Damian Sendler: Recently, research has increasingly focused on the interactions between multiple social determinants, interventions aimed at upstream causes of mental health issues, and the use of computer simulation models to represent complex systems. However, methodological difficulties and inconsistent findings prevent a definitive understanding of which social determinants should be addressed to improve mental health, and within which populations these interventions may be most effective.
Damian Jacob Sendler: Frameworks for social determinants study how people’s health is influenced by their living and working conditions [1]. Inequalities in life expectancy, child mortality, and disease burden among disadvantaged populations are thought to be driven by these conditions (i.e., social determinants) [1]. Conceptual frameworks for social determinants are built on the premise that there is a “social gradient,” in which people with lower social status face greater health risks and have a shorter life expectancy than those with higher social status [2]. Observed differences in social determinants can be reduced through targeted social and economic policies and programs.
Dr. Sendler: People’s socioeconomic status affects their risk of mental illness and their ability to access mental health services, which in turn improves outcomes. Allen and colleagues [4] used a multi-level framework that included: a life-course approach covering prenatal periods through old age; community-level contexts such as environment and health care systems; and country-level contexts such as political and economic factors, cultural norms, and specific policies [4]. Mental illness disproportionately affects the poor and disadvantaged, and stress and physical health act as multipliers of the negative effects of social determinants over time [4]. Other studies describe the impact on health over multiple generations of cumulative advantages and disadvantages [5].
Damian Sendler
A distinction between “upstream” and “downstream” determinants has emerged as social determinants frameworks have evolved. Upstream social determinants (e.g., economic opportunities) act as “fundamental causes” and typically have an impact on health through downstream social determinants, according to Braveman and colleagues [5]. (e.g., living conditions). Gender and race/ethnicity, as well as educational attainment, occupational status, and social support, are included in the definition of social determinants of health (p. 383). Racism and daily stress are also highlighted in this work [5]. Chronic stress affects mental health outcomes via biological pathways, according to Fisher and Baum [6]. Low socioeconomic status has been linked to mental health problems in lower socioeconomic groups, including stress from navigating everyday circumstances and anxiety about insecure and unpredictable living conditions.
There has been an increase in evidence in the last three years that social determinants have an effect on mental health outcomes in particular populations. Even in countries with universal healthcare, where employer-provided health insurance is less essential to accessing services, unemployment, precarious employment, and employment conditions continue to be linked to increased psychological distress [7, 8]. An increase in the rates of serious mental illness among migrant workers in Singapore was linked to hostile interactions with employers (i.e., injury disputes, threats of deportation). Similarly, nursing assistants who worked for profit-making companies and who were subjected to management dominance and emotional stress were more likely to endorse depressive disorders [12]. Other social determinants can be moderated by one’s employment status. According to some research, men’s mental health suffers more than women’s when they are unemployed [13]. According to researchers, nativity status and mental health among women working in Spain were linked by occupational social class (i.e., manual or non-manual labor) [14].
Damian Jacob Markiewicz Sendler: People with low incomes and a lot of financial stress are more likely to have poor mental health, according to studies in Sweden. It has been found in Korea, Europe and North America that people with other disadvantages have similar findings [7, 8, 10]. Katz-Wise and colleagues [17] found that transgender adults in the United States who had lower incomes were more likely to engage in self-harm, suicide attempts, and depression. Low-income pregnancies were linked to higher rates of depressive and anxious symptoms [18], but this link was partially mediated by financial stress (e.g., insufficient food, transportation, or housing). Chronic exposure to low-quality housing conditions (e.g., inadequate heating, overcrowding) has been shown to have a negative impact on the mental health of youth and adults [19, 20]. Mental health issues have also been linked to food insecurity and a poor diet [21–23] in the United States and Canada.
Damian Jacob Sendler
It has been shown time and time again that discrimination based on a person’s race/ethnicity, immigration status, sexual orientation, or employment status is harmful to one’s mental health [24–28]. Additionally, among African asylum seekers in Hong Kong [29] and Iraqis in Sweden [16], higher levels of depression and worse mental health were found to be associated with reported discrimination experiences. Discrimination that people in the United Kingdom believe they have experienced has been linked to increased levels of psychological distress over time [30]. According to Khan and colleagues [31], the “fundamental cause” of depression and a predictor of anxiety is multifactorial discrimination (i.e., based on multiple minority identities).
The quality of one’s mental health can be greatly influenced by the quality, or lack thereof, of one’s family relationships. Families that have a high level of satisfaction and connectedness are less likely to suffer from depression [7, 29]. As “reduced involvement” fathering (as opposed to “authoritative” fathering) has been linked to more internalizing and externalizing symptoms among Mexican youth in the United States [32], it is possible that parenting styles influence mental health. PTSD, anxiety, and aggression have been linked to a history of abuse and neglect from a family member [33, 34]. Social support, belonging to a group, and trust in others have all been linked to better mental health [9, 35, 7, 10, and 9], and it has been found that having a large social support network, including family and friends, reduces the risk of developing mental health problems, personality disorders, and psychotic experiences [36]. Migrants, refugees, and transgender people, for example, may find it especially helpful to participate in their communities and receive social support.
Damien Sendler: Community characteristics, such as urbanicity or neighborhood safety, have been frequently examined in contemporary social determinants research. Gay and bisexual men who live in rural areas are more likely to suffer from depression and other mental health issues than those who live in urban areas. Mental health outcomes can be predicted in large part by one’s perception and experience of one’s neighborhood’s safety [40, 41]. Dissatisfaction with living conditions and neighborhood safety were linked to lower depression levels among Chinese urban residents [42], while conflicts with local government over neighborhood planning were linked to higher depression levels [43]. U.S. researchers found no correlation between the quality of a neighborhood and the mental well-being of young people, even when controlling for other relevant factors [20]. Experiencing community violence as an adolescent has been linked to an increased risk of depression, anxiety, and post-traumatic stress disorder (PTSD) symptoms [40, 41]. Major depressive disorder and generalized anxiety disorder may be more common in people who live in areas with high prison admission rates in the United States [44]. According to Bor and his colleagues [45], a unique community-level predictor was examined—police killings of unarmed Black Americans in the state. People of color who lived in states where at least one such killing had occurred in the previous three months reported an increased number of days during which their mental health was “not good” in that research [45]. The same was found in a study of the 2014 unrest that erupted in Ferguson, Missouri, following the death of Michael Brown, who was shot and killed there [46].
Furthermore, recent studies have confirmed the importance of several fixed characteristics such as race/ethnicity, nationality, gender and sexual orientation in addition to studying dynamic social determinants of mental health. Certain mental health symptoms have been linked to racial/ethnic minority status [40, 30], while other studies have examined how race/ethnicity interacts with other variables to affect mental health. Racial/ethnic minorities frequently report poorer mental health than White respondents among LGBT adults in the United States, for example [31]. Black and Latinx ethnicity predicted higher post-traumatic stress in New York City residents affected by Hurricane Sandy [47]. The direction and magnitude of the relationship between race/ethnicity and psychological well-being depends largely on whether other social and health variables are included in the analysis, suggesting that race/ethnicity may play more of an indirect role in influencing mental health.
Nationality and status as a migrant have been shown to have negative effects on mental health around the world [15]. Further evidence suggests that, although migrants initially show better mental health than the native population, this advantage usually diminishes over time [10]. Parents of Latinx children in the United States who have experienced negative immigration-related impacts since January 2017 were more likely than other parents to report high levels of psychological distress [49]. A number of studies have found that women consistently report lower levels of mental well-being than men. Neurodevelopmental and disruptive and impulse control disorders may be less likely to meet diagnostic criteria for these individuals [13]. Transgender identity and sexual orientation continue to be associated with a range of behavioral health outcomes, including self-harm, suicide, depression, and other serious mental illness [17]. Although the differences in mental health based on fixed characteristics such as race/ethnicity and gender are important to note, they are more likely to be a result of oppression or discrimination rather than an inherent vulnerability.
Mental illness can affect social determinants such as homelessness, school dropout, marital instability, and economic insecurity, although this is less frequently discussed [52–54]. When one’s mental well-being is compromised, it affects one’s personal choices and, in turn, the living conditions that limit one’s options. The World Health Organization [55] has described how mental health symptoms can have a cumulative and dynamic impact on socioeconomic status and other social determinants at each stage of life using a life-course approach. I Multiple social determinants can intersect and contribute to the onset of behavioral health disorders in young adulthood, which is critical. Individuals’ ability to navigate social norms and structures, such as school, work, or the justice system, can be negatively impacted by mental health symptoms at this stage of life [56, 57, 58]. These risk factors can then limit future earnings, create barriers to socioeconomic advancement, and increase the likelihood of developing a mental illness. Another factor that contributes to social inequality is the lack of access or long-term connection to behavioral health services for young adults [59].
It has been shown that interventions aimed at increasing housing stability, community functioning, perceived well-being and quality of life, and increased self-esteem are effective for people with mental illness [61]. Programs like Individual Placement and Support (IPS) have been shown in a meta-analysis of interventions aimed at increasing employment rates to be highly effective [62]. However, IPS program implementation is hindered by a lack of funding [63]. Although Housing First programs have been linked to better housing outcomes, lower rates of inpatient hospitalization, and a more stable use of health services for people experiencing homelessness and mental health challenges, these programs did not significantly reduce clinical symptoms [61, 64]. Before providing housing, mental health needs should be addressed, according to studies [61].
In Canada, food insecurity rates have decreased as a result of social policies aimed at improving housing stability [65]. Individuals’ perceptions of government assistance may moderate the benefits of national programs like the Supplemental Nutrition Assistance Program, which has been linked to poor mental health outcomes [21–23]. National efforts to reduce poverty, such as the Earned Income Tax Credit, may reduce depressive symptoms and improve self-esteem among beneficiaries [67].
Mental health disparities can be reduced by community-based interventions that build community trust and safety, mitigate violence and crime, or improve neighborhood deprivation. [68, 69]. Several studies have found a link between better mental health and fewer depression symptoms when regional and national programs focused on urban planning (e.g., increasing public access to green spaces). Stress reduction, increased physical activity, and/or a stronger social network are all possible explanations for these findings [73]. Social inclusion and connection initiatives have also yielded promising results [74]. A number of behavioral health outcomes have been linked to programs that encourage participation in social media, social marketing, schools, primary care, and parental relationships [75]. An evidence-based prevention strategy that aims to reduce youth substance use, violence, and other problem behaviors community-wide is implemented through stakeholder coalitions [76]. In addition to reducing drug use initiation and delinquency rates, this strategy has also resulted in a reduction in overall projected health and justice costs [76].
Investing in and integrating social services with mental health care has been shown to have a positive impact on outcomes. When it comes to improving patient engagement and treatment utilization, community health workers (CHWs) have been credited [77] and CHWs have also successfully implemented interventions targeting the social determinants of mental health conditions [78]. When primary care providers link socioeconomically disadvantaged patients to appropriate social and cultural activities, there have been mixed results in terms of mental health benefits [79–80]. Many social prescribing studies, however, were found to be small in scale and to employ poor study designs in a recent systematic review. It is suggested that reducing disparities in mental health can be achieved by providing universal primary health care access, given that people in countries with universal health care have better emotional well-being [81, 82].
Social determinants affect mental health in a variety of ways, often over a long period of time, and at various levels within complex systems [69]. Beyond generalized linear models, researchers have used analytical strategies to estimate these non-linear and time-varying relationships. Complex systems can be simplified by using simulation models, for example, and they can be useful in understanding system dynamics related to social factors.
When clinical event timing (i.e. incidence, relapse) is critical, state-transition and network models can be particularly useful for simulating disease progression among populations [84]. For example, Scata and colleagues [85] used a state-transition model to model the spread of suicidal ideation among people with psychological distress. Their findings suggested that educating people about the dangers of suicidal behavior through prevention programs and social media campaigns could reduce the spread of suicidal thoughts. Heterogeneous social networks, made up of people from different socioeconomic backgrounds, may help to strengthen the network’s resistance to disease transmission [85].
Agent-based models (ABMs) simulate and evaluate the combined effects on a system of the actions and interactions of multiple agents (e.g., patients and providers) [86–88]. According to Silverman [88], an ABM model of a behavioral healthcare system was used to examine how changes in the patient’s employment status and/or provider workflow affected re-hospitalization rates and the number of days spent in the hospital. ABMs are more time and cost-effective than randomized control trials for testing intervention effectiveness and assessing policy impact [84].
Researchers must exercise caution when distinguishing “allowable” from “non-allowable” [89] differences in social determinants, especially when measuring mental health disparities, even though simulation techniques allow for this. Simulation should not be used to alter “allowable,” or “justifiable,” social determinants of health (such as age and sex). To account for differences in health caused by “non-allowable” determinants (such as employment or education), simulation adjustments may be necessary. Reweighting and propensity score matching were employed by Alegra and colleagues [90] to correct for “non-allowable” social determinants. There was a strong correlation between better mental health outcomes and an increase in employment; however, there was a weak correlation between an increase in education or income. Existing survey designs can easily incorporate these weight-based approaches, making them useful for weighted survey analysis.
Social determinants research in recent years has also used descriptive (unsupervised) and predictive (supervised) machine learning algorithms to interpret existing patterns and behaviors as well as to predict future events. Systems can learn the structure of input data without explicitly providing outputs, enabling the identification of previously unknown data patterns [93]. It has been shown that LGBT adolescents’ interview responses can be clustered into distinct social determinant clusters using k-means clustering methods [94, 95]. Mental health problems (such as anxiety, depression, suicidality, and other forms of psychological distress) were more common in children and adolescents who had little or no family support. While this is not the case, supervised learning methods do allow systems to learn a mapping function (such as classification or prediction models) when both input and output data are present [96]. There have been numerous studies that use penalized regressions, random forests, and neural networks to show how factors like income and social support affect health [92]. There is some evidence that nonclinical data about individual and community-level social factors can be used to predict mental health outcomes and service need, but performance improvement results have been mixed [97, 92]. When compared to traditional models, machine learning algorithms are able to account for more complex, dynamic relationships, and thus identify new social factors [98].