Does IQ decrease with age?

The question of how our IQs change has been researched in numerous research investigations, often using the various versions of the Wechsler Adult Intelligence Scale (WAIS), the most recent being the 2008 WAIS-IV. The results have been confirmed many times, but the research is a little tricky to conduct for two reasons: 

(A) IQs are always calculated relative to a person’s age, whether that age is 10, 15, 25, 50, 72, or 88. So 25 year olds are compared to other 25 year olds in terms of the number of items they answer correctly on any given task just as 50 year olds are compared to other 50 year olds. For every age group, the average or mean IQ is set at 100. So we can’t directly compare the mean IQs across the adult age range because—by definition—every group will average 100. 

(B) the second hitch is that different age groups differ in their education level. It is more common now than 25 or 50 years ago for people to attend college and to earn degrees. Since education is related to IQ, that variable serves as a confound in the research. If IQs go down with age, how can we be sure that any decrease is due to age rather than the lower level of education, on average, for older than younger adults? 

Both of these problems are easily handled by researchers (I have conducted a number of these investigations with my colleagues). The first thing we have to do is to find a common “yardstick” to compare adults. We can compare the performance of 70 year olds, 60 year olds, 50 year olds, 40 year olds, etc. to the norms (reference group or standards) established for young adults. 

In my research, we define young adults as about age 30 (usually ages 25-34). In that way, young adults will have an average IQ of 100 because that is the way the norms are developed. When we compare adults across the life span to young adults that will tell us how IQ changes as we get older. But first we have to take care of the inequity in education across the age range. That can be done statistically by “controlling” for education (even though many more 30 year olds graduated college and many more 70 year olds dropped out of high school, this statistical procedure controls for the age to age differences. 

Now we can compare Full Scale (global) IQs for adults of different ages. A clear decline is evident. The mean WAIS-IV IQ is 100 for ages 20-24 and is 99 for ages 25-44. Then it drops to 97 for ages 45-54, to 94 for ages 55-64, to 90 for 65-69, to 86 for ages 70-74 and to 79 for ages 75+. 

But global IQ is an amalgam of different kinds of intelligence, the most popularly studied being fluid intelligence and crystallized intelligence which together—along with abilities called working memory and processing speed—are combined to yield global or Full Scale IQ. Fluid intelligence or fluid reasoning (abbreviated Gf) reflects the ability to solve novel problems, the kind that aren’t taught in school, whereas Crystallized intelligence or crystallized knowledge (Gc) measures learning and problem solving that are related to schooling and acculturation. And they have very different aging curves. 

Gc averages 98 at ages 20-24, rises to 101 by ages 35-44, before declining to 100 (ages 45-54), then 98 (55-64), then 96 (65-69), then 93 (70-74), and 88 (75+). 

The decline with age in Gf—solving novel problems—is even more precipitous. Gf peaks at ages 20-24 (100), drops gradually to 99 (25-34) and 96 (35-44) before starting a roller coaster plunge to 91 (45-54), 86 (55-64), 83 (65-69), 79 (70-74), and 72 (75+). 

These values are just averages for the entire US population of adults, with the mean IQs for each age higher for more educated individuals. But the same rate of decline across the age range seems to occur for all adults, on average, whether they are semi-skilled workers or university professors. 

Two sources to consult for those interested: Essentials of WAIS-IV Assessment—2nd edition (Elizabeth Lichtenberger & Alan Kaufman), 2013, John Wiley publisher and IQ Testing 101 by Alan Kaufman, 2009, Springer publisher.