Validade de critério de uma forma abreviada da escala Wais-III em uma amostra de idosos brasileiros

Abstract

The use of instruments that contribute to the detection and differential diagnosis of impairment is highly relevant for psychological evaluation. Wechsler Adult Intelligence Scale-III (WAIS-III) has received worldwide recognition as the gold standard. Yet, the time needed for its application may lead to fatigue, inattention and loss of motivation, when the elderly population is tested. Thus, abbreviated versions have been proposed and defended in the specialized literature. This study had the general objective to assess criterion validity evidence of a short form of the scale, composed of eight subtests (SF8). The study specifically aimed to: a) describe the epidemiologic profile; b) characterize the performance of the groups with and without cognitive decline on SF8; c) compare the performance in SF8 and subtests among the elderly with hypertensive chronic disease and non hypertensive, according to the schooling; d) provide estimates of IQs and factorial indexes of the scale; e) compare the qualitative classification proposed by the WAIS-III manual with the results obtained from the cut-off points; f) investigate the association between SF8 and tests of cognitive screening and; g) identify variables predictive of cognitive performance on SF8. The convenience sample was composed of 192 individuals from both sexes, aged 60 years or above, living in the community or in institutions devoted to the long stay of elderly, in the city of Juiz de Fora (MG). The interviews, composed of health and sociodemographic questionnaires, cognitive screening tests, and the SF8, were applied in the elderly´s places of residence, and lasted for approximately 40 minutes. Participants signed an informed consent form. The female sex predominated (n = 144; 75%). Mean age was 75.87 years (SD = 9.14) and mean schooling was 6.40 years (SD = 4.81). As fort marital status, 43.2% (n = 83) lived in widowhood, and 33.3% (n = 64) were married. In a sub-sample, composed of 118 elderly, 53.4% (n = 63) had hypertension. As for cognitive performance, the mean for the whole sample was 107.43 points (SD = 29.28), on SF8. On dividing the sample in groups with (G1) and without (G2) cognitive decline, performance on SF8 was observed to be positively and significantly associated with schooling. The highest cut-off point on SF8 was 142. Among elderly aged from 60 to 69 years, from 70 to 79 years, and 80 or above, the results were 160, 129 and 129 points, respectively. SF8 had good sensitivity (80.0%) and specificity (77.3%), with 95% confidence interval. Positive and negative predictive values were 77.3% and 80%, respectively. The regression model was composed of the variables schooling, age and marital status, accounting for 49.2% of the cognitive performance. From the psychometric parameters assessed, SF8 was relevant for the identification of cognitive impairment in the elderly. The instrument can be applied to the elderly, with a time reduction of approximately 50% compared to the full scale. It is suggested that the SF8 should be more frequently used, both in the clinic and in research, aiming at a better refinement of the knowledge on the usefulness of the cut-off points.