Neuropsychological test norms controlled for physical health: Does it matter?

Abstract

The objective of the present study was to investigate the effects of physical health on neuropsychological test norms. Medical and neuropsychological data from 118 healthy volunteer controls, aged 26-91 years, were collected during five recruitment occasions. The examinations included a clinical investigation, brain neuroimaging, and a comprehensive neuropsychological test battery. Test-specific statistical regression-weights for age, education and gender were calculated to establish preliminary test norms. Hierarchical regression analyses demonstrated that control in addition for physical health moved best performance from age 60 to 65 for abstraction; replaced a plateau above age 70 for verbal fluency, with a continued rise in performance; eliminated significant negative influences of age on auditory learning, spatial reasoning and complex copying; reduced them on wordlist recall, psychomotor speed, visual scanning and mental shifting; and slightly reduced negative influences of low education on most verbal tests, several memory tests, and psychomotor speed, indicating rises in normative scores of up to 0.8 SD at age 80 and 0.4 SD at age 60. No differences were found at age 40. Although the sample size is not adequate to be used for normative data, the findings indicate that norms uncontrolled for health overestimate the negative influence of advanced age and low education, implying a risk of drawing false diagnostic conclusions.

Health‐adjusted neuropsychological test norms based on 463 older Swedish car drivers

“…The increases in normative scores due to the control for health status displayed no consistent pattern related to level of education. This finding seems inconsistent with the finding in the recent study (Bergman & Almkvist, 2015) where the control for health slightly reduced negative influences of low education. In both studies, however, the effect of the control for health on the impact of education was very small, suggesting that the impact of education on cognitive test scores largely holds up against agerelated morbidity.…”

“…A notable similarity between the findings of the recent study (Bergman & Almkvist, ) and the present one was the relatively high susceptibility to age‐related morbidity of verbal learning and free recall, as measured by the Rey Auditory Verbal Learning test (Lezak et al ., ) in the previous study and SGRC free recall in both studies. The previous study found a susceptibility mainly to the diseases of the circulatory system and in the present one the susceptibility was to global health, implying close to ⅓ SD (residual) in impaired performance at age 80.…”

“…There was, nonetheless, a significant incidence of age-related morbidity and clinical signs explaining impairments in test performance in old age up to 0.36 SD (residual) in magnitude. This is a more modest influence than reported in a recent study based on a different sample (Bergman & Almkvist, 2015), where a health-related influence of up to 0.8 SD was found. The largest differences between the recent study and the present one concerns Rey-Osterrieth copy and TMT A, where health-related influences of up to 0.6 SD and 0.4 SD, respectively, were found previously, whereas we found no influence for Rey-Osterrieth copy and an influence of only 0.2 SD (residual) for TMT A.…”

“…of 31.0 for the copy task and 15.5 for the immediate recall task. The mean scores were higher, but not excessively so, for SGRC free recall (i.e., 5.95 vs. 5.43) and recognition (i.e., 3.33 vs. 3.14) when applying the means of age and education to predict preliminary normative values as recently published by Bergman and Almkvist (). We found no published reference data for the APT tests.…”

“…This position entails the risk of including persons with undetected cognitive impairments associated with chronic medical problems, which may lead to an underestimation of truly healthy functioning in old age. Indeed, a recent study implementing control for physical health by Bergman and Almkvist (2015) showed rises in preliminary normative scores of up to 0.8 SD at age 80 as compared to scores uncontrolled for physical health. Norms controlled for physical health might thus assist in drawing correct diagnostic conclusions.…”