Wealth and depression: A scoping review

Abstract Introduction The inverse relation between income and depression is well established. Less is understood about the relation between wealth and depression. We therefore conducted a scoping review to answer the question: What is known from the existing literature about the relation between wealth and depression? Methods We searched for studies and articles in Medline (via PubMed), Embase, PsycINFO, PsycArticles, EconLit, and SocINDEX from inception through July 19, 2020. Ninety‐six articles were included in our review. Key article characteristics were year of publication, sample size, country, study design, definition of depression, definition of wealth, and association between wealth and depression. Thirty‐two longitudinal articles were included in a detailed charted review. Results Depression was defined in a relatively standard manner across articles. In contrast, definitions and measurements of wealth varied greatly. The majority of studies in the full review (n = 56, 58%) and half of the studies in the longitudinal charted review (n = 16, 50%) reported an inverse relation between wealth and depression. The longitudinal charted review showed that (1) macro‐economic events influenced depression, (2) wealth status influenced depression across the lifecourse, (3) wealth protected against depression in the face of stressors such as job loss, (4) subjective or psychosocial factors such as perception of wealth, relative comparison, and social status modified the relation between wealth and depression, and (5) savings interventions were successful in reducing depression and varied by context. Conclusion These findings suggest that wealth should be included in our consideration of the forces that shape mental health.

contexts that shape individual lives (Allen et al., 2014). Having financial resources can protect people from the stressors that drive depression and can lighten the blow of disruptions to daily living (Hammarström & Virtanen, 2019). Although there is abundant evidence that higher income is associated with better mental health (Kawachi et al., 2010;Patel et al., 2018), the relation between wealth and mental health is far less clear. By wealth, we refer to the total non-income accumulated assets that persons own or have access to, including financial assets such as money in savings accounts or stocks and physical assets such as home ownership. Wealth can be passed intergenerationally and may be a better indicator of financial stability than income and may also better document inequities than income.

Rationale
There is little question that low income is one of the central determinants of health (Kawachi et al., 2010;Krieger, 1994;Krieger et al., 1997;O'Donnell et al., 2013). Relatedly, but not directly addressing the question of interest here, Patel and colleagues conducted a systematic review of the literature on income inequality and depression and found that increases in income inequality were associated with increases in depression. The effects varied across subpopulations, with women and low-income persons experiencing worse impacts of income inequality on depression (Patel et al., 2018). However, a growing literature suggests that wealth may be an even more important factor in shaping health (Erixson, 2017). As a representation of accumulated assets over time, potentially shared intergenerationally, wealth reflects a feature of financial stability and socioeconomic advantage that can protect individuals throughout their lifecourse. While there have been several reviews aimed at better understanding what is known about the relation between wealth and physical or mental health (Baum, 2005;Braveman & Gottlieb, 2014;Pollack et al., 2007), we are not aware of any review that has specifically focused on the relation between wealth and depression. In a review of studies about wealth and a range of health outcomes, Pollack and colleagues identified six studies between 1990 and 2006 that explored socioeconomic indicators and depression (Pollack et al., 2007). However, the review did not focus specifically on depression and provided only high-level summaries.
Two population level trends in particular motivate this review: an increase in wealth inequality due to the unequal economic burden created by the coronavirus disease 2019 (COVID-19) pandemic and an increase in reporting of depressive symptoms during the COVID-19 pandemic (Ettman et al., 2020a;Luo et al., 2020). Recent studies suggest that persons with low wealth may be at greater risk of depression when the COVID-19 pandemic ends (Ettman, Cohen, Abdalla, et al., 2020;Ettman et al., 2020b;Ettman, Cohen, Vivier, et al., 2020). There remains, however, a lack of clarity about the potential effect of wealth on depression. Understanding how wealth relates to depression will be an important next step, and a useful tool for researchers and policy makers, as countries grapple with a high prevalence of depression and economic distress in the coming years.

Objectives
We conducted a scoping review to systematically understand the research about wealth and depression and to identify gaps in the literature about wealth and mental health. In particular, the goal of this study was to better understand the relation between depression and wealth as a liquid asset, as opposed to wealth in the form of home ownership or debt, which have been covered in other reviews (Richardson et al., 2013;Tsai, 2015;Tsai & Huang, 2019). Our research question was the following: what is known from the existing literature about the relation between wealth and depression?

Protocol
We evaluated the peer-reviewed published literature from the beginning of search functions through July 19, 2020 on financial assets and depression. We followed the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) (Tricco et al., 2018) guidelines.

Eligibility criteria
Eligible studies were English-language, quantitative analyses on human subjects that focused on objective financial indicators (e.g., wealth, assets, savings) and depression. We excluded articles solely about income, nonliquid assets (e.g., home ownership), and debt. The relation between home ownership and mental health has been well documented in other reviews (Tsai, 2015;Tsai & Huang, 2019;Vásquez-Vera et al., 2017). Similarly, the relation between debt and mental health is complex and has been addressed elsewhere (Richardson et al., 2013).
Studies were excluded if they did not feature depression or psychological distress. Studies were also excluded if they featured postpartum depression, since postpartum depression is a distinct clinical event that is separate in nature from other forms of depression (Di Florio & Meltzer-Brody, 2015). The literature suggests differentiating between postpartum depression and depression experienced during other times of life, given their different pathways and unique features (Batt et al., 2020). We did not include studies with ecologic data for depression (i.e., we excluded studies that did not assess depression at the individual level).

Information sources
We searched the following databases from inception through July 19, 2020: Medline via Pubmed, Embase, PsycINFO, PsycArticles, Econ-Lit, and SocINDEX. We exported the final search results to Zotero and EndNote, where duplicates were removed. Abstract screening was conducted using Abstrackr (Rathbone et al., 2015). The search strategy was created in partnership with a Medical Sciences Librarian; the F I G U R E 1 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flowchart of study selection process full search strategy can be found in Appendix SA. The database search was supplemented by additional searches using Google Scholar and the snowball technique of reviewing references of seminal papers.

Selections of sources of evidence
We reviewed all abstracts for relevance. CKE sequentially reviewed titles, abstracts, and full articles for inclusion in the final full review.
We excluded duplicate articles and articles that did not meet eligibility criteria. Reasons for exclusion from the full review were the following: studies did not report on the relation between depression and wealth (n = 31), duplicates (n = 27), articles were not quantitative studies (n = 8), studies were not conducted on human subjects (n = 2), postpartum depression was measured (n = 1), and could not locate the full article (n = 1). A flowchart of the study selection process can be seen in

Data charting process
A chart for mapping the data was created to determine variables to extract. The data were charted and revised in an iterative fashion. The "charted review" refers to the subset of articles selected for detailed charting, in this case 32 longitudinal studies selected.

Data items
We organized the full review data (n = 96) by the year the article was written, sample size, country where the study was conducted, study design, definition of depression, definition of wealth, and key findings. We focused on a subset of the full review to include in the featured charted review, selecting the longitudinal studies (n = 32). In the longitudinal charted review, we mapped the year of publication, research question, study population, study design, definition of depression, depression of wealth, direction of association between wealth and depression, and conclusions. While cross-sectional studies provide information about the association between wealth and depression, longitudinal studies with at least two time points can provide potential causal inference and interpretations. Therefore, we chose to focus on longitudinal studies for the charted review to better understand temporal and potential causal relations between wealth and depression.
We present findings on the full review, which included both crosssectional and longitudinal studies, and the charted review, which provided summarized information about the longitudinal studies.

Full review characteristics
We identified 96 published articles on wealth and depression through July 2020 (Appendix SB). Table 1 shows the characteristics of the articles in the full review. The number of such articles published increased over time. The first article included in the review was published in 1990 (n = 1). In 2019, the last full year included in the review, 11 articles on wealth and depression were published. Figure 2 shows the TA B L E 1 Characteristics of articles on wealth and depression included in full review (n = 96)
Measures of depression used in fewer than five studies are listed in The majority of studies in the full review (n = 56, 58%) found an inverse relation between greater wealth and depression. Thirty-one articles (32%) reported complicated findings, showing multiple associations in potentially different directions among subgroups. Six articles (7%) reported nonsignificant associations between wealth and depression. Three articles (3%) reported a direct relation between wealth and depression; thus, these three articles showed that more wealth was associated with more depression.

3.2
Summary of charted review findings

3.2.1
Charted review characteristics between wealth and depression, where depression increased as wealth increased. All but one article (n = 31, 97%) examined the relation between baseline wealth and subsequent depression (finding in more than half of articles that greater baseline wealth was associated with less subsequent depression); one article (n = 1, 3%) found that greater baseline depression was associated with less subsequent wealth. Thus, all but one article assessed how wealth predicted depression; one article estimated how depression predicted wealth.

3.2.3
Charted review themes Table 3 shows a charted review of findings from the longitudinal studies on wealth and depression. Studies assessing depression and wealth often used changes in wealth as their analytic focal point, either following a decrease due to an economic recession or an increase due to experimental savings initiatives. Studies could generally be grouped into five themes: observational studies following macro-economic events (n = 7); observational studies measuring trends in depression across the lifecourse (n = 7); observational studies following job loss (n = 4); observational studies measuring general patterns between wealth and depression (n = 7); and experimental studies implementing a specific savings intervention (n = 7). We report findings by these themes.

3.2.4
Observational studies following macro-economic events Seven studies used observational designs following macro-economic events such as shifts in the stock market, housing prices, or interest rates. The majority of articles using macro-economic events as focal points of their design used older adult study populations (n = 5) (ages 50 years and old); only two articles (n = 2) using macro-economic as focal points of their design had study populations ages 18 years and older. Among studies following macro-economic events, three reported an inverse relation between wealth and depression (Hamoudi & Dowd, 2014;Pool et al., 2017;Yilmazer et al., 2015). Pool et al. (2017) reported that adverse wealth shock was associated with an increase in depressive symptoms up to 4 years after the wealth shock. Yilmazer et al. (2015) reported that losses in housing and nonhousing wealth were associated with an increase in psychological distress. Meanwhile, Hamoudi and Dowd (2014) found that increased housing values were associated with improved psychological health, with greatest effects for home owners.
Four studies focusing on macro-economic events reported "complicated" findings (Boyce et al., 2018;Cagney et al., 2014;McInerney et al., 2013;Wilkinson, 2016). In general, these studies found that certain subgroups reported an inverse relation between wealth and depression and that other subgroups reported nonsignificant relations. McInerney et al. (2013) found that the 2008 stock market crash had no effect on depressive symptoms of nonstock holders, had a significant negative effect on depressive symptoms among above-median (wealthier) stock-holders (leading to more depressive symptoms), and had a positive effect on depressive symptoms of below-median (lower wealth) stock-holders (leading to less depressive symptoms). Cagney et al. (2014) found that living in a neighborhood with high foreclosure rates was associated with more depressive symptoms among older adults following the 2008 stock market crash and subprime mortgage crisis; however, they found no significant relation between having more assets and depression for people reporting default, auction, or transition to real-estate. Wilkinson (2016) found that having greater objective indicators of wealth did not appear to predict depressive symptoms following the 2008 Great Recession, but subjective indicators such as financial strain did. Boyce et al. (2018) found that low interest rates were inversely related to depressive symptoms among persons with high-debt but the relation between interest rates and depressive symptoms did not vary significantly by savings status.

3.2.5
Observational studies across the lifecourse Seven studies explored how wealth influenced depression across the lifecourse. Among these studies, four found an inverse relation between wealth and depression (Bogan & Fertig, 2018;Celeste & Fritzell, 2018;Mossakowski, 2008;Smith et al., 2005). Mossakowski (2008) found that childhood wealth was inversely related with depressive symptoms in early adulthood; in particular, having zero family net wealth or negative family wealth in childhood led to significantly greater depressive symptoms in adulthood. In a study that followed Swedish persons for 43 years, being poor at least once in life led to greater psychological distress across the lifecourse; they also found that while relative differences in psychological distress between poor and never-poor persons decreased over time, they persisted through old age (Celeste & Fritzell, 2018). Smith et al. (2005) studied the effects of disability onset on depression over time. They found that having wealth protected persons from worse depressive symptoms following onset of disability; these effects appeared to fade after 2 years, with persons having lower wealth recovering mental health following adaption to the disability. Bogan and Fertig (2018) focused specifically on onset of depression and found that persons who developed depression were more likely to have less retirement savings over time than their peers who did not develop depression.
Complicated. An increase in wealth between 1995 and 1989 was associated with a reduction in depression between 1995 and 1999 for women but was not associated with a significant change in depression for men.
Silveira et al. Participants were asked, "If you and your spouse/partner were to sell all your major possessions (including your home), turn all of your investments and other assets into cash, and pay all of your debts, would you have money left over, break even, or be in debt?" Three dichotomous indicator variables were created: "Break even" for zero net worth, "in debt" for negative net worth, and "have money left over" for positive net worth.
Yes. Having zero or negative net worth relative to positive net worth was associated with greater depressive symptoms for young adults in the study (adjusting for family background, race, age, and marriage). Wealth also partially mediated the effect of race and ethnicity on depression. Duration of childhood poverty significantly predicted depressive symptoms for adults ages 27-35 years.
(Continues) share of housing equity at baseline (housing debt; indicator measured equity stake as less than two thirds of purchase value and less than full purchase value of the house); nonhousing debt at baseline (with indicators as none, more than one fifth, or more than three quarters of total nonhousing wealth).

References
Yes. Among home owners, steeper growth in housing value was associated with improved psychological health. Rising prices were less beneficial for psychological health for home renters than for home owners.
(Continues) Yes. Mean scores of depressive symptoms were significantly lower among mothers in the treatment arm than the control arm. The effect of the treatment on mothers was significantly larger on the low-income and low-education subsamples than for the whole sample. Two savings variables were used to test for mediation; the two savings variables were negatively associated with CES-D but neither association was statistically significant (potentially due to the small number of women who contributed to savings accounts). The two savings indicators did not mediate the relation between the OK SEED intervention and depressive symptoms.

References
(Continues) in all waves of the study were coded as "never poor." Persons who responded "no" at least once were coded as "at least once poor."

References
Yes. Persons who had been poor at least once in their life had higher psychological distress across the lifecourse than persons who had never been poor. Absolute differences in psychological distress were observed across the lifecourse. While relative inequities got smaller over the lifecourse, they still persisted into old age. Slater et al. (2018)  assets, physical assets, and housing wealth) and excluded pension wealth. Household wealth calculated by summing assets, including assets held in bank accounts in England, the value of owner-occupied housing (minus outstanding mortgage), the value of business properties or holiday homes (minus outstanding mortgages), and the value of physical assets such as jewelry, artwork, and antiques, and by subtracting debts. Yes. Persons in wealth quintiles 3,4, and 5 had significantly lower odds of developing elevated depressive symptoms at 2-year follow-up than persons in quintile 1 (p < .01). Persons in quintile 2 had lower odds of depressive symptoms at follow-up than persons in quintile 1 but the results were not statistically significant (p < .07). (Continues)

Study design
Definition of depression Did the study define how wealth was measured?

Was wealth associated with less depression?
Boyce et al. (2) psychological case-ness, as a binary indicator score above 3 on a 0-12 scale. Yes. Household savings defined as a binary indicator in response to this question, "Do you save any amount of your income for example by putting something away now and then in a bank, building society, or Post Office account other than to meet regular bills?" Household debt position defined by response to two questions: (1) "Do you or anyone in your household have to make repayments on hire purchases or loans?" where participants were asked to include home mortgage loans but to exclude Department of Social Security social fund loans.
(2) Participants were asked "to what extent is the repayment of such debts and the interest a financial burden on your household?" where responses included "heavy burden," "somewhat of a burden," and "not a problem." Interest rates defined by the Bank of England base-rate on the day each person was interviewed. As an alternative interest rate indicator, authors also looked at the average interest rate in the year up until the person's interview.
Complicated. Being a saver was associated with a lower prevalence of psychological distress or psychological case-ness but the relation was not statistically significant. Interest rates on average were not linked to mental health; however, the influence of interest rates on mental health varied by household debt position.
While there appeared to be a negative effect on mental health for persons with heavy debt burden when interest rates were high, the results were not statistically significant.
(Continues) Symptoms of psychological distress. Self-report scales (0-4 scale) of depression measured at provider health checkup. Yes. Bundled treatment for intervention included free health checkups and a free personal savings account at neighborhood local banks. Participants in the treatment group received an initial deposit of 10,000 pesos ($5 US dollars) and matching 1:3 funds up to a limit.

References
Yes. Women in the savings intervention arm had statistically significant lower rates of depression than women in the control arm.
The treatment effects were largest for women whose baseline surveys did not reflect intimate partner violence. From the 1364 women allotted to treatment, 49% created an account, 33% made one deposit or more, and 21% made one withdrawal or more. The median total deposited across the project of $95 US dollars. (Continues)

Study design
Definition of depression Did the study define how wealth was measured?

Was wealth associated with less depression?
Karimli et al.

3.2.6
Observational studies following job loss Four studies observed the association of wealth and depression following job loss. One study found an inverse relation between wealth and depression: Gallo et al. (2006) reported that among persons with below median nonhousing wealth, depressive symptoms 4 and 6 years after job loss were significantly greater among persons than among persons who had not lost their jobs; among persons with above median nonhousing wealth, there was no significant difference in reporting of depressive symptoms between persons who experienced job loss and persons who did not. Two studies reported complicated findings. Rodriguez et al. (1999) found that wealth was associated with depressive symptoms following job loss for persons who were white but not for African-American persons. Riumallo-Herl et al. (2014) assessed depressive symptoms following job loss in the United States and Europe; they found that depressive symptoms scores were higher following job loss across both regions, but the effect was greater in the United States. They found no significant interaction between job loss and wealth in depression in Europe but they did find a significant interaction in the United States; namely, persons with little or no wealth before losing their jobs experienced significantly more depressive symptoms than their counterparts with wealth.
One study found no significant relation between wealth and depression following job loss: Berchick et al. (2012) found no significant relation between wealth and depression after unexpected job loss; they did find that having lower education and higher occupation prestige were associated with greater risk of depression following job loss.

3.2.7
Observational studies measuring general patterns between wealth and depression Seven studies assessed general trends between wealth and depression to understand how wealth was associated with depression in the normal course of living. Three studies reported an inverse relation (Carter et al., 2009;McGovern & Nazroo, 2015;Slater et al., 2018), two studies reported a complicated relation (Dew, 2007;Hounkpatin et al., 2015), one study found no significant relation (Yoshikawa et al., 2008), and one study found a direct relation (Nieuwenhuis et al., 2017). In New Zealand, Carter et al. (2009) specifically sought to answer the question of whether wealth was associated with mental health (and whether that relation held when controlling for other socioeconomic indicators). They found that having more wealth was significantly associated with having fewer psychological symptoms over time; this association was greater than that between household income and psychological symptoms. McGovern and Nazroo (2015) sought to explore the relation between wealth and mental health to understand potential mechanisms for the relation; in a study of UK older adults, they found a direct effect of wealth on depression. Further, the relation between objective indicators of wealth and depression was partially mediated by subjective indicators such as social status. Slater et al. (2018) focused their research question on metabolic health, obesity, and depression, but they did report relations with covariates and in doing so found that persons in the three highest wealth quintiles were less likely to report depressive symptoms after 2 years than persons in the lowest wealth quintile. Of the studies that found a complicated relation, Dew (2007) found that wealth and depressive symptoms were inversely related in unadjusted models but that the association was no longer significant when controlling for other factors, suggesting that the relation was mediated by other factors. In a study of older adults in the United States and the United Kingdom, Hounkpatin et al. (2015) found that wealth was associated more strongly with depressive symptoms in the United Kingdom than in the United States; they also found that wealth rank was more important than absolute wealth in predicting depressive symptoms over time. In a study assessing access to institutional resources (owning a savings account), Yoshikawa et al. (2008) found no significant relation between having access to a checking account or savings account and depressive symptoms; they did find that institutional resource access was related to reduced financial distress, which in turn was associated with depressive symptoms. In the one longitudinal study that showed a direct relation between wealth and depression, Nieuwenhuis et al. (2017) found that teens who moved to higher wealth neighborhoods had a significantly higher likelihood of reporting depression than teens who had not moved.

Experimental studies
Seven experimental studies reported on savings interventions that aimed to improve mental health. Among the seven experimental studies, five reported an inverse relation between wealth and depression, where participants reported significantly lower depression following the savings interventions (Han et al., 2013;Huang et al., 2014;Karasz et al., 2015;Ssewamala et al., 2012;Tankard et al., 2019). The five studies showing an inverse relation were conducted among adults (n = 3) and children (n = 2) and were conducted in the United States (n = 2), Uganda (n = 2), and Columbia (n = 1). The five successful interventions had multiple components; in addition to savings incentives (through matching contributions in savings accounts) they included educational or social components through mentorship programs (Han et al., 2013;Ssewamala et al., 2012) or peer coaching (Karasz et al., 2015) to encourage saving. Mothers in the treatment arm of the SEED Oklahoma intervention, which provided initial funds for their young children's college accounts and matching contributions, reported lower depressive symptoms than women in the control arm (Huang et al., 2014). Of the two articles that reported more complicated results, one showed a treatment effect (significantly lower depression in the savings intervention group) at 12-and 24-month of follow-up that attenuated at 36-and 48-month follow-up (Karimli et al., 2019). The second article that reported complicated results found that the savings intervention resulted in significantly lower depression at follow-up for women with below median household income but not for women with above median household income.
In the experimental studies, context mattered for the relation between wealth and depression. Among the seven experimental interventions, household context affected the treatment effect in four of the studies (Han et al., 2013;Huang et al., 2014;Kilburn et al., 2019;Tankard et al., 2019). At times, this could improve or weaken the intervention effect. In the case of a conditional cash transfer in Ghana, the women who benefited most from the cash savings intervention were those who had below median income households (Kilburn et al., 2019). In the SEED Oklahoma study, treatment effects were significantly larger for women who were low-income and had low education (Huang et al., 2014). In Colombia, the effects of participating in a savings intervention were greatest among women who had not experienced intimate partner violence at baseline (Tankard et al., 2019).

Summaries of the characteristics of the full review and charted review
There is a growing literature on wealth and depression. We identified 96 articles that have been published from the start of online search history through July 19, 2020 that featured the relation between wealth and depression. The number of publications on wealth and mental health has increased over time, and the number of longitudinal studies also increased over time.

Comparison of characteristics of longitudinal versus cross-sectional studies
The characteristics of longitudinal studies and cross-sectional studies were similar. Both featured majority adult populations, with greater than one third of studies featuring only older adults. One difference between longitudinal articles and cross-sectional studies was that the proportion of studies conducted outside of the United States was greater among cross-sectional studies than among longitudinal studies. Although cross-sectional studies do not allow for causal inference, they nevertheless provide important information about the link between wealth and depression. And, in particular, they can help us to understand how the relation between wealth and depression may vary across countries until more longitudinal studies are established globally. Both longitudinal and cross-sectional studies signaled consistent patterns, namely, that greater wealth had a significant association with less depression across a range of contexts, populations, and scenarios.

Summaries of the findings of the full review and charted review
The majority of studies in both the full review and the charted review showed an inverse relation between wealth and depression. In general, having more assets was associated with less depression across the studies reviewed. Across the world, age groups, genders and races, having more assets was related to less depression throughout the lifecourse. More than one half of studies reported an inverse relation between wealth and depression and one third of studies reported complicated relations between wealth and depression. While the associations between wealth and depression were not always statistically significant, the relation between the two constructs, with rare exception, showed a dose response pattern, where more wealth being associated with less depression.

Charted review
The

4.3.1
Observational studies following macro-economic events Economic events influence mental health; shifts in the stock market, interest rates, and value of housing influenced mental health for years to follow. These findings were consistent with other reviews that have looked at the effects of economic crises on mental health (Haw et al., 2015;Uutela, 2010;Vásquez-Vera et al., 2017

Observational studies following job loss
Existing studies suggest that having wealth protected against depressive symptoms up to 4-6 years following job loss; differences in depressive symptoms following job loss between Europe and the United States may indicate that broader social safety net policies in Europe may have softened the blow of job loss. Given the much less robust economic and social safety net following job loss in the United States, people must rely more on personal savings, which may, therefore, explain the significant interactive effects of wealth and job loss on depression in the United States. These findings are consistent with a crosssectional study that suggests that the reduced influence of personal wealth on depression in Nordic countries may be due to generous social and economic safety nets in those countries (Kourouklis et al., 2019).
Future policies, therefore, may wish to focus on savings interventions to reduce the mental health effects of job loss.

Observational studies measuring general patterns between wealth and depression
These studies suggested that subjective or psychosocial factors may be driving part of the relation between wealth and depressive symptoms.
The studies together suggest that rank (Hounkpatin et al., 2015), social status (McGovern & Nazroo, 2015), and relative wealth compared to others in one's neighborhood (Nieuwenhuis et al., 2017) all influence the relation between wealth and depression.

Experimental studies
Savings interventions made a significant difference, particularly among specific subpopulations. The experimental studies showed that interventions to increase wealth can work, and that their effectiveness is in part determined by context. Treatment effects varied by level of social or economic well-being prior to intervention. As with all other determinants of health, wealth is one of many structural and significant factors that shapes health outcomes. While these studies showed that wealth was a predictor of depression, and one that could be intervened upon, they also highlighted the complex set of causal factors that shape the contexts people live in and that shape health. Wealth interventions should balance the full suite of factors that contribute to health, and the trade-offs inherent in investing in particular interventions over others.

Other reviews on wealth and health
This review is consistent with other reviews assessing the relation between wealth and mental health. The most relevant example is a systematic review of studies on wealth and health from 1990 through 2006 conducted by Pollack and colleagues. Of the six studies they included that examined the relation between wealth and mental health, five reported significant links between wealth and mental health within at least a subgroup of the population. One study found no significant association between wealth and depression once controlling for other factors. They concluded that wealth was a significant factor in health, and recommended that wealth and income be measured separately (Pollack et al., 2007).
Our findings that assets across the life course were associated with reduced depression were consistent with reviews on other socioeconomic indicators and mental health outcomes. Muntaner (2004) reviewed the literature on mental disorders and socioeconomic position. The author found in general that having a higher socioeconomic position was associated with better mental health; in particular, they found that persons in lower social strata consistently had higher depression. Similar to our findings, they identified exposure across the

Limitations
This scoping review has several limitations. First, we limited our reviews to articles that included an objective wealth indicator (e.g., financial assets or savings). We did not include articles that only had subjective indicators of wealth (e.g., financial strain) although subjective indicators may well explain some of the relation between assets and mental health. Second, we also did not include studies that featured debt without other objective indicators of wealth, given that debt as a concept has a complex relation with depression (Richardson et al., 2013), meriting a separate investigation. However, understanding objective indicators is an important part of the complex set of factors that contribute to depression, so these findings can help pave the way for future work on other aspects of financial assets. Third, our results reflect the literature as of July 2020 which may change and evolve over time. Finally, the studies included were limited to English language studies, which may exclude findings conducted globally. Second, studies on wealth should define how the concept is operationalized, and strive towards consistency with established work in order to build a more robust body of evidence. Studies should include contextual factors, potentially measuring objective and subjective indicators of wealth, to better understand the mechanism and causal drivers of mental health. This review showed that more than one third of articles concluded that the relation between wealth and depression is complicated, suggesting that having more information on the definition and types of assets may provide clarity on the role of assets in shaping mental health.

Recommendations for future research
Third, more intervention studies should be conducted to assess how savings initiatives can improve mental health in different contexts across different populations. Interventions can reduce depression, and studies show several examples that have been met with success. There is much work to be done in the future to explore the relative benefits of particular interventions, weighing considerations such as return on investment and equity versus efficiency trade-offs.

CONCLUSIONS
The preponderance of evidence reviewed here makes the case that persons ending the pandemic with record job loss and financial insecurity. Rising inequalities globally make the call for this work even more urgent, for the research and policy community to better understand the influence of assets towards the end of mitigating growing mental health gaps.

DATA AVAILABILITY STATEMENT
Data sharing is not applicable to this article as no new data were created or analyzed in this study.

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1002/brb3.2486