Accuracy of the National Adult Reading Test and Spot the Word Estimates of Premorbid Intelligence in a non-clinical New Zealand sample

This study investigated National Adult Reading Test (NART) and Spot the Word (STW) premorbid estimates compared to Wechsler Adult Intelligence Scale (3rd ed; WAIS-III) performances in 89 healthy New Zealand adults. STW and NART Full Scale Intelligence (FSIQ) estimates were strongly correlated to WAIS-III FSIQ for Europeans (n=75; rNART = .70, p ********** Estimates of premorbid intelligence are an essential aspect of neuropsychological assessment, with the contrast between premorbid level of functioning and current level of performance forming the basis of judgements about areas of preserved versus impaired functioning (Baddeley, Emslie, & Nimmo-Smith, 1993; Klesges & Troster, 1987; Lezak, Howieson, & Loring, 2004; Vanderploeg & Schinka, 1995; Watt & O'Carroll, 1999). As people are not often assessed prior to disease/injury onset, no direct comparison to premorbid functioning can be made, and indirect estimations of premorbid abilities are required (Bright, Jaldow & Kopelman., 2002; Hart, Smith, & Swash, 1986; Klesges & Troster, 1987; O'Carroll, 1995; Petito, 2000). Predictive methods must provide estimations that correlate highly with actual functioning (typically measured as intelligence [IQ]) in normal populations and must also be resilient to decline in cognitive function (Schoenberg, Scott, Duff, & Adams, 2002; Sharpe & O'Carroll, 1991). Such methods include demographic based regression equations, and use of tests of overlearned skills thought to be resilient to brain damage (Lezak, et al., 2004). In New Zealand, regression formulae to estimate premorbid intelligence are, as yet, unavailable. In clinical practice, judgements about premorbid level of functioning are often made using qualitative information about overall ability in conjunction with tests of overlearned skills such as the National Adult Reading Test (NART) and Spot the Word (STW). As such, the literature around these will each be reviewed briefly below. The NART and STW The NART (Nelson & Willison, 1991) is one of the most commonly used premorbid estimation methods. It consists of 50 irregularly spelled words that examinees are asked to pronounce correctly, as a measure of reading ability (Collins, 2000; Crawford, Parker, Stewart, Besson, & De Lacey, 1989; Hart et al., 1986; Nelson & McKenna, 1975; Sharpe & O'Carroll, 1991). Reading ability is a potentially useful method of predicting overall functioning as it requires a complex interaction of cognitive functioning, with inclusion of irregular words ensuring that it tests familiarity of a particular word rather than the ability to work out words phonetically (Collins, 2000; Nelson & O'Connell, 1978). A regression equation to predict premorbid IQ from the NART has been developed (Nelson & McKenna, 1975), and is a better predictor of premorbid functioning than the WAIS-R Vocabulary subtest (Collins, 2000; Petito, 2000). In a retrospective study, NART IQ estimates in people aged 77 years correlated significantly with IQ scores obtained from them at age 11 years, indicating only minimal effects of age and life experience on NART estimates (Crawford, Deary, Starr, & Whalley, 2001). The NART also remains highly stable over time (Morrison, Sharkey, Allardyce, Kelly & McCreadie, 2000; Smith, Roberts, Brewer, & Pantelis, 1998), and has been reported to be an effective predictor of premorbid 1Q in closed head injury and orthopaedic trauma patients (Watt & O'Carroll, 1999). …