Cognitive functions decline with age, including the speed of information processing and memory performance (Salthouse et al., 1996; Bisiacchi et al., 2008; Johnson et al., 2009). The severity of age-related cognitive decline varies considerably between individuals, leading researchers to coin the term cognitive reserve (CR) to refer to differences in cognitive function between individuals of the same age (Stern, 2002, 2009). CR has been theorized to comprise passive and active components (Stern, 2009). The passive component reflects pre-existing CR due to brain characteristics such as size or the numbers of neurons and connections. The active component reflects how the brain exploits its cognitive abilities to compensate for aging, such as relying on higher general intelligence or using alternative cognitive strategies. Some cognitive activities are more efficient and resilient to the effects of aging than others (Stern, 2009; Park and Bischof, 2013). Numerous studies have used educational level as a proxy measure of the active component CR. This literature suggests that prior education may protect against age-related cognitive decline and risk of dementia (for a review, Summers et al., 2013). For example, a meta-analysis showed that education reduces the risk of incident dementia (Valenzuela and Sachdev, 2006), while a longitudinal study of Catholic clergy found that education modulates the risk of Alzheimer's disease (Bennett et al., 2003). These findings have been confirmed in the general population (Roe et al., 2007; Santos et al., 2013). In sum, the CR hypothesis states that individual differences in cognitive processes as a function of lifetime intellectual activities explain differential susceptibility to functional impairment in the presence of age-related changes and brain pathology (Barulli and Stern, 2013).

While studies have begun to provide a convincing picture of the ability of education to prevent age-related cognitive decline, few studies have examined the effect of education on age-related decline of emotional intelligence (EI). While researchers agree on the importance of EI for mental health and social and professional functioning, they have conceptualized it in quite different ways. The two most frequent EI models are mixed models and the ability model (Mayer et al., 2008). Mixed models view EI as a conglomeration of characteristics, including empathy, motivation, persistence, optimism, and social skills. Mixed EI is typically measured through self-report instruments, and it overlaps extensively with personality traits and measures of emotional/psychological well-being (Oconnor and Little, 2003; Grubb and McDaniel, 2007; Webb et al., 2013). The ability model, in contrast, defines EI as the integration of several abilities: “the ability to perceive accurately, appraise, and express emotion; the ability to access and/or generate feelings when they facilitate thought; the ability to understand emotion and emotional knowledge; and the ability to regulate emotions to promote emotional and intellectual growth” (Mayer and Salovey, 1997). Ability EI is assessed in adults using the Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT; Mayer et al., 2002). MSCEIT is a comprehensive performance test that requires individuals to solve tasks pertaining to each of the four abilities (“branches”) defined by the theory, which are referred to as perceiving, facilitating, understanding, and managing emotions (Mayer et al., 2003). MSCEIT scores show a positive relationship with various domains of daily life, including mental and physical health, social functioning, and academic and workplace performance (e.g., Mayer et al., 2008; Brackett et al., 2011).

Studies of how age affects ability EI have given mixed results. While some studies reported significantly better scores on all four EI branches for older adults (Mayer et al., 1999), other work found age to correlate negatively with emotional perception (Day and Carroll, 2004; Palmer et al., 2005). In fact, a more recent meta-analysis concludes that older people have difficulty recognizing emotions (Ruffman et al., 2008). Still other studies have found no significant association between age and MSCEIT branches (Farrelly and Austin, 2007; Webb et al., 2013). One significant gap in these studies is that they fail to explore a wide range in age; most focus on university populations and enroll participants with an average age younger than 30 years. Thus, studies examining EI in much older individuals are sorely needed.

Even less is known about whether educational level protects against age-related decline in EI. Goldenberg et al. (2006) found educational level to correlate positively with ability EI, but they did not analyze whether EI reserve (equivalent to CR) modulates the relationship between ability EI and age. Specifically, the EI reserve hypothesis is that individual differences in EI as a function of lifetime intellectual activities explain differential vulnerability to functional impairment in the presence of age-related changes. Therefore, the main aim of the present research was to examine educational level and ability EI in older adults from the general population in order to assess whether education can help prevent age-related EI decline. More concretely, our objectives were the following:

Explore the relationship of ability EI with age and educational level.

Numerous studies have demonstrated that CR protects against age-related decline in cognitive function. To examine whether the equivalent concept of EI reserve protects against age-related decline in EI, we carried out the present study in adults from the general population that were substantially older than the under-30 undergraduates typically examined in the EI research. Since the CR literature routinely uses educational level as a proxy measure of active CR (Bennett et al., 2003; Roe et al., 2007), we used the same variable to represent EI reserve. Specifically, the EI reserve hypothesis is that individual differences in EI as a function of lifetime intellectual activities explain differential vulnerability to functional impairment in the presence of age-related changes. We show that in our population, educational level appears to protect against age-related EI decline, providing the first evidence that educational level helps to counteract gradual losses not only in cognitive function but also in EI.

To examine whether EI indeed declines with age, we measured ability EI in older adults using the MSCEIT. We found that older people scored lower than younger people for total EI and for the three EI branches of perceiving, facilitating, and understanding emotions. In contrast, age was not significantly associated with the EI branch of managing emotions. These findings that age negatively affects numerous aspects of EI are consistent with reviews of clinical and theoretical studies that confirm the inevitability of age-related cognitive decline (e.g., Luszcz, 2011), as well as with a meta-analysis concluding that older people have difficulty recognizing emotions (Ruffman et al., 2008). Our finding that age does not significantly affect emotion management is consistent with behavioral and neuroscience studies associating old age with more stable and satisfying emotional well-being (Carstensen and Mikels, 2005; Scheibe and Carstensen, 2010; Carstensen et al., 2011), and with studies reporting that regulation abilities remain intact in older adults (Lamonica et al., 2010; Ochsner et al., 2012). Therefore, our results suggest that the EI reserve hypothesis applies to three of the four EI branches, but not to the branch of managing emotions, when considered independently of educational level. If educational level is taken into account, then the EI reserve hypothesis may apply even to managing emotions, since we found that people with a university education (who tended to be older) scored higher on this EI branch than did people with a lower educational level.

In fact, we found that the ability EI of adults correlated positively with educational level, and that this variable predicts several dimensions of EI over and above the effects of gender and age, namely, total EI and the EI branches of facilitating, understanding, and managing emotions. These findings are consistent with a previous study showing that educational level improves ability EI in older adults (Goldenberg et al., 2006).

We examined whether educational level mediates and/or moderates the relationship between EI and age. Educational level was found to mediate, and explain some of the age-related variance in, total EI and the EI branches of facilitating and understanding emotions. These findings for EI reserve mirror the results of studies on CR that concluded that education level is associated with better cognitive performance in later life (Gatz et al., 2001; Stern, 2009; Summers et al., 2013). Our moderation analyses showed that higher educational level attenuates age-related EI decline, insofar as older adults with a university education had higher EI scores than older adults with only primary or secondary education. This moderating effect of education has already been reported for cognitive abilities (Bosma et al., 2003). The implication of our work in EI and previous studies with cognitive intelligence is that a higher education level (proxy for EI reserve and CR) increases the likelihood that cognitive-emotional structures remain functional and stable during aging.

Despite the insights from the present work, it has several limitations. Its cross-sectional nature and reliance on a single instrument to assess ability EI precludes inferences about causality in the relationships among age, educational level, and EI. While our mediation analyses support the notion that educational level leads to higher ability EI, prospective studies are needed to address this question rigorously. Our sample was recruited from adult education centers, and the oldest participant was 76 so future studies should seek to include even older individuals. Our study examined exclusively educational level to the detriment of other variables that may help explain the observed variance in cognitive-emotional aging, including socioeconomic status, previous careers, lifestyle engagement, and quality of social relationships. Future studies should take these factors into account to provide a complete picture of the interaction of educational level, EI reserve, and age. Eventually this line of research should be extended to both theory of mind and social cognition (Fernández-Abascal et al., 2013), and EI should be assessed using mixed models or self-reported emotional regulation strategies to complement the performance-based activity EI that we have measured here (Cabello et al., 2013).

Finally, longitudinal interventional studies should be undertaken to test whether EI training can help older adults both with and without university level education against age-related decline in EI and to have a better mental and social life. Such research may provide more insight into how the association between age, EI, and educational level evolves over a lifetime. Evidence suggests that training in social and emotional competencies is crucial and should begin in the first years and be continuous throughout each life-span (Ruiz-Aranda et al., 2012; Rivers et al., 2013).

Our findings contribute to the growing literature on factors that may help protect against age-related decline in cognitive and emotional functioning, which affects quality of life, social adjustment, and professional performance. Educational level has already been shown to protect against age-related cognitive decline by modulating the relationship between age and cognitive abilities (Bennett et al., 2003; Valenzuela and Sachdev, 2006; Roe et al., 2007). Here we show that the same appears to be true for age-related EI decline, at least when older adults with a university education are compared with older adults with a primary or secondary education, or with younger adults of any education level.