Chile has 2.8 million people 60 years of age and older [16.2% of the total Chilean population; (7)]. According to the Chilean National Health Survey 2009–2010 (8), prevalence of cognitive impairment in this group was 10.4% and it rises rapidly with advancing age (12.8% for people 70–79 years old and 20.9% for 80 years and over). Moreover, this prevalence is much higher among older adults living in contexts of social and economic vulnerability and with low educational levels; 39% had some degree of disability. From a socioeconomic standpoint, 44% of disabled people in Chile belong to low socioeconomic sectors. Older adults account for 44.3% of disabled people, with a high percentage having low levels of education [69.6% according to (9)]. The same trend was seen in the educational level, where prevalence of cognitive impairment was 5.6 times higher among less educated adults than among those with higher educational levels (8). Disability to perform in daily life among the elderly has the same distribution by socioeconomic and educational levels; which means that the poorest and least educated have a higher prevalence of disability to perform in everyday life (9). There is also a high prevalence of depression in Chile (17.5% according to the latest National Health Survey ENS Chile from 2009 to 2010), especially so among older adults (10) in vulnerable conditions (11). This is relevant, given the association between depression and development of dementia (12), for example. Among diseases associated with cognitive impairment, in Chile Alzheimer's disease is the leading cause of dementia in older adults. Its incidence increases with age (1–2% of the people aged 60 years, 3–5% of people 70, in 15–20% of people 80 years and one-third or half of those over 85). Its clinical course generally begins with failures in recent memory and ends with total dependence (13). Chilean authorities consider factors such as low education and low levels of culture (low-income patients from vulnerable and rural contexts) to be social determinants of this kind of disease (14). Social vulnerability is understood as a set of social characteristics that put a group of people who live in contexts of lower economic resources and at high risk of falling into poverty in a situation of structural, material and personal inequality (15, 16). Typically, these populations have low cultural and educational capital, limited access to goods and services, lower quality of social benefits (health, education, housing), higher levels of social deprivation, more impoverished social environments, and unsafe neighborhoods, effectively lowering the quality of life of their inhabitants compared to the general population (17, 18). For example, it is known that environments of lower social capital or poverty are more likely to be exposed to higher toxins and pollutants, crime and traffic and have fewer chances to participate in physical activities, less access to healthy food, greater likelihood of living in chaotic homes, more violence and lower parental sensitivity (19). Thus, exposure to multiple stressors, that is, various risk factors, has greater side-effects on cognitive abilities than exposure to a single risk (20, 21).

The term psychosocial adaptation (PSA) refers to quality of life in terms of social and relational activities. This concept includes multiple dimensions such as social behavior, emotional regulation, and the development of social habits (4, 5). PSA represents an ecological approach to everyday and cognitive-contextual adaptation, in which cognition and emotion are interrelated (6). Within PSA, social cognition (SC), fluid, and crystalized intelligence (FI; CrI) play a highly significant part, particularly in decision-making, emotional processing, and the way we relate to others (empathy and theory of mind). Research suggests that the prefrontal cortex plays a major role in such adaptability, given its involvement in the flexibility of behavior, executive functions (EF), FI and SC (22, 23). It should be noted that not all EF are related to FI (24–26). Similarly, SC tasks and FI (27, 28) or mental flexibility have been associated with this area (29, 30). Damage or alterations in the frontal lobe have direct implications on these functions, resulting mainly in maladaptive behaviors. SC includes the ability to make decisions, emotional processing, the ability to understand others' intentions and to develop in the social world (31). Studies in this line have shown that the social context exerts a profound influence on SC (32, 33). Meanwhile, FI has been defined as the ability to think logically and solve problems in new situations, regardless of the acquisition of knowledge. This reflects the ability to reason and think abstractly, in contrast to what is called crystallized intelligence (CrI) (34), which depends on cultural and academic learning. From a neuroanatomical viewpoint, FI has been associated with frontal lobe functions (27). Injuries in this area affect the performance of these cognitive abilities (28, 35) and studies that have measured FI with functional neuroimaging have shown activation of frontal areas (36, 37). FI has been linked as a protective factor for mental health, violent conditions and PSA (38). On the other hand, CrI is known as the ability to use skills, knowledge, and experience (39), relying on accessing information from long-term memory. As McGrew (40) establishes, it “is primarily a store of verbal or language-based declarative (knowing what) and procedural (knowing how) knowledge acquired through the investment of other abilities during formal and informal educational and general life experiences” (p. 5). According to this, CrI is indicated by a person's depth and breadth of general knowledge, vocabulary, and the ability to reason using words and numbers. As such it is also viewed as the product of educational and cultural experience in interaction with FI. One relevant element is that it changes with age (41, 42). For example, research on individuals over the age of 60 showed that many abilities indeed show average decline rates (43–45). Thus, given the description of SC, FI and CrI, this gains relevance as protective factors in aging, in addition to cognitive reserve (CR). In this sense, the CR can be defined as “differences in cognitive processes as a function of lifetime intellectual activities and other environmental factors that explain differential susceptibility to functional impairment in the presence of pathology or other neurological insult” [(46) p. 502]. CR acquires importance since it proposes two forms of protective actions (1, 2). Neural reserve refers to pre-existing brain networks that are more efficient or have a greater capacity but are less susceptible to change. It can also operate like an alternative neural network that compensates for the disruption of pre-existing pathology networks. Despite the fact that CR has mainly been linked to the protective role in everyday activities, as well in education, work, or socio-economic status against cognitive impairment (47), literature on SC shows that very little attention has been paid to its moderating role. It has even been suggested that there is no relationship (48, 49). This is interesting, considering that it incorporates environmental aspects as well as intrinsic body actions [concept differs “brain reserve capacity” (BRC) of (3)]. Epidemiological studies have shown that individuals with higher FI, educational and occupational levels or leisure-related activities have a lower risk of developing AD and other neurodegenerative diseases (2). In the same way, SC has been associated with positive outcomes in aging. On the one hand, factors such as positive life events and social support, have proven to influence and increase survival rates in patients with dementia compared to subjects living in poor conditions (50). Similarly, Fratiglioni et al. (51) have found that the construction of social networks with poor or limited support increased the risk of dementia by 60% and that there is a link between variables such as isolation, feelings of social isolation, educational level, among other lifestyles, that accelerate or delay the occurrence dementia (52, 53). These can be seen as protective factors, but with an active role (1) based on neural networks. On the other hand, evidence from epidemiological research suggests that higher FI, high educational and occupational levels (better jobs), being active in leisure activities or larger social networks, have a lower risk of developing, for example, Alzheimer's disease (AD) or dementia (54, 55). Therefore, we understand that people with better social adaptability, given their greater social cognition, will be more resistant to neurodegenerative diseases, either due to the existence of a more functional network or a larger brain capacity to involve alternative networks. There are known risks factors for neurodegenerative diseases such as genetic or medical conditions, other neurological pathologies, or brain injuries. Nevertheless, the role of social variables have been studied less formally and are less understood, especially considering only individual factors (56). One possible hypothesis about the relationship between CR and FI is that other protective social and cultural factors could modulate the adaptation process, benefiting from the social and cognitive conditions that people possess, such as social support and social networks, demographic aspects, educational level, among others. Although the concepts of CR, CrI and FI differ in important respects, they are complementary as opposed to competing. Accordingly, the study of social factors as predictors of the delay or acceleration of neurodegenerative diseases is essential and we sought to fill a gap in previous research by conducting a study to find the relationship between fluid intelligence (FI) and crystallized intelligence (CrI), and as complementary factor with cognitive reserve (CR), as well as their relationship to social cognition (SC), emotional recognition (ER) and executive function (EF) in older adults in socially vulnerable contexts. In addition, we want to go into greater depth and answer the question of which of these two types variables (FI or CrI) better predicts social adaptive capacity. We suggest that FI and CrI, are associated with aspects of SC, such as emotional recognition and theory of mind (ToM), in addition to EF; and could strength cognitive process that are crucial for CR. All these aspects could influence the processes of social adaptation and thus the possibility to the adjustment of the elderly in vulnerable contexts. CR is often estimated using proxy variables for lifetime exposures and cognitive activity: years of education, measures of crystallized intelligence, such as vocabulary or knowledge, literacy level, number of intellectually stimulating leisure activities, degree of occupational complexity, and socioeconomic status are all commonly used to create an estimate of CR (1). According to the above, FI and Crl are variables that we propose to use as a synthetic way to explain a broader construct of CR (since both variables are components of CR).

All twenty-seven participants in this study were in healthy conditions, over 60 years old (M = 66.44, SD = 6.59; 55.6% male) and recruited from contexts of social vulnerability. Given that they come from contexts of high social risk (poverty and insecurity), they are participants who are very difficult access. They come from families currently participating in a Chilean Social Security Program—CSPP—implemented by the Chilean Ministry of Social Development. All participants signed an informed consent, following the protocol of the Declaration of Helsinki. Participation was voluntary, protecting participants' anonymity. The following exclusion criteria were considered: individuals with visual and/or hearing impairment that prevents them from performing the various tasks and measurements in the study; and a psychiatric or neurological background representing an impediment to the evaluation of the protocol, assessed in an initial interview.

We obtained descriptive statistics (central trend and dispersion measurements) and subsequently calculated Pearson's correlation coefficients to evaluate the independency between variables. A Kolgomorov-Smirnov Test was conducted to test for normal distribution. We worked with p level of <0.05 (two-tailed) to confirm bivariate correlations. In addition, we quantified the evidence to support the alternative hypothesis by computing a Bayes factor for the specific effect of crystallized and fluid intelligence on: (a) EF score; (b) total emotional recognition score; (c) ToM; and (d) inhibitory verbal control. Bayes factor analysis for each prediction will yield very strong evidence to support the alternative expected effect on main interest variables when BF10 > 30 and ≤ 100; strong when BF10 > 10 and ≤ 30; and moderate evidence if it is between BF10 > 3 and BF ≤ 10 [assuming a uniform distribution of priors; see (67, 68)].