Problems in Detecting Lifestyle Effects
on Intellectual Functioning

In an interesting debate in the literature, researchers from two major longitudinal studies on intellectual change in older adulthood (the Victoria Longitudinal Study and Canadian War Veterans Study) found conflicting results regarding the effects of an active lifestyle on intellectual functioning. On the one hand, Pushkar, Etezadi, Andres, Arbuckle, Schwartzman, and Chaikelson (1999) claim that they show evidence that engaged lifestyles have significant but small causal effects on verbal performance across the adult life span. In other words, an engaging and intellectually active adult lifestyle is associated with a more complex social environment in which intellectual abilities can flourish.

However, Hertzog, Hultsch, and Dixon (1999) did not find this relationship in their study. They argue that the verdict is still out, given conflicting findings in the literature and underdeveloped methods of assessing engaging lifestyles. Current research has weighed in on this debate suggesting that those studies that find cognitive enrichment effects do not find them on all cognitive measures (Hertzog et al., 2009). Some astounding findings suggest that mentally stimulating activities during adulthood are associated with a reduced incidence of mild cognitive impairment and dementia later on in adulthood (Wilson, Scherr et al., 2007).

It is also important to unravel the causal direction of these effects. Does activity facilitate cognitive functioning, or are those individuals with high cognitive functioning the ones who engage in an active lifestyle (Mackinnon et al., 2003)? Finally, none of these research teams would discourage older adults from actively engaging in intellectual activities and social lives such as crossword puzzles, reading, and adult education. Instead, Hertzog and colleagues (1999, 2009) argue that the current research suggests that mentally stimulating and cognitive enrichment activities have the effect of delaying the onset of cognitive decline, not preventing it. The positive aspect of this is that it thereby reduces the proportion of old age spent in a cognitively disabled state (Hertzog et al., 2009).

intellectual development has ignored this distinc­tion; we have been concerned only with normal changes. However, not everyone is healthy, expe­riencing only normal cognitive aging. Moreover, disease is a hit-or-miss proposition, affecting some people primarily physically, as in arthritis, and oth­ers primarily cognitively, as in the early stages of dementia. Thus we need to consider how specific aspects of health influence intellectual ability.

The most obvious relationship between health and intelligence concerns the functioning of the brain itself. We noted in Chapter 3 that several normative changes in brain structure with age affect functioning. Disorders such as Alzheimer’s disease and head injuries may damage the brain

250 CHAPTER 7 and interfere with its functioning. In some cases these problems get worse as the individual ages. Obviously, the more extensive the damage, the more significant the impairment of intellectual ability.

The connection between disease and intelligence has been established fairly well in general (Rosnick et al., 2004) and in cardiovascular disease in par­ticular (Spiro & Brady, 2008). Vascular diseases are linked to a pattern of cognitive impairment that is typically observed in “normal” cognitive aging. Some researchers suggest that the effects of age on intelligence and cognition are related at least in part to vascular disease that selectively affects the prefrontal brain (Spiro & Brady, 2008; Waldstein et al., 2005). Findings concerning hypertension are
complex. Whereas severe hypertension has been associated with earlier-than-usual declines, mild hypertension may actually have positive effects on intellectual functioning (Sands & Meredith, 1992).

Finally, other research indicates that health-related factors are related to very specific types of intellectual functioning. For example, in a longitudinal study the absence of an APOE genotype (associated with Alzheimer’s disease discussed in Chapter 3) and self­reported sensory functioning were associated with spatial functioning whereas reports of current health symptoms, depression, and anxiety were associated with perceptual speed (Christensen et al., 1999; Christensen et al., 2001; Hofer et al., 2002). Similarly, Ulman Lindenberger and Paul Baltes (Baltes & Lindenberger, 1997; Lindenberg et al., 2001) found that visual and auditory acuity (measures of sensory functioning) were related to fluid intelligence, which may explain why we see age-related decline in these intellectual abilities. In fact, they suggest sensory functioning is a better predictor of intellectual ability than variables such as educational level or occupa­tional success. Given these findings, it appears that a variety of variables, including health, sensory func­tioning, lifestyle, and education, predict the variabil­ity in individual trajectories of intellectual change in adulthood (Christensen et al., 1999).

Relevancy and Appropriateness of Tasks. All the psy­chometric tests used today trace their origin to Binet’s
(1903) original attempt to measure academic per­formance. Some researchers argue that the academic settings and skills that led to the development of these tests may not be equally important or relevant to adults. Consequently, they argue that we need new tests based on the problems adults typically face.

To build new tests requires understanding what types of skills adults use in everyday situations. Given that a major concern of older adults is to maintain independent living, Willis and colleagues have been conducting a program of research exam­ining the relationship between mental abilities and older adults’ competence in everyday situations (Diehl et al., 1995; Willis & Schaie, 1993). They drew on the seven domains of daily living out­lined by the Instrumental Activities of Daily Living (IADL) scale (Fillenbaum, 1985). These include t aking medications, managing finances, shopping for necessities, using the telephone, managing transportation, preparing meals, and housekeep­ing. The responses provide a very different starting point for developing tasks for an intelligence test than trying to figure out how to measure classroom learning potential.

Willis and colleagues (Allaire & Marsiske, 2002; Diehl, 1998; Marsiske & Willis, 1995, 1998) exam­ined the relationships among seven primary mental abilities, measured by traditional tests, and eight cat­egories of everyday tasks, measured by the ETS Basic Skills Test. The categories of everyday tasks included

understanding labels on medical or household articles, reading street maps, understanding charts or schedules, comprehending paragraphs, filling out forms, reading newspaper and phone directory ads, understanding technical documents, and comprehending news text. Three scores on the skills test were calculated; two scores reflected different levels of comprehension and information processing (literal and inference), and the third was the total score. Correlations between the scores for primary abilities and basic skills were very high for the older adults, indicating that the two tests were measuring similar things.

When Willis and colleagues examined their data to see which of the primary abilities best predicted each of the eight categories of everyday tasks, some interesting findings emerged. They had expected their measures of crystallized intelligence to be the best predictors, on the basis that these everyday skills reflect cultural knowledge and should not decline with age. Much to their surprise, the mea­sures of fluid intelligence, especially figural relations, were the best predictors most of the time. Moreover, older adults did not always perform as well as the younger adults on the everyday skills test. In fact, the younger adults obtained near-perfect scores, whereas the older adults were significantly below

ceiling, on average. Finally, a longitudinal analysis indicated that crystallized and fluid measures of intelligence predicted everyday functioning 7 years from the first time of measurement.

The Willis study is important for two reasons. First, it shows that traditional tests of primary men­tal abilities may predict performance on everyday tasks. For supporters of the psychometric approach, these data show that a total rejection of traditional tests on the ground that they are inadequate may be unwarranted. Second, the findings also show that tests consisting of what appear to be more relevant tasks may tap some of the same components of intelligence that the traditional tests are thought to measure. This suggests that we might develop new tests that consist of familiar tasks but those skills still tap the components of intelligence identified in psychometric research. Later we will return to this point when we discuss everyday problem solving.

The issue of task relevancy is still far from being settled, however. As we will see, many researchers and theorists argue quite strongly that only by abandoning a purely psychometric approach and moving to a focus on everyday uses of intelligence will we advance our understanding of intellectual aging. As with most con­troversies, there is something to be said for both sides.