Synonyms

Aptitudes; Cognitive abilities; Intellectual development; Intelligence

Definition

Cognitive abilities are defined as a person’s mental capacity to do or act; broadly considered, cognitive abilities include attention, reasoning abilities, memory, and knowledge (Salthouse 2012).

Answers to questions about the development of cognitive abilities with age have implications for work performance, socioeconomic success (i.e., income and education, SES), and even mortality (the likelihood of mortality at earlier ages increases at lower ability levels, even after controlling for SES) (Salthouse 2012). Cognitive ability facilitates the execution of an array of tasks associated with a successful life, such as registering and completing courses in school, completing job applications and successful execution of job tasks, and simply getting from one place to another. Although not the only important factor, cognitive ability is a central determinant of life success.

Answers to questions about age-related changes in abilities are complex. For one, ability changes throughout the lifespan vary by person. For instance, two 50-year-olds may have extremely different intellectual profiles: one may have the same measured cognitive abilities as an average 30-year old and the other may resemble an average 70-year old. Moreover, within the same person, different abilities decline and/or grow at varying rates. These changes are a function of the continuous use of some skills, which serves to preserve skill-related abilities and the decay of unused skills. As such, there is significant between- and within-person variability in age and abilities. Because of this variability, there is not an agreement on the age at which a person becomes an “older” person. In this review, general changes in abilities are described. Research suggests that these changes are a function of regular aging (memory impairment that is a function of psychopathology such as dementia or Alzheimer’s disease is not considered). Nonetheless, it is important to note that the trends described herein will not occur at the same age for every person (Hertzog et al. 2008). Moreover, ability is not a monolithic construct and different types of abilities have different patterns of growth and decline throughout the lifespan.

Cognitive Abilities

There are two categories of cognitive abilities most relevant to aging: one related to reasoning abilities associated with generating, transforming, and manipulating information and the other related to knowledge accumulated throughout the lifespan. These abilities have different names depending on theoretical orientation; they have been referred to as fluid and crystallized abilities representing the reasoning and knowledge components, respectively, and the process (reasoning) and products (knowledge) associated with cognition (Carroll 1993; Horn and Cattell 1966). They are thought to represent, for example, a person’s ability to acquire new information compared to the information already known (Salthouse 2010). For simplicity, the terms reasoning and knowledge are used to denote these different types of cognitive abilities.

Measures of reasoning and knowledge abilities are positively correlated in the general population; that is, a person who has relatively higher reasoning capacity is also likely to acquire more knowledge. This relationship reflects the idea that reasoning ability is a major determinant of learning and knowledge acquisition throughout the lifespan. Indeed, the development of knowledge and expertise within a domain is often described as a function of the investment of reasoning ability such as when a student works with full attention to complete a calculus problem in a unit he/she is learning or when an accountant learns a new spreadsheet program to increase his/her productivity (Ackerman 2014).

Despite this positive relationship, however, reasoning and knowledge have different trajectories over the lifespan. The trends differ slightly depending on how the abilities are measured and depending on the design of the research study (as discussed below), but both reasoning and knowledge increase up to early adulthood, when their paths begin to diverge. Reasoning abilities begin to decline early – some studies suggest as early as late adolescence or early adulthood – and continue the downward trend throughout older ages. The size of the effect varies by study, but generally research shows a decline of about 1.5–2 sample standard deviation units from when a person is in their 20s to when they are in their 70s in reasoning and related abilities (e.g., memory, speed, and working memory tests, Salthouse 2010). By contrast, knowledge levels remain stable and may even increase, up until age 70 or so (Salthouse 2010). Patterns of reasoning and knowledge abilities are shown in Fig. 1, which is derived from research conducted with thousands of participants using an array of measures and study designs (Ackerman 2014; Salthouse 2010). The dashed line represents the growth and stability of knowledge throughout the lifespan, while the solid line represents the growth and subsequent decline of reasoning abilities.

Age-Related Changes in Abilities, Fig. 1
figure 13figure 13

Hypothetical trajectories of knowledge (crystallized abilities shown with the dashed line) and reasoning (fluid abilities shown with the solid line) by age group. The figure represents a compilation of research findings on age-related changes in abilities using an array of measures and both cross-sectional and longitudinal research designs (Ackerman 2014)

Full size image

Some theoretical perspectives place a greater emphasis on reasoning abilities than knowledge as representative of intelligence (Spearman 1904). These perspectives either consider knowledge to be a product of intelligence, but not an essential component of it, or they ignore knowledge completely. Given that reasoning abilities start declining relatively early in life and continue a downward trajectory, this perspective provides a relatively pessimistic view of intellectual development at middle and older ages. Furthermore, this view neglects compelling evidence – available through everyday encounters with smart and successful people – that intellectual abilities continue to develop throughout life. For instance, the overwhelming majority of CEOs of fortune 500 companies in the United States is between the ages of 45 and 70. Similarly, with few exceptions around the globe, heads of states are likely to be older versus younger. Given the ability trajectories shown in Fig. 1, these leaders would be considered long past their intellectual peak if reasoning were the sole or central cognitive ability important in adult intellect (Salthouse 2012). In the context of aging, theories that emphasize reasoning abilities over knowledge paint a relatively pessimistic picture of adult intellectual development; a picture that is not aligned with lay observations and common sense.

Theoretical perspectives that consider adult intellect to be comprised of both reasoning and knowledge give credit to adults for their knowledge and expertise (Ackerman 2014). And although there is little research on the topic of how adults might continually develop their knowledge and expertise even with declining reasoning abilities, it seems likely that people typically choose environments (i.e., for education, work, home, hobbies) that align with their established knowledge and skills. This strategy increases people’s reliance on their vast repertoire of knowledge and expertise and also reduces the need for people to reason through every problem in their environment as if it were new. Indeed, research suggests that even though declining reasoning abilities with age can make learning novel information difficult, domain-specific knowledge facilitates the acquisition of new knowledge in that particular domain (e.g., an extensive understanding of investment products facilitates learning about managing investments within a retirement account) (Ackerman and Beier 2006). In this way, the age-related trajectories of abilities shown in Fig. 1 can be considered somewhat adaptive; that is, people have less need to reason through difficult problems as they age because they have developed vast stores of knowledge through experience that they can bring to bear on an array of adult situations. A middle-aged or older engineer, for instance, might work on a variety of projects during a year – learning something new from each of them – and this learning may not seem very effortful. Nonetheless, it would be more difficult, although probably not impossible given enough time and effort, for the middle-aged or older engineer to learn a completely new field, like psychiatry.

Assessment

There are a variety of methods used to assess reasoning and knowledge abilities, and a researcher’s choice of measure will undoubtedly affect the outcome of the research. Reasoning abilities are typically measured with abstract problems such as pattern completion with figures and numbers (e.g., number series tests where test takers complete a pattern of numbers and Raven’s advanced progressive matrices) (Raven et al. 1991). These tests are designed such that performance is relatively knowledge and context free (although it is certainly the case performance is affected by a person’s familiarity with test taking and that practice in this regard can affect performance). Assessments of working memory capacity – also shown to be related to reasoning ability – are relatively free of knowledge and focus on a person’s ability to simultaneously process and store information. Examples of such tests are the backward digit span test, which requires test takers to recall – in reverse order– a set of three or more numbers that are read aloud, and the operation span test, which requires test takers to make decisions about the veracity of an equation while remembering the equation’s numerical outcome (Ackerman et al. 2002).

Because no individual measure is perfectly reliable – or a perfect reflection of a concept as complicated as cognitive ability – researchers typically use a battery of multiple measures to assess reasoning abilities (e.g., spatial, numerical, symbolic). Reasoning ability is then estimated by aggregating – or averaging – people’s scores on these multiple measures. This approach is similar in concept to factor analytic approaches, which derive an ability factor by pooling the common variance among measures (Ackerman et al. 2002; Carroll 1993). Aggregation helps control for the influence that the measurement error or content associated with any one test has on the assessment of reasoning ability, which can be substantial. For instance, if the only test used to assess reasoning ability is a number series test that is only somewhat reliable, a person’s score on that test would be a function of their reasoning ability, but also a function of their numerical ability and the measurement error associated with the particular test used. It would also be impossible to separate the amount of variance associated with each of these factors (reasoning ability, numerical ability, and error). To avoid these issues and to get a reliable assessment of reasoning ability, scores derived from most commercially available intelligence assessments are a function of an aggregation of individual items and measures over a range of content (e.g., digit symbol, block design, matrix reasoning, and letter number series in the Wechsler Adult Intelligence Scale) (Wechsler 1997).

Knowledge is typically measured with vocabulary tests or general information tests that include questions about widely available information within a cultural context (e.g., What is the capital city of France? Who was Benjamin Franklin?). As discussed above, performance on general cultural knowledge tests remains relatively stable across the lifespan, but performance on these tests does not typically show increases in knowledge with age. This is somewhat puzzling given the expectation that knowledge will continue to grow as a function of professional and life experiences. One reason for this discrepancy is that, because knowledge develops in ways that are unique to a person’s experiences, knowledge acquisition is idiosyncratic. As such, a complete picture of what a person knows would include a lot more than general cultural knowledge; it would include knowledge about his or her job, hobbies, and unique life experiences – essentially anything encountered and learned throughout the life course (e.g., the length of time a whole chicken needs to roast, when a child should be taken to the doctor, how to operate a forklift). As implied by these examples, capturing the whole of knowledge through the lifespan – giving adults credit for what they know – would require an impossibly elaborate knowledge battery. Indeed, researchers endeavoring to assess knowledge growth with age have measured knowledge across multiple academic (e.g., 20 academic domains including natural science, business, social science, and humanities) and nonacademic (e.g., current events, health, financial, and technology knowledge) domains (Ackerman 2014). In this research, age was positively correlated with knowledge possessed across all domains, with the exception of those domains most related to natural science (e.g., physics and chemistry). Nonetheless, these elaborate knowledge assessments will still underestimate what adults actually know because assessments can never account for the idiosyncratic nature of adult experiences that lead to knowledge and expertise.

Research Designs

Most research on age and abilities is cross-sectional in nature, meaning that people of different ages are assessed simultaneously. Inferences about age-related changes are made by examining the test scores for people of different ages (e.g., comparing performance on an ability battery for 20- versus 70-year-olds or correlating ability scores with age). Though informative, these studies are limited in that differences between age groups may not represent age-related changes within a person. A classic anecdote illustrates this point (Salthouse 2010). A scientist examining age-related changes who finds himself/herself in Miami in the year 2014 might observe that younger people are more likely to be of Hispanic/Latino or African-American descent, while older people are more likely to be of European descent. Based on this observation of an age-diverse cross section of the population, the researcher might conclude that people tend to become increasingly European looking (i.e., white) with age. This is absurd of course, but it is meant to illustrate that cross-sectional studies may lead to erroneous conclusions about age-related changes because they do not actually assess the changes within a person that are a result of aging; rather, they assess differences between people and presume that these differences are a function of age. Moreover, these designs do not control for environmental, societal, or other extraneous factors that might affect people differently by age group.

Cohort effects are an example of a societal influence on cross-sectional studies in aging. A cohort is a generational group that presumably shares a cultural identity. Factors that affect one cohort differently than others can influence the development of abilities. For instance, millennials are generally defined as those people who reached young adulthood around the year 2000 (i.e., they were born around 1980 or so). In developed and developing countries, millennials have grown up with access to technology that allows them to communicate globally in minutes and that provides them access to a wealth of information at the press of a button. In this example, access to technology would affect the development of knowledge differently for millennials relative to older cohorts. As such, cross-sectional studies on aging and knowledge would capture differences in knowledge that are a function of age and cohort and importantly, the variance associated with each could not be separated (a researcher could not determine what differences between people were a function of age vs. cohort). In cross-sectional designs, cohort essentially introduces a third variable (or confound) in the study. For this reason, there is considerable debate about the value of cross-sectional studies for examining age-related changes in abilities, with some researchers taking the extreme position that the value of cross-sectional research in aging is limited (Salthouse 2010). Rather than discounting all cross-sectional studies, however, it is probably important to understand the influence of cohort vis-à-vis the constructs and variables in question. For instance, the discussion above highlights that cohort might be an important influence on knowledge development, particularly as related to millennials versus older generations. It is less clear, however, how cohort effects might influence the development (growth and/or decline) of reasoning ability.

In contrast to cross-sectional studies, longitudinal research tracks the development and decline of abilities within a person by administering the same (or similar) measures periodically over time. Most of these studies include the periodic inclusion of a new sample of younger participants to ensure a continuous sample given attrition and mortality. Examples of significant longitudinal studies in cognitive aging include the Seattle Longitudinal Study (Schaie 2013), which was started in the 1950s with a sample of about 500 people ages 20 to 69. Participants were assessed on a battery of reasoning and knowledge measures on 7-year intervals, and every 7 years until 2005, a new cohort was added to the study. The Victoria Longitudinal Study (Hultsch et al. 1998) is similar to the Seattle Longitudinal Study, but the sample is somewhat older (55–85) with new cohorts starting every 10 years or so. Each of these studies has assessed the abilities of literally thousands of participants.

Although longitudinal studies are rare because of the time and resources involved, they provide information about within-person change in abilities and can control for cohort or other influences. Fortunately, the results of longitudinal studies tend to echo those of cross-sectional studies; that is, most of this research shows the growth of both reasoning and knowledge until early adulthood, the subsequent decline of reasoning abilities, and the relative stability of knowledge. Longitudinal studies show a more optimistic picture of cognitive aging than do cross-sectional studies, however. That is, the decline of both reasoning abilities and knowledge tends to be relatively later in longitudinal research (e.g., reasoning abilities begin to decline closer to age 30 in longitudinal studies vs. around age 20 in cross-sectional studies) (Ackerman et al. 2002; Schaie 2013). In summary, the age-related trajectories of cognitive abilities shown in Fig. 1 reflect trends found in cross-sectional and longitudinal research designs.

Ability Preservation

Important questions have been raised about the factors that affect changes in cognitive abilities throughout the lifespan, and the answers to such questions can inform interventions to preserve abilities. To date, many possibilities have been investigated (e.g., gender, personality traits, initial levels of abilities, and environmental influences such as education and health, Ackerman et al. 2002), but there is generally little evidence that any one factor exerts a strong effect on the course of age-related changes in abilities. There is some research to suggest that a person’s initial level of ability, overall health, and education will differentiate people by ability level throughout the lifespan (Salthouse 2010). For instance, a person who starts out with significantly lower scores on reasoning ability tests relative to others in the population of the same age will likely continue to have relatively lower scores compared to the same population throughout the lifespan; a person who is healthier will likely have higher reasoning ability and knowledge scores throughout their life compared to someone who is less healthy.

Recent research has focused on the preservation of abilities throughout adulthood (into older ages). This preservation is indeed important as most people will tend to experience some form of intellectual decline, even in knowledge and expertise, in late life (e.g., age 80 and beyond). The aging of the global population, coupled with the daunting prospect of the loss of cognitive abilities, has increased the urgency of finding remedies to age-related cognitive decline. Common ability preservation strategies include both cognitive (e.g., brain training) and physical (e.g., exercise) approaches.

Brain training. Brain training typically employs cognitive exercises to enhance a person’s working memory. Based on models of physical fitness that target exercises to specific muscles for strengthening, brain training is designed to strengthen memories or reasoning abilities through mental drills. At least in the United States, brain training is developing into a profitable industry, with advertisements extolling the virtues of online brain training exercises for people of all ages. Unfortunately, little empirical evidence has shown brain training to be effective; meta-analytic studies examining training effectiveness found little benefit to using these programs (Melby-Lervag and Hulme 2013). Some research has shown that direct training on working memory measures can be effective for increasing cognitive performance. These effects have typically been small, temporary, and limited to already cognitively healthy individuals, however. Moreover, these short-term improvements tend to exist only for the specific working memory tasks practiced in training (or similar tasks), meaning that the effects of working memory training are relatively narrow and have not been found to transfer to more generally complex life tasks (Hertzog et al. 2008). Nonetheless, because of the importance of preserving cognitive abilities into older ages, many researchers continue to work on developing effective strategies for preserving mental abilities through brain training. The bottom line is that current brain training activities are not likely to improve general memory or mental functioning in a measureable way, but they may not do any harm either. Moreover, to the extent that remaining cognitively engaged leads to learning and skill acquisition (i.e., expertise in an area), these exercises may increase levels of knowledge.

Physical exercise. Research on physical exercise has shown promise for its effect on preserving cognitive abilities into later life. These findings extend to both short- and long-term exercise interventions and have been most compelling for aerobic exercises (i.e., those that increase heart rate such as brisk walking/jogging vs. stretching) (Hertzog et al. 2008). The key to cognitive benefit appears to be enhancing cardiorespiratory functions that lead to myriad health benefits related to increased tissue oxygenation (healthier muscles, heart, and brain). Studies examining short-term aerobic and high intensity exercise interventions suggest better performance at simple cognitive tests postexercise. Effects are largest for people with lower cognitive ability predating exercise interventions. Long-term effects of exercise are a bit more complex to study. In younger cohorts, regular aerobic exercise has been shown to predict improvement in various tasks related to reasoning ability and working memory (Guiney and Machado 2013). For healthy older adults, however, regular physical activity does not appear to improve cognitive ability, so much as maintain it. That is, people who engage in regular aerobic exercise across the lifespan are expected to optimize cognitive ability when young and maintain ability longer and more effectively as they age.

Conclusion

Medical science has succeeded in expanding life expectancy across the globe. According to the World Health Organization, people born in 2012 can expect to live 6 years more, on average, than people born in 1990, and average life expectancies are now around age 80 for developed countries (such as Japan and the United States) (World Health Organization 2014). Cognitive abilities are essential for healthy aging – they permit people to travel, work, engage in hobbies, and enjoy life. Preserving abilities into late life will help ensure that people can take advantage of the additional years granted by medical science by remaining mentally active and engaged. Age-related changes in abilities are inevitable, and these changes will depend on myriad factors: the person, initial levels of ability, and the ability in question. There are well-established general trends, however, as shown in Fig. 1. As people age, they can expect a relatively early decline in reasoning abilities (and other related abilities such as working memory) and stability and even improvement in those abilities associated with the acquisition of knowledge and expertise.

Research in cognitive aging is moving toward an understanding of the outside factors – such as mental and physical exercise, lifestyle, and education – that influence the relationship between age and cognitive abilities. Research in this area promises the development and testing of interventions designed to help maintain and even increase cognitive abilities into old age. In this way, researchers are simply responding to the demands of a rapidly aging global population to stave off pending declines. Although the research is currently inconclusive, the best evidence suggests some promise for remaining mentally and physically active throughout the lifespan. The brain, after all, is an organ that benefits from physical activity just as do other organs in the body. And although the research on mental exercise is still inconclusive, brain training activities are unlikely to do any harm, especially if people refrain from spending excessively on unproven techniques (e.g., brain training software programs). There are, after all, plenty of relatively inexpensive ways to stay mentally engaged (e.g., crossword and other word puzzles, math games, reading a book). For both mental and physical health, however, cognitive benefits are most evident when people start early and remain consistently active.

Cross-References