Depressive Symptomology, Identity and Religious Practices among Catholics and Evangelicals: Differences between the Mapuche and Non-Indigenous Chilean Population

: Background: Chile is a highly religious country. Although a majority of the population describes itself as Catholic, there has been a substantial growth in Evangelism, especially among indigenous people. In this context, the aim of this study is to analyse the relationship between Catholic and Evangelical religiosity in terms of identity and practices and depressive symptoms in the Mapuche and non-indigenous Chilean population. Methods: The study was conducted using secondary data from the Longitudinal Intercultural Relations Study of 2017, estimating linear regressions to explain variations on the PHQ-9 scale between the adult Mapuche and non-indigenous Chilean population by ﬁrst including the controls variables, followed by religious identiﬁcation, churchgoing, and prayer. Results: Social support, good health, and age showed a negative association with PHQ-9 in both groups. Being a woman and not having a partner were only positively related with depression in the non-indigenous group. A negative association was found between Evangelical religious identity and depressive symptoms among the Mapuche population, while churchgoing was negatively associated and prayer was positively associated with depression in the non-indigenous group. Conclusions: The ﬁndings conﬁrm that religiosity is a protective factor against depressive symptomology in the Chilean population. However, the analysis reveals signiﬁcant ethnic differences.


Introduction
Catholicism has been considered the dominant religion historically in Latin America, and particularly in Chile. However, the 20th century saw a growing process of religious conversion and mobility (Camargo Martínez 2019) at the same time as a rapid development of secularization (Morello and Rabbia 2019). The new religious landscape in Latin America reflects an increased number of Evangelicals, particularly Pentecostals and Neo-Pentecostals.
It is well known that Catholicism became dominant in Latin America during the Hispanic and Portuguese colonial era that spanned over three centuries, when the Catholic Church gave its support to the Crown, which in turn offered the Church its protection. This made the Catholic Church a cornerstone of the colonial social order, and Catholicism retained its supremacy during the creation of the various Republics. The shift toward different religions in Latin America started with the arrival of Protestantism at the beginning of the 19th century (Morello and Rabbia 2019). The region currently hosts a range of religious persuasions, to the extent that it can even be described as religiously pluralistic (Camargo Martínez 2019;de Roux 2012). There has been a notable growth in the number of Protestants, which has doubled or even tripled in some countries. Brazil is an example of this, with the percentage of Protestants in the population growing from 7. 8% in 1960to 17.4% by 1985, and to 23% by 2010(Pew Research Center 2017.
This situation transcends age groups and even ethnic groups, including indigenous peoples, producing a need to examine the repercussions of religious diversity for psychosocial wellbeing in Latin American societies. It is in this context that the present study will specifically consider how Catholic or Evangelical religiosity is related to mental illness in the indigenous and non-indigenous Chilean population. We start by describing the context of religiosity in Chile before briefly explaining the empirical evidence on the impact of religiosity on psychosocial wellbeing in different groups and the new religious structures among the indigenous Chilean population.

Background: Chile as a Religious Country
Chile is a religious country, as shown by the finding reported in the latest Bicentenary Survey (UC 2019) that 76% of Chileans believed in God. Catholicism remained in a strong position among the religions practised, but there was an abrupt fall from 70% of the population professing Catholicism in 2016 to 45% in 2019. The second-largest religion in Chile is Evangelism, representing 18% of the population, while 5% stated that they had another religion (Mormon, Adventist and so on) and 32% declared that they did not follow an organised religion.
Catholicism is also dominant among indigenous peoples. However, there has also been a fall in its prevalence among this group, with a trend toward Evangelism or Protestantism. Specifically, 72% of Andean peoples (including Aymara, Quechua, and Atacameño) described themselves as Catholic in 2016, falling to 65% in 2018. Evangelism or Protestantism in this ethnic group rose from 14 to 15% over the same period of time. The change is more marked in the context of the Mapuche, with Catholicism falling from 52 to 43% (between 2016 and 2018), while Evangelism or Protestantism rose from 32 to 36% (Centro de Estudios Interculturales e Indígenas 2019).
Together with this change in religious beliefs, increasing migration into Chile, for example, from Haiti, has also given rise to a change in religious diversity (Orellana 2021). In light of all of this, Chile can be seen to be experiencing an increasing process of religious diversification, with a substantial increase in the secularization process among its population.

Religiosity in Chile and Its Relation to Mental Illness
It is necessary to specify that mental health was operationalised by Keyes (2002, p. 208) as a syndrome of symptoms of positive feelings and positive functioning in life. In this sense, mental health is not only focused on the absence of illness, but also as the presence of positive functioning; therefore, it includes positive elements such as emotional, psychological and social wellbeing and, on the other hand, mental illnesses such as depression, anxiety, panic disorder, alcohol dependence, among others (Keyes 2005).
Understanding that mental health is a broad construct, however, in this study, we focus on mental illness. Studies conducted in Chile regarding the impact of religiosity on the mental illness of the population tend to focus on the social groups at greater risk of experiencing psychosocial problems. This includes people with suicidal thoughts, groups at high risk of abusing drugs and/or alcohol, people who have experienced stressful life events, and groups with a higher predisposition to suffer depressive symptoms. These studies consider religiosity and/or spirituality as a protective factor that aids coping with and mitigating the consequences of stressful events, as well as reporting that stressful experiences can lead to a development of religion or spirituality (Shaw et al. 2005).
In line with the classic approach of Pargament (1997), religious coping involves the use of religious beliefs and behaviours to prevent and/or mitigate the negative consequences of stressful life events and to help resolve problems. These strategies may be individual, as in the case of prayer, or take place in group spaces, such as group prayer or participation in religious rituals or services. As observed by Pargament et al. (1998), it is important to distinguish positive religious coping, which includes seeking spiritual support, religious forgiveness, collaborative religious coping, spiritual connection, religious purification, benevolent religious reappraisal, and religious focus, from negative religious coping, which encompasses spiritual discontent, punishing God's reappraisals, interpersonal religious discontent, demonic reappraisal, and reappraisal of God's powers. People appear to use various methods of religious coping in combination with each other; that is, they apply different configurations of religious thought, feeling, behaviour, and relationships in their efforts to deal with major life stressors (Pargament et al. 1998, p. 720).
In this regard, a study by Florenzano et al. (2014) involving women with depressive symptoms who had experienced suicidal thoughts or attempted suicide suggested that participants who stated that they were not believers had more extensive histories of suicide attempts and more family members who had committed suicide. Meanwhile, Delva et al. (2015) reported spirituality to be an unexpected protective factor against the consumption of alcohol, given that religious practice in and of itself, without significant faith, was not associated with the consumption of alcohol among Chilean adolescents. In the context of highly stressful experiences, research has been conducted in Chile with people who lost their homes following the 2010 earthquake. In this respect, a study performed by García et al. (2014) confirmed that positive religious coping together with perceived social support had a positive influence on post-traumatic growth (PTG). In the same study, positive religious coping was reported as mediating the relationship between subjective severity and PTG; in other words, following a stressor such as an earthquake, the seriousness of the experience does not have a direct effect on PTG, but one that is mediated by positive religious coping.
However, most empirical evidence regarding this topic is focused on older adults and the implications of religiosity/spirituality for their wellbeing. Older adults in Chile face ageing with higher levels of socioeconomic disadvantages owing to their small pensions, giving rise to precariousness in old age (Moyano-Díaz and Mendoza-Llanos 2020), in addition to a social protection system that is mainly inefficient in the area of health and dependency (Palacios et al. 2020). Following this line of argument, a study by Gallardo-Peralta and Sánchez- Moreno (2014) suggested that a combination of religious beliefs and practice was related to a lower incidence of depressive symptoms among older adults. Moreover, being a member of a high-cost or doctrinally strict religion (Jehovah's Witnesses, Mormons) did not represent a protective factor against depressive symptoms when compared with a group of non-religious older adults. The same authors published a later article confirming that only spirituality was related to fewer depressive symptoms; they defined this term as a "singular, specific and personal experience with a higher being, which can be identified as a psychological process in the personal search for meaning and purpose in life" (Gallardo-Peralta and Sánchez-Moreno 2020, p. 172). A recent study by Fernández and Valenzuela (2020) confirmed that religiosity has a positive impact on subjective wellbeing in old age, in particular, attendance at religious services improves access to social networks and support structures, and the prayer enables the resignification of failures and misfortunes and allows a comprehensive framework for the interpretation of world events, all of which promote wellbeing. Finally, Fernández and Rosell (2021) analysed the different types of religiosity and depression and anxiety among older people, observing a direct effect between organisational religiosity (beliefs, practices, and rituals in the religious community) and depressive symptomology, and between intrinsic religiosity (religious relevance, commitment, and motivation) and anxiety-related symptomology.

Indigenous Communities in Chile: The Mapuche and Religiosity
Ten indigenous groups and one Afrodescendant tribal group are recognised in Chile; 9.5% of the population state that they are a member or a descendant of these groups. The Mapuche are the largest group, equivalent to 84.8% or 1,437,308 personas (Ministry of Social Development and Family 2017).
Mapuche means "people of the earth", and the Mapuche people were indeed originally hunter-gatherers. Their territorial identity emerges from the territories in which their population has historically been located: nagche are the northern people, wenteche are the people of central valleys, lafkenche are the coastal people, pewenche are from the foothills, and huilliche are the southern people, which includes the urban Mapuche identity. The Mapuche live in rural areas that are connected to large cities, having experienced significant migration to urban settings.
The Mapuche people occupied the central and southern territories of present-day Chile until the arrival of the Spanish. They were subsequently confined to south of the Bío Bío Region as a result of the so-called Pacification of La Araucanía by the Chilean State toward the end of the nineteenth century. The Chilean State then imposed settlement and division processes on communities, breaking up the territory into small communities that today form a fundamental part of the social, cultural, economic, and identity dynamic of their population. Changes in urban/rural composition from the 1980s and 1990s onwards meant structural shifts in terms of location and sociocultural organization (Briceño-Olivera et al. 2020).
The Mapuche are currently characterised by this significant migration from rural areas, abandoning their ancestral territory (Wallmapu) to move to various cities in Chile, mainly Santiago. This has produced various consequences in terms of inter-and intra-ethnic relations, the restructuring of the Mapuche cosmovision, and even a loss of identifying features (such as the language, Mapuzugun) and urban segregation processes, among others (Gissi 2001;Imilan 2017;Imilan et al. 2016;Valenzuela and Aranis 2017). This has been combined with an increasing shift in religious culture from Catholicism toward Evangelism (Centro de Estudios Interculturales e Indígenas 2019).
As Valenzuela and Aranis (2017) state in response to the loss of identity linked to the land (tuwün), religious practices offer the Mapuche the opportunity to know their history and retain a sense of community (lovche). Mansilla and Orellana (2019, p. 114) add that Pentecostalism facilitates the reproduction of the traditional order of indigenous rural populations in urban environments, acting as a form of sociocultural refuge. To understand this process, we shall briefly examine the history of Evangelism and the social and cultural foundations that enabled it to grow rapidly among the Mapuche.
Evangelism in Chile is strongly represented by the Pentecostalist movement. Historically, it developed through the Pentecostal Methodist Church around 1909 and 1910, mainly driven by the female spouses of pastors or leaders who preached in the streets. It arose through popular protest and the struggle for the right to participate in religious work, not merely as "workers", but also as "ministers" (leaders). There was a growing demand for social mobility in religious work and a call for the inclusion of popular expressions, symbols, and customs in Protestant liturgy. From these origins, Pentecostalism would be interpreted as a religion of the poor (Mansilla and Orellana 2019, p. 103).
Pentecostalism has been linked to the Mapuche since 1911 as a result of Pentecostal missions to territories with large indigenous populations, as well as the migration of indigenous peoples to cities. Urban indigenous peoples tended to live in poor and marginalised neighbourhoods, where Pentecostalists had their churches or preached in public. Whenever an indigenous person converted to Pentecostalism, the movement would push them to take the new religion to their land and particularly to their immediate family, with the result that Pentecostalism spread rapidly among these indigenous peoples (Mansilla and Orellana 2019, pp. 113-14).
When examining the social foundations of this extensive spread, Mansilla et al. (2014, p. 155) suggested that, as Pentecostalism seeks to incorporate excluded or segregated social groups, such as the urban indigenous population, these groups see a chance for dignity, survival, and social mobility through the Pentecostal message. Pursuing this line of argument, Kessler (1967, p. 324) argued that indigenous Pentecostalism finds human warmth in religious gatherings. The participation of the congregation in religious services provides an opportunity to belong to a group, while extensive secular work offers the chance for employment and social mobility.
From a cultural perspective, Pentecostalism is presented as a replacement indigenous community. In other words, indigenous people who have migrated to the city find a place that enables them to relive a lost space and time through the Pentecostal community (Kessler 1967;Mansilla et al. 2014). In this regard, D'Epinay (1968, p. 67) argued that the Pentecostal community reproduces what was found in Mapuche society: the priest and elders correspond to caciques (indigenous leaders) and heads of family, while the Pentecostal body of prophets corresponds to machis (indigenous doctors), who are both prophets and healers.
The growing process of the conversion of the Mapuche community to Pentecostalism continues to have a sociocultural explanation, since this religion offers a substitute community space that allows the reproduction of the traditional order of the rural Mapuche community, a united, self-sufficient, closed community, with solid values and alien to the processes of hybridisation (Mansilla and Orellana 2019, p. 114).

Present Study
It is useful to briefly examine the context of mental health symptoms in the Chilean population to understand the importance of investigating religiosity as a psychosocial resource for enhancing it. Chile has a high prevalence of depressive symptoms and the ensuing repercussions for secondary morbidity (Dagnino et al. 2017), meaning it is considered one of the main public health issues (Andrade and Espinoza 2019; Martínez et al. 2017). The most recent National Health Survey (2016-2017) reported that 15.8% of the population aged over 18 years presents signs of depression, while the national prevalence of depression is at 6.2%. There is a marked difference based on sex, with a prevalence of 2.1% among men and 10.1% among women (Ministry of Health 2017).
Considering this context, this study seeks to focus its analysis on one aspect of mental health, namely depression. However, based on the available empirical evidence, there is a need to continue to examine the positive or negative effects of Catholic and Evangelical identity and practices among Chile's population, distinguished by ethnic group. As we have noted, the indigenous peoples of Chile are highly religious, but there has been an abrupt shift in religious adherence in recent years, linked to aspects of identity and social integration (Gavilán 2019;Lindhardt 2011;Mansilla and Muñoz 2017). As a result, the aim of this study is to analyse the relationship between Catholic and Evangelical religiosity (identity and practices) and depressive symptomology in the Mapuche and non-indigenous populations, controlling for different variables including social support, perceived health, income, sex, age, and partner. We expect to observe certain differences in the groups subject to analysis (Mapuche and non-indigenous), in line with the following hypotheses: Hypothesis 1 (H1). Evangelical religious identity is negatively associated with depressive symptomology in the Mapuche population.
Hypothesis 2 (H2). Attendance at religious services is negatively associated with depressive symptomology in the Mapuche and non-indigenous populations.

Hypothesis 3 (H3).
Prayer is positively associated with depressive symptomology in the Mapuche and non-indigenous populations.

Hypothesis 4 (H4).
There is a negative association between the control variables-social support, positive perceived health, higher income, being male, age and having a partner-and depressive symptomology in the Mapuche and non-indigenous populations.

Data Source and Sample
This study used secondary data from the Longitudinal Intercultural Relations Study (ELRI for its initials in Spanish) performed by the Centro de Estudios Interculturales e Indígenas. ELRI is a longitudinal survey with the main purpose of reporting on intercultural relations between the main indigenous groups and the non-indigenous population in Chile. It covers themes of identity, family, intergroup contact, conflict, public policy, migration, and wellbeing. More specifically, this study used ELRI second-wave data. This survey was conducted in person with indigenous and non-indigenous Chilean people aged 18 years and over from 14 September to 18 December 2018.
Probabilistic, random and stage-stratified sampling was used. First, 120 municipalities in Chile were selected (from a total of 346), chosen due to having an indigenous population of over 9%, or more than 0.04% of the total national indigenous population. A "mirrormatch sample" design was then used, involving the random selection of indigenous and non-indigenous participants aged 18 years or over in the same selected blocks or rural areas within the municipalities. The final sample size was 2879 cases, with an 80% response rate.

Dependent Variable
The dependent variable corresponded to the Patient Health Questionnaire (PHQ-9) for depressive symptoms (Kroenke et al. 2001). This scale scores the nine criteria of the Diagnostic and Statistical Manual Fourth Edition (DSM-IV), with scores ranging from 0 (not at all) to 3 (nearly every day): (i) little interest or pleasure in doing things; (ii) feeling down, depressed, or hopeless; (iii) having trouble falling asleep or staying asleep, or sleeping too much; (iv) feeling tired or having little energy; (v) having a poor appetite or overeating; (vi) feeling bad about yourself, or that you are a failure, or that you have let yourself or your family down; (vii) having trouble concentrating on things (e.g., a newspaper or the television); (viii) moving or speaking so slowly so that other people have noticed, or alternatively being fidgety and restless, and (ix) having thoughts that you would be better off dead, or considering hurting yourself in some way. The depression severity is graded as: None (0-4), mild (5-9), moderate (10-14), moderately severe (15)(16)(17)(18)(19), and severe (20-27). This questionnaire has been validated in Chile for a sample of adults at primary care centres (Saldivia et al. 2019).

Control Variables
The control variables were available social support (Multidimensional Scale of Perceived Social Support (MSPSS) from (Zimet et al. 1988)), perceived health (What score would you give to your current state of health on a scale from 1 to 7? with responses re-codified as 1. Poor/regular health (5 or lower) and 2. Good health (6 and 7)); and income (which is an ordinal variable that measures the range of average monthly household income, with responses ranging from 1. Up to CLP 215,000 to 10. CLP 1,800,000 or more) 1 . The evidence was considered to show a strong association between these variables and psychosocial wellbeing (Diener et al. 2018;PNUD 2012). Sociodemographic variables such as sex, age, and partner were also included as control variables.

Independent Variables
The following variables were included in relation to religiosity: 1.
Religious identification: Which religion or church do you belong to or feel closest to? Responses to this question were re-codified as 1. Catholic; 2. Evangelical; 3. None or other religion (reference category).

2.
Attendance at religious services: How often do you attend religious services at your church or practise your faith? Responses to this question were re-codified as 1. Never or a few times a year (reference category); 2. A few times a month; 3. Weekly or several times a week.
Prayer: How often do you pray outside of religious services? Answers here were re-codified as 1. Never or only on special occasions (reference category); 2. On some days but not others; 3. Every day or several times a day.

Statistical Analyses
The statistical programme STATA (version 17) was used to carry out analyses in this study. First, the authors performed a descriptive bivariate analysis (of percentages, means, and standard deviations). Different linear regression models were then calculated to explain variations in the PHQ-9 among Mapuche and non-indigenous Chileans aged 18 years and above. More specifically, the sample was sub-divided into two groups, the Mapuche and the non-indigenous populations, based on a self-identification question regarding whether the participant belonged to that indigenous group. As the analysis in this study was focused on this indigenous group in particular (compared with the non-indigenous group), cases where respondents self-identified with other indigenous groups were excluded from the analysis and the sample size was reduced from 2879 to 2413 cases. An initial model was subsequently calculated for each group (Mapuche and non-indigenous) that included only the control variables. The second model incorporated the religious identification, attendance, and prayer. To evaluate the goodness-of-fit of the models, the authors used a variance analysis (ANOVA) in addition to the adjusted R 2 , which considered the number of independent variables in each model. Table 1 presents the descriptive statistics for the variables that were included in both groups of models. The total sample comprised 66.1% women, with an average age of 50.8 years (SD = 17,135), while 66% of respondents stated that they had a partner. Self-declared non-indigenous respondents made up 48% of the sample, with the remaining 52% selfdefining as Mapuche. The average PHQ-9 scale score was 4.5 points (SD = 5,440). With respect to the religiosity variables, 55% of the sample defined themselves as Catholic and 32% as Evangelical. Frequent attendance at religious services (weekly or several times a week) was reported by 23% of respondents, and over half prayed every day or several times a day. There appeared to be generally positive levels of perceived social support (an average of 47.8 on a scale from 12 to 60), while around half of respondents perceived themselves to be in good health. Income was reported with an average score of 3.37 (SD 2.122), where category 3 had a value of between CLP 300,001 and CLP 390,000 and category 4 was between CLP 390,001 and CLP 470,000.

Descriptive Analysis
There were some significant differences in these variables when the Mapuche group was compared with the non-indigenous group. First, the Mapuche group had a lower average for the PHQ-9 scale than the non-indigenous group (4.28 vs. 4.8 points, respectively). A higher percentage of Catholic identification was found among the non-indigenous group (62% vs. 49%), in contrast to a higher percentage of Evangelical identification among the Mapuche (38% vs. 24%). The Mapuche group also reported higher levels of frequent attendance at religious services than the non-indigenous group (26% vs. 20%). In relation to living conditions, although the non-indigenous group reported higher income levels (3.7 vs. 3.0), the Mapuche group had higher perceived health (57% vs. 47%), and availability of social support (48 vs. 47 points).  Table 2 presents the results of the estimated regressions on the PHQ-9 scale for Mapuche and non-indigenous Chileans aged 18 years and above. All the models were significant according to ANOVA. In both cases, the full models (models 2a and 2b) had the best goodness-of-fit, achieving 13% and 11%, respectively, of the explained variance for the PHQ-9 scale.
Then, it was observed in model 2b that self-identifying as Evangelical compared with having no religion or another religion was negatively associated with the PHQ-9 scale, with a value of β = −1.572, p < 0.01, 95% CI [−2.689-−0.4553] being the only significant religiosity variable.

Discussion
In this study, we have observed that religiosity, whether through identity or religious practices, has a protective role against depressive symptoms in the Chilean population, which is consistent with previous research (Fernández and Rosell 2021;Florenzano et al. 2014;Gallardo-Peralta and Sánchez-Moreno 2014). However, religious experiences differ according to ethnic identity (Gavilán 2019;Mansilla and Muñoz 2017), showing the importance of examining the various ways of experiencing religiosity according to cultural ethnic identity. However, research into the relationship between religion and ethnicity has a prominent place in social sciences, with much of it exploring the symbiotic nature of the relationship: these two social forces can be closely interlinked and generally reinforce each other (Calvillo and Bailey 2015, p. 57). Curiously, few studies analyse the impact of religiosity, particularly, the adherence to new religions, on the mental illness of indigenous Chilean communities.
Returning to the hypotheses of this study, our findings confirm that an Evangelical religious identity is negatively associated with depressive symptoms in the Mapuche population (H1). This is perhaps the key finding of this research, given that Evangelical religious identity is a social integration mechanism for a group that has historically faced discrimina-tion and even violence from the Chilean State (Manríquez-Hizaut et al. 2018;Moloney 2010). In this religious space, the Mapuche people have found a social and cultural recognition that promotes some aspects of their mental health. As stated by Lim and Putnam (2010, p. 285), religious beliefs provide a sense of meaning and purpose during difficult periods in life, which help with psychological integration, generally promote a positive, optimistic, and hopeful worldview, and offer a support community that does not exclude people due to their financial, social, physical, or mental circumstances. More specifically, the Evangelical religion is a resource with which to cope with difficult circumstances (Weber and Pargament 2014) for the Mapuche people, who experience situations involving exclusion, segregation, and the loss of their cultural identity (Mansilla et al. 2014).
The findings also confirm that religious practices are associated with depressive symptoms, but only among the non-indigenous population (H2 and H3). However, there are different relationships for group and individual practice. While attendance at religious services is negatively associated with depression, prayer is positively associated. These findings confirm the multidimensional nature of religiosity and that it can be positively or negatively associated with mental health (Weber and Pargament 2014). Religion has traditionally been analysed in terms of its social nature, notably in sociological classics, such as Durkheim (1982) and Weber (1987). In other words, it has been examined as a social and participative mechanism (Lim and Putnam 2010) that is organised and implemented within a community of people who have common beliefs and practices about the sacred (Koenig 2009, p.284). Maintaining active practices such as attendance at religious rituals (mass, services) hence becomes a resource to prevent depressive symptoms (Balbuena et al. 2013;Weber and Pargament 2014) because it involves a common experience with similar people, with whom one can share interests, beliefs, and values, establish relationships of friendship and emotional intimacy, and receive social support (Ellison and George 1994). A mutually supportive community is therefore created. As stated by Lim and Putnam (2010, p. 927), people with religious ties are more satisfied with their lives because they frequently attend religious services and construct intimate social networks in their congregations. However, it is to be expected that people who have more mental illness problems, in this case with more depressive symptomology, will tend to pray or worship more frequently and make use of this resource to obtain guidance in overcoming their circumstances, taking into account that religion provides models in the form of sacred texts that help with the acceptance of suffering and give people a sense of indirect control over their circumstances, reducing the need for personal control (Lim and Putnam 2010, p. 285). Developing this idea, the positive beliefs that religion offers provide people with consolation, meaning, and a sense of control and hope (Weber and Pargament 2014, p. 360). This means that religious practices thrive in situations of great personal trauma (Shaw et al. 2005).
As we have discussed, the results of this study, in the case of religious practices, such as service attendance and prayer, are only related to depressive symptoms in the non-indigenous population. This reaffirms the idea that religiosity among the indigenous population is mainly a community experience linked to identity, rather than religious practices per se, as it has been suggested that Pentecostal religion is a space of reproduction and continuity to the sociocultural and identity organization of the indigenous community in urban spaces (Mansilla and Orellana 2019). However, it is necessary to further deepen this analysis between identity and religious practices to understand in a deeper way the growing rise of Pentecostalism among indigenous people.
With respect to the control variables (H4), social support showed a negative association with depressive symptoms in both ethnic groups, confirming that it is a psychosocial resource, whose direct or buffering effect promotes health and wellbeing (Barrón and Sánchez-Moreno 2001), ensuring a sense of social integration and satisfaction of basic social needs through interaction with others (Thoits 1986(Thoits , 2011. These results are consistent with prior research in Chile (Jiménez-Molina et al. 2021;Sandoval et al. 2019), including among indigenous communities (Gallardo-Peralta et al. 2015). A negative association between good perceived health and depressive symptoms was also observed for both ethnic groups.
Self-perceived health is known to be one of the most valuable and reliable indicators of health, given that it predicts outcomes such as mortality and impairment in functional capacity (Idler et al. 1999); this has been confirmed for the Chilean population ) and among indigenous communities in Colombia (Tuesca-Molina and Amed-Salazar 2014). Age is another control variable that was present in both groups: depressive symptomology was lower at older ages. These findings contradict previous studies (Sivertsen et al. 2015), which reported that depression was more common among older adults, including older Chileans (Madero-Cabib et al. 2021). There was a significant association with depression among the non-indigenous population for the sex and partner control variables. Being a woman was associated with having more depressive symptoms. A recent study by Jiménez-Molina et al. (2021) confirmed significant gender-based differences in the prevalence of depression in Chile; specifically, depression rates were 23.2% among women and 13.4% among men. Having a partner was also associated with fewer depressive symptoms, and the empirical evidence has generally shown that having a partner and being satisfied with that relationship are associated with lower depression (Whitton and Whisman 2010).
It is also worth analysing how the Catholic religion is losing popularity in the Chilean population and especially in indigenous communities. Respecting the findings of this study shows that being Catholic is not a protective factor for mental illness (depression) among the Mapuche. Both elements are undoubtedly connected, i.e., Catholicism has decreased its adherence in indigenous communities because it loses its strength as a space of salvation and protection. This growing process of conversion is partly explained by the current crisis experienced by the Chilean Catholic Church due to the numerous allegations of child sexual abuse, which situate it as an institution that promotes abuse and cover-up (Ramírez 2018), to which Silva (2019) adds that the depraved and scandalous behaviour of many of its members, precisely those who lead and govern the Catholic Church, is long-standing and that there are still no effective and credible responses to reverse this distrust.
The limitations of this study are mainly related to the transversal nature of the data used, which prevents the establishing of causality between religiosity variables and psychological wellbeing. Along the same lines, some of the relationships, such as the one between depressive symptoms and prayer in the non-indigenous population, may be endogenous in the sense that, as mental health worsens, prayer may become a resource to cope with situations and find consolation. Future studies could contribute a longitudinal perspective to help shed light on the direction of these relationships. Together with this, there is a need to examine the various religious experiences and multiple practices (beyond prayer or attendance at services) among ethnic minorities, as in the case of indigenous peoples, and how these new religious structures are impacting their wellbeing in terms of the construction and maintaining of their ethnic identity.
Finally, to be able to delve deeper from qualitative research through observation in worship services and in-depth interviews with people with a high religious commitment. What mechanisms does Pentecostal religion offer to promote mental health in groups at social risk such as indigenous people? By this, we mean the possibility of addressing the cognitive mechanisms of mind/body dualism, teleological thinking, or the attribution of causality to nonsensory agents. This line of research, which has not been carried out among indigenous peoples in Latin America, may provide clues to the substantial increase in evangelical religiosity in this mystical encounter of supernatural beliefs between indigenism and Pentecostalism

Conclusions
We have confirmed in this study that religiosity is a resource that promotes mental health, specifically depressive symptoms, through positive religious coping, community and support, and positive beliefs (Weber and Pargament 2014), and that its multidimensional nature makes it a mechanism that enhances wellbeing in accordance with the psychosocial needs of different ethnic groups. As we have observed in this study, an Evangel-ical religious identity is a protective factor for the Mapuche people, while mental health among the non-indigenous population is promoted by religious practices, attendance at services, and prayer.