037). We conducted additional analysis and found no correlation between NPI-depression sub-scores or any of the presented variables (data not shown). Finally, we grouped together all patients from the two study phases in order to compare the distribution of the 31 AS scores (Figure 3). The histogram shows a fairly wide range of results, at least in this study in which subjects at mild dementia stages predominated.

Discussion
In general, the interviewed caregivers displayed good comprehension for all the questions of the AS. Clarifications were sometimes necessary, especially regarding quantitative issues, such as "interests". In this case, we usually attained to the scale question and instructed the caregiver to compare the number of current interests with the ones the patient had before the memory impairment was noticed. In a few instances, caregivers with low schooling needed a brief explanation of what "apathetic" meant. In this case, a broad definition such as "uninterested, unmotivated, indifferent and unconcerned" was used.
The principal strength of this study is the confirmation of content validity for the translated version. Strong correlations were found between AS, NPI-apathy and DAD scores. Although the NPI is the most widely used tool for quantifying apathy in dementia research, it also has significant limitations and is heavily dependent on examiner experience. 27 In contrast, DAD is a functional scale for assessment of activities of daily living. It does not seem to depend on examiner expertise and has little influence from subjective bias. The robust correlation between AS and DAD clearly shows that we were measuring clinically meaningful behavioral disorder. Furthermore, depressive symptoms assessed with the NPI do not to explain the above findings.
It seemed that AS fulfilled most of the expected requirements. There is little doubt that this tool measures apathetic symptoms. In most cases, it took on average no more than 10 minutes to complete the scale and a wide range of scores were observed in the studied population. An obvious limitation of this study is that we did not present several important psychometric properties from the scale, such as inter-rater and test retest reliabilities. In response to growing calls from the Brazilian research community in Cognitive Neurology for a Portuguese version of a specific scale for evaluating apathy in dementia, we decided to publish our version in this preliminary paper prior to formal validation.     (3) slightly (2) some (1) a lot (0) 11. Are you unconcerned with many things? not at all (3) slightly (2) some (1) a lot (0) 12. Do you need a push to get started on things? not at all (3) slightly (2) some (1) a lot (0) 13. Are you neither happy nor sad, just in between? not at all (3) slightly (2) some (1) a lot (0) 14. Would you consider yourself apathetic? not at all (3) slightly (2) some (1) a lot (0) Total (0-42)