Overindebted, unsupported and lonely, and mentally stressed and depressed—exploring the relationship between overindebtedness, social disconnectedness and poor mental health in a population-based Swiss study Authors: Oliver Hämmig Published: 2025 ABSTRACT AbstractBackgroundOnly little is known about the social and health conditions of overindebted individuals in Switzerland. Very few studies on the mental health outcomes or correlates of overindebtedness have been performed so far, and to date, hardly any study has been conducted on the social well-being and connectedness of overindebted people.MethodsSurvey data collected in 2019 from clients of all four official debt advisory centers in the Canton of Zurich were used and merged with a subsample of the Swiss Health Survey 2017 of the same age and canton of residence. The combined dataset and pooled study sample covered a total of 2,216 individuals, 219 overindebted adults and 1,997 respondents of the Swiss Health Survey and residents of the Canton of Zurich. Contingency tables were constructed, and multiple logistic and linear regression analyses were performed to estimate different measures of effect or association (bivariate frequency distributions, multiple-adjusted odds ratios, standardized beta coefficients).ResultsOverindebted individuals show a markedly increased frequency and risk of being socially disconnected or isolated. They are much more likely to be insufficiently supported (55%) and to frequently feel lonely (43%) than not overindebted representatives of the general population (10%/4%). Overindebted individuals accordingly have – even after adjustment for control variables – a strongly increased risk of being comparably unsupported (aOR = 12.6) and frequently feeling lonely (aOR = 15.4) in comparison with their counterparts. However, this association does not fully explain the strong negative mental health impact of overindebtedness. Half or less of this effect was indirect and thus was mediated by perceived social support and felt loneliness.ConclusionsThe very poor mental health status of overindebted people cannot be completely attributed to the strong association between overindebtedness and social disconnectedness that was found in this study and shown for the first time.Keywords: Overindebtedness, Social connectedness, Loneliness, Social support, Switzerland Introduction Household indebtedness has grown considerably in most developed and high-income countries for many years [1]. In the wake of the global financial crisis of 2008, an epidemic of overindebtedness was observed [2–5], a “hidden” epidemic that “overshadows the lives of millions in rich countries” [3]. Recently, another crisis, a global health crisis, has occurred and revealed a loneliness epidemic that has noticeably increased but was present and observed even before the COVID-19 pandemic [6–9]. Either way, both overindebtedness and loneliness are on the rise – in North America and Europe [1–5, 9] and, hence, in Switzerland. The two phenomena or epidemics seem to be partly related to one another. Overindebtedness can be defined as the precarious financial situation of being unable – permanently or temporarily – to meet financial obligations and commitments and to repay and/or service all debts and outstanding accounts fully and on time [2, 10, 11]. Overindebtedness is characterized as having multiple loans, being in payment arrears or having outstanding payments, spending an overly large proportion of personal or household income on repayments and/or debt services, feeling highly burdened by debts and debt service, and seeking professional help or advice from a debt counseling service. It is estimated that 6.5% or more of the Swiss resident population, a substantial (and rising) proportion, meet one or even several of these criteria or preconditions and therefore are overindebted or at risk of becoming overindebted in the near future [10, 11]. Being overindebted in Switzerland is particularly hopeless, burdensome, stressful and health-threatening because there is no real prospect of complete debt relief and hence lifelong no way out of the situation [10, 11]. Overindebtedness cannot be equated with poverty or material deprivation. The phenomenon is observed even among the better earning individudals and households. Overindebted individuals in Switzerland are underproportionally but none the less substantially represented among the better educated and higher earners [10, 11]. Around one quarter or more of the overindebted has a fairly high personal net income per month of more than CHF 6’000 and a higher vocational education or even a university degree [10, 11]. Overindebtedness, although still a rather poorly recognized and explored phenomenon in wealthy countries such as Switzerland, has been studied and proven to be a major risk factor for general, musculoskeletal and particularly mental health [4, 5, 10, 12–20] and for unhealthy or even self-destructive behaviors [17, 19, 21–25]. Overindebtedness without doubt is a serious and highly stressful experience in life, or a so-called critical life event [26], that places the health of affected people at great risk apparently [10, 11, 26]. However, it is not well understood if overindebtedness is associated with social disconnectedness and lack of social support, and if and how that mediates the impact of overindebtedness on (mental) health. This still unanswered question is of particular relevance for public health because social support is an important and proven health resource [27], and perceived social isolation or disconnectedness and loneliness have been found to be significant health risk factors [28]. The present study therefore focuses on social disconnectedness and perceived social isolation as a possible consequence or correlate of overindebtedness and the assumed “missing link” in the observed strong relationship between overindebtedness and poor mental health. This might constitute the true causal path or mechanism through which overindebtedness affects the mental health of those concerned. However, overindebtedness has not been considered or has scarcely been explored as a potential cause, consequence or correlate of social isolation and disconnectedness. It seems plausible that overindebtedness accompanies social decline and a loss of status and therefore leads to a weakening of social ties, relationships and support and, finally, to perceived social isolation or, in other words, loneliness. Of course reverse causation may also be true. Isolated and/or lonely individuals may end up being overindebted because they lack social ties and social (and financial) support. In spite of such possibility, the main reason why overindebtedness is strongly and negatively associated with mental health [11] is presumably that overindebtedness lowers or weakens social connectedness, and social disconnectedness, in turn, is an important determinant of mental health [29]. Loneliness and lack of social support are indicators of social isolation and disconnectedness [13, 30] and predictors of or risk factors for poor mental health [30–33]. As such, they are presumably mediating (or confounding) the relationship between overindebtedness and mental health. In other words, social (dis)connectedness may be a mediator (or confounder) of the strong negative association between overindebtedness and mental health, in addition to the sense of control as has been found already previously [26]. The research gap and lack of evidence regarding the effect of overindebtedness on social connectedness were the reason and starting point for this cross-sectional study designed to answer the question of whether this largely unexplored relationship between overindebtedness and social connectedness – if present – explains the previously identified strong negative relationship between overindebtedness and mental health [11, 26]. In light of these arguments and considerations, the present study was based on the premise that perceived social isolation or loneliness and lack of social support are indicators of social disconnectedness and on the following two hypotheses (see Fig. 1) and the subsequent research questions: Path model illustrating the postulated hypotheses and the underlying causal paths and steps of analysis Hypotheses: Overindebtedness is an important cause of social isolation and disconnectedness. (H1) Overindebtedness is an important cause of social isolation and disconnectedness. (H1) The relationship between overindebtedness and poor mental health is mediated by social disconnectedness. (H2) The relationship between overindebtedness and poor mental health is mediated by social disconnectedness. (H2) Research questions (operationalized hypotheses): Is being an overindebted client of a debt advisory center a strong predictor or risk factor for frequently feeling lonely and comparably less socially supported? Is being an overindebted client of a debt advisory center a strong predictor or risk factor for frequently feeling lonely and comparably less socially supported? Can elevated symptoms of depression and psychological distress among overindebted individuals be attributed to or largely explained by their loneliness and relative lack of social support? Can elevated symptoms of depression and psychological distress among overindebted individuals be attributed to or largely explained by their loneliness and relative lack of social support? Figure 1 shows all the relevant study variables or constructs and their relationships and illustrates the 3-step approach to test for mediation according to Baron and Kenny [34]. Methods Data and study sample Two different sets of survey data based on partly overlapping questionnaires were linked together to obtain a pooled dataset with comparable measures (identical wording of questions and response scales) and surveyed individuals (same age and residential area). The pooled data consisted of the following two samples: exposed (overindebted individuals) and nonexposed (largely not overindebted representatives of the general adult population): a full sample of 219 overindebted clients (aged 18 and older) from the four official debt counseling centers in the Canton of Zurich who were seeking advice during an 8-month period in 2019, had sufficient mastery of the German language, and were willing to participate subsequently in a survey on an anonymous and voluntary basis (as the exposed group) a full sample of 219 overindebted clients (aged 18 and older) from the four official debt counseling centers in the Canton of Zurich who were seeking advice during an 8-month period in 2019, had sufficient mastery of the German language, and were willing to participate subsequently in a survey on an anonymous and voluntary basis (as the exposed group) a subsample of 1,997 randomly selected respondents of the Swiss Health Survey of 2017, i.e., a representative and comparable sample of adults permanently living in the Canton of Zurich (as the nonexposed or comparison group) a subsample of 1,997 randomly selected respondents of the Swiss Health Survey of 2017, i.e., a representative and comparable sample of adults permanently living in the Canton of Zurich (as the nonexposed or comparison group) Overindebted individuals in the study population and compared to the representatives of the general adult population are relatively often middle-aged, frequently single, separated or divorced, less educated, lower earning and much more frequently unemployed [10, 11]. Nevertheless, there are substantial proportions of better paid and well educated people among overindebted individuals. Measures Exposure variable Overindebtedness was measured by the newly constructed dichotomous variable of being or rather having been an overindebted client of a debt advisory center in 2019 or a (presumably not overindebted) participant and respondent of the Swiss Health Survey 2017. Primary outcome variables (mediator variables) Perceived social isolation or loneliness, as one indicator of social disconnectedness, was assessed identically in both surveys and merged study samples by a single and simple question asking about frequent feelings of loneliness, which were measured by the following question and answers: How often do you feel lonely? (1 ‘seldom/never’, 2 ‘sometimes’, 3 ‘fairly often’, 4 ‘very often’). Social support, another indicator of social disconnectedness, was assessed via a multiple-item measure of the self-reported level of social support, namely, the 3-item Oslo Social Support Scale [33, 35, 36]: How many people are so close to you that you can count on them if you have great personal problems? (4 response categories: none, 1–2, 3–5, 5+) How many people are so close to you that you can count on them if you have great personal problems? (4 response categories: none, 1–2, 3–5, 5+) How much interest and concern do people show in what you do? (5 response categories: none, little, uncertain, some, a lot) How much interest and concern do people show in what you do? (5 response categories: none, little, uncertain, some, a lot) How easy is it to obtain practical help from neighbors if you need it? (5 response categories: very difficult, difficult, possible, easy, very easy) How easy is it to obtain practical help from neighbors if you need it? (5 response categories: very difficult, difficult, possible, easy, very easy) Response categories from each of the three items were numbered consecutively from 1 to 4 or 5. The 3-item Oslo Social Support Scale (OSSS-3) was constructed by summing up the scores of the three items to a sum score ranging from 3 to 14. A maximum score of 14 is indicating the highest degree of social support while a minimum score of 3 stands for the lowest degree of support or social connectedness. Secondary outcome variables The indicators of (poor) mental health used to test the mediational hypothesis (H2) in this study and its population were psychological distress and depression. These two indicators are commonly used as mental health measures or outcomes and were developed as instruments for mental health [34] long before a recently published study used them to examine mental health correlates of overindebtedness [14]. Psychological distress was identically measured in both surveys and subsamples by the 5-item Mental Health Inventory (MHI-5), a widely used brief screening instrument and subscale of the established and validated MOS 36-item Short Form Health Survey (SF-36) [36]. The SF-36 was originally developed for use in clinical practice and research and was constructed to survey and screen health status in the Medical Outcomes Study [37]. The MHI-5 in the present study contains five questions about the frequency of being or feeling very nervous, down in the dumps, calm and balanced, downhearted and depressed or happy over the previous 4 weeks. These questions had to be answered on a 5-point scale (1 ‘always, 2 ‘mostly’, 4 ‘sometimes’, 5 ‘seldom, 6 ‘never’). In the original English version of the MHI-5 a 6-point answer scale was used. The originally predefined answer ‘a good bit’ (of the time) with a score of 3 and located between ‘mostly’ (score of 2) and ‘sometimes’ (score of 4) is a quite unusual term which can barely be translated into German and therefore was left out as a response option in the Swiss Health Survey. The total score of the MHI-5 in the Swiss Health Survey just like in the present study and pooled data set was computed by reversing the polarity of two items (calm and balanced, happy), adding the scores of all items and transforming the sum score from 5 to 30 to a 0-100 point scale by subtracting 5 from the sum score, then dividing it by 25 and finally multiplying it by a factor of 100. A total score of 52 or below in the transformed scale is commonly used as a cutoff for a clinically relevant mental disorder. Scores between 53 and 72 indicate a moderate degree of psychological distress. Depression was measured by a 9-item subscale of the 59-item Patient Health Questionnaire (PHQ-9). The PHQ-9 is the depression module of a self-administered version of the PRIME-MD (Primary Care Evaluation of Mental Disorders) diagnostic instrument for common mental disorders or, in short, a brief, reliable and valid measure of depressive disorders and depression severity [38, 39]. The PHQ-9 contains questions about the frequency of having little interest or pleasure in doing things, feeling down, depressed, or hopeless or feeling tired or having little energy. The nine questions had to be answered on a 4-point scale from 0 ‘not at all’ to 3 ‘almost every day’ over the last two weeks. These scores were added to a sum score ranging from 0 to 27. A severe or major depression is indicated by a total score of 15 or higher and can be diagnosed if a person has five symptoms or more for at least “more than half of the days”. In screenings already a score of 10 or more is often used as a general cutoff for a potentially treatable mental health condition. Control variables Sex, age and education were used as control variables. These variables were assessed by assigning oneself to a gender (0 ‘female’, 1 ‘male’), one of eight predefined age categories (1 ‘18–20 yrs.’, 2 ‘21–30 yrs.’, 3 ‘31–40 yrs.’, 4 ‘41–50 yrs.’, 5 ‘51–60 yrs.’, 6 ‘61–70 yrs.’, 7 ‘71–80 yrs.’, 8 ‘80 + yrs.’) and the highest educational level achieved thus far (from 0 ‘no compulsory education’ to 7 ‘university degree’). Originally assessed on a 8-point ordinal scale, education was grouped into four degrees or levels (due to partly small and very unequal numbers of cases of the educational qualifications): low (0–1), medium (2), high (3–5), very high (6–7). Statistical analyses To characterize the study population regarding the most important study variables (overindebtedness, social support, loneliness, psychological distress, and depression), marginal distributions for both the overindebted individuals and controls from the general population were calculated separately. Furthermore, multiple logistic regression analyses were performed to estimate (un)adjusted odds ratios and to test for causation or rather association (Hypothesis 1) between overindebtedness and social support and loneliness. Finally, stepwise multiple linear regression analyses were performed to calculate standardized beta coefficients and to test for mediation (Hypothesis 2) with cross-sectional data according to the 3-step approach of Baron & Kenny [34]. Results The main and most frequently reported reasons for the precarious financial situation of the 219 overindebted clients of the official debt advisory centers in the Canton of Zurich were unemployment (38.1%); lack of spending discipline (35.3%); disease or accident (29.4%); addiction problems, including alcohol, drug, shopping and gambling (28.4%); divorce or separation (26.6%); and other people (21.6%). Sometimes health problems seem to be more a cause rather than a consequence of overindebtedness. This suggests that the assumed causal direction from overindebtedness to poor health can also be reversed. A bi-directional relationship is even more likely. However, the calculated two-way frequency or contingency tables shown in Table 1 revealed strongly different relative frequencies stratified by overindebtedness status, i.e., very unequal prevalence rates between the two subsamples of the study population (clients of debt advisory centers, representatives of the general population). 55% of the overindebted individuals had no or minimal support, whereas only 10% of the comparison group showed an equally low level of social support. Almost 43% of the overindebted participants reported frequent feelings of loneliness, whereas fewer than 4% of the representatives of the general population felt equally. In addition, overindebted clients of debt advisory centers were much more often highly distressed (57% vs. 4%) and severely depressed (24% vs. 2%) compared to the general population. In other words, the two subpopulations differ greatly with respect to measures of social disconnectedness to the disadvantage of the overindebted, as expected. Table 1Health-related characteristics of the study population, stratified by relevant subsamplesClients of debt advisory centers(N = 219)aRepresentativesof the general population(N = 1,997)bTotal study population(N = 2,216)Overindebtedness (dummy: 0/1)Yes100.0%–9.9%No–100.0%90.1%Social Support (OSSS-3: 3–14)No/little support (3–8)54.8%10.0%14.5%Moderate support (9–11)37.9%43.9%43.3%Much support (12–14)7.3%46.2%42.2%Feelings of loneliness (1–4)Seldom/never (1)23.9%63.7%59.8%Sometimes (2)33.5%32.8%32.9%Fairly often (3)25.2%2.0%4.3%Very often(4)17.4%1.5%3.1%Psychological distress (MHI-5: 0-100)No/low (73–100)18.1%86.5%79.9%Moderate (53–72)24.8%9.7%11.1%High (0–52)57.1%3.8%8.9%Depression (PHQ-9: 0–27)None/mild (0–9)46.1%92.9%88.2%Moderate (10–14)30.0%4.8%7.3%Moderately severe (15–19)18.9%1.6%3.4%Severe (20–27)5.1%0.7%1.1%aOverindebted adults seeking advice from one of the official debt advisory centers in the Canton of Zurich and participating in the overindebtedness study (and survey) of 2019bSubsample of the nationally representative Swiss Health Survey of 2017, restricted to adult respondents (aged 18 yrs. and older) from the Canton of Zurich; weighted data Health-related characteristics of the study population, stratified by relevant subsamples aOverindebted adults seeking advice from one of the official debt advisory centers in the Canton of Zurich and participating in the overindebtedness study (and survey) of 2019 bSubsample of the nationally representative Swiss Health Survey of 2017, restricted to adult respondents (aged 18 yrs. and older) from the Canton of Zurich; weighted data Odds ratios and standardized beta coefficients as calculated multiple adjusted measures of association clearly showed and confirmed an extraordinarily strong relationship between exposure (overindebtedness) and the two primary outcomes under study (social support, loneliness). Even after adjusting for different control variables, the risk of having no or only little social support and often feeling lonely increased more than twelvefold and fifteenfold, respectively, for overindebted individuals compared with the predominantly not overindebted representatives of the general population (see Table 2). Table 2Logistic regression analyses with (un)adjusted odds ratios as measures of association to test the causational hypothesis (H1)No/little social support (3-8)Very/fairly often feelings of loneliness (3-4)%OR95% CI%OR95% CIOverall prevalence in study population14.57.4 MODEL 1 (unadjusted) Overindebtedness No (general population)10.013.51 Yes (clients of debt advisory centers)54.810.968.08–14.8742.720.4214.26–29.24No. cases in model2,1472,209 MODEL 2 (adjusted) Overindebtedness No (general population)10.013.51 Yes (clients of debt advisory centers)54.812.568.95–17.642.715.4410.6–22.6Adjusted for Sex (male)1.371.05–1.790.730.50–1.06 Age (1–8)1.231.14–1.340.880.78–0.98 Education (1–4)0.740.65–0.850.640.53–0.79No. cases in model2,1412,200 Logistic regression analyses with (un)adjusted odds ratios as measures of association to test the causational hypothesis (H1) Stepwise multiple linear regression analyses further revealed that the strong negative relationship between overindebtedness (exposure variable) and psychological distress or depression as the two mental health indicators (secondary outcome variables) can be partly but not fully explained by perceived social support and feelings of loneliness as the two assumed mediators and indicators of social disconnectedness (primary outcome variables). Standardized beta coefficients as measures of the association between overindebtedness and mental health were reduced from − 0.56 to − 0.38 for psychological distress and from 0.40 to 0.21 for depression when the potential mediators were included in the regression models (see Table 3). However, the beta coefficients decreased substantially but did not disappear completely. This means that the relative lack of social support and relatively frequent feelings of loneliness found among overindebted individuals only partly explain their very poor mental health. Table 3Stepwise multiple linear regression analyses with standardized beta coefficients as measures of association to test the mediational hypothesis (H2)Outcome variables:Social support (3–14)Loneliness (1–4)Psychological distress (0-100)1)Depression (0–27)Step 2Step 2Step 1Step 3Step 4Step 1Step 3Step 4ßßßßßßßßExposure variable: Overindebtedness (0/1)− 0.27***0.36***− 0.56***− 0.38***0.40***0.21***Mediator variables: Social support (3–14)0.19***0.12***− 0.15***− 0.11*** Feelings of loneliness (1–4)− 0.55***− 0.43***0.49***0.42***Control variables: Sex (male)− 0.06**− 0.10***0.11***0.05**0.07***− 0.11***− 0.06***− 0.08*** Age (1–8)0.01 (n.s.)− 0.08***0.09***0.09***0.05**− 0.12***− 0.10***− 0.08*** Educational level (1–4)0.00 (n.s.)− 0.07***0.04*0.06***− 0.01 (n.s.)− 0.07***− 0.07***− 0.04*R square adjusted0.0780.1680.3450.4220.5330.2060.3370.372No. cases in model2,1402,1992,1742,1102,1102,1522,0872,0871) A higher score indicates a lower level of mental stress and a negative beta coefficient indicates a risk factor for mental stress (and a positive one a protective factor)*p ≤ .05; **p < .01; ***p < .001; (n.s.) = not significant (p > .05) Stepwise multiple linear regression analyses with standardized beta coefficients as measures of association to test the mediational hypothesis (H2) 1) A higher score indicates a lower level of mental stress and a negative beta coefficient indicates a risk factor for mental stress (and a positive one a protective factor) *p ≤ .05; **p < .01; ***p < .001; (n.s.) = not significant (p > .05) Discussion The present study aimed to fill a research gap in Switzerland with respect to the social connectedness or well-being of overindebted individuals. Social disconnectedness was expected to be a consequence of overindebtedness and the explanation why overindebted individuals have such poor mental health, as shown in earlier studies [11, 26]. Overindebtedness not only goes along with poor mental health, as previously shown, but also comes along with perceived social isolation (or felt loneliness) and lack of social support. In short, overindebtedness is accompanied by social disconnectedness, which is at least coresponsible for the very poor mental health of overindebted individuals already observed. In brief, this was the research interest and main finding of this study. Overindebted individuals were found to have a strongly increased risk of being socially less supported (aOR = 12.6) and feeling lonely frequently (aOR = 15.4), as was assumed in Hypothesis (1) However, this only partly explains why overindebted individuals are comparably strongly mentally distressed and depressed, as additionally postulated in Hypothesis (2) An independent and significant direct effect of overindebtedness on psychological distress and depression (as indicators of poor mental health), remained (ß=-0.38/0.21) even after adjusting for the two potential mediators and indicators of social disconnectedness (no/little perceived social support, frequent feelings of loneliness). In sum, overindebtedness was strongly negatively associated with social disconnectedness, as expected, but this newly identified or previously largely unexplored association by far did not completely explain the already known very poor mental health of overindebted individuals. There are no other studies in the international literature on overindebtedness that have explored the social connectedness of overindebted individuals, as far as is known. Therefore, the findings of this study cannot be compared with those of previous studies. However, a nationally representative survey (Social Diagnosis Survey) and longitudinal study from Poland among almost 8,800 respondents aged over 15 incidentally revealed a statistically significant and increased relative risk of feeling lonely for the approximately 200 overindebted individuals included in the study population [40]. This finding is supported by the present study, although the magnitude of the effect or association in the two studies were quite different. Overindebted individuals in the Polish study showed only a slightly increased relative risk for loneliness (aOR = 1.63), whereas those in the present Swiss study presented a far greater relative risk of loneliness (aOR = 15.44). This may be attributable, at least partially, to the fact that loneliness in the Polish study was measured by a simple dichotomous yes/no-question and, in the Swiss study, also by a single item but with a 4-point response scale on the frequency of feeling lonely, which allowed us to differentiate between the occasionally lonely and the frequently lonely. However, it seems more plausible that overindebted individuals in a rich and high-income country such as Switzerland, with comparably very low rates of unemployment (which often precedes overindebtedness) and with no legal possibility of complete debt relief, are more stigmatized, disadvantaged and socially isolated and therefore feel much more lonely and less supported by others than overindebted people in other countries where overindebtedness, unemployment or working poverty may not be not such rare phenomena or have such a heavy burden and strong social stigma. We have now learned that the fact that overindebted people in Switzerland are in very poor mental health, i.e., strongly psychologically distressed and depressed, can at least partly be attributed to the study’s main finding that overindebtedness makes affected people feel isolated and lonely and less socially supported and appreciated. The burden of overindebtedness itself and independently (direct effect) and the perceived social isolation and lack of support as a result of overindebtedness (indirect effect) in turn strongly predict psychological distress and depression among adults living in the Canton of Zurich. The primary outcome of social disconnectedness (low social support and frequent feelings of loneliness) predicts the secondary outcome of poor mental health (psychological distress and depression) even more than exposure (overindebtedness) directly and independently. In particular, feelings of loneliness were not only highly correlated with or associated with overindebtedness (ß=0.36) but also the strongest predictor of poor mental health (ß=-0.43/0.42), even prior to overindebtedness itself (ß=-0.38/0.21). In other words, the indirect effects of overindebtedness on mental health seem to be even stronger than the direct effects. This recognition has implications with respect to prevention measures (see conclusion). Strengths and limitations The present study focused on and provided initial empirical evidence on an unexplored or rather underexplored condition of a largely ignored population in public health and research in Switzerland (as in most other countries), namely, the social connectedness or rather disconnectedness of overindebted individuals, measured by perceived social support and feelings of loneliness. This and the use of pooled and population-based survey data with an oversampled and comparably large number of overindebted individuals (“exposed”), which are usually underrepresented in population surveys, and a representative sample of the general population as a comparison group (“unexposed”) are clear strengths of the present study. The limitations of the study, in contrast, are (a) the use of cross-sectional data limiting causal conclusions from associations, (b) the lack of information on the exposure status (overindebtedness) of at least one of the two pooled study samples, potentially allowing measurement error or misclassification bias to occur, and (c) the recruitment of clients from debt advisory centers instead of directly and randomly selecting overindebted individuals from the general population, making a selection bias possible. There is a fundamental principle in epidemiological research that a potential cause (exposure) must occur before the effect (outcome) is observed. To conclude causation from association in a studied relationship between exposure and outcome, it is required that the exposure precedes the outcome in time. When cross-sectional data collected at one point in time are used, this so-called criterion of temporality cannot be verified; therefore, one must beware of overinterpreting associations as cause‒effect relationships. Nevertheless, causal conclusions based on cross-sectional data cannot be excluded either. This is especially the case when other Hill’s criteria for causality are fulfilled and provide additional evidence for cause and effect, such as a strong association, a clear gradient or dose‒response relationship, high plausibility, the existence of consistent findings, supporting experimental evidence or coherent common knowledge [41, 42]. Besides having no (empirical) proof of causation, there is also no clear (theoretical) indication for the direction of causality. The causational and mediational hypotheses postulated in the present study imply a unidirectional relationship from overindebtedness to social disconnectedness and to poor mental health. But the causality could also be reversed. Instead of being a cause for social isolation and loneliness and subsequently for poor mental health, overindebtedness just as well could be a consequence of social disconnectedness and poor mental health. Moreover, reciprocity or bidirectional causality in the relationship between overindebtedness and social disconnectedness or poor mental health likewise appear plausible. Used cross-sectional data do not allow to draw any conclusions on causation or reverse causation. There is a fundamental principle in epidemiological research that a potential cause (exposure) must occur before the effect (outcome) is observed. To conclude causation from association in a studied relationship between exposure and outcome, it is required that the exposure precedes the outcome in time. When cross-sectional data collected at one point in time are used, this so-called criterion of temporality cannot be verified; therefore, one must beware of overinterpreting associations as cause‒effect relationships. Nevertheless, causal conclusions based on cross-sectional data cannot be excluded either. This is especially the case when other Hill’s criteria for causality are fulfilled and provide additional evidence for cause and effect, such as a strong association, a clear gradient or dose‒response relationship, high plausibility, the existence of consistent findings, supporting experimental evidence or coherent common knowledge [41, 42]. Besides having no (empirical) proof of causation, there is also no clear (theoretical) indication for the direction of causality. The causational and mediational hypotheses postulated in the present study imply a unidirectional relationship from overindebtedness to social disconnectedness and to poor mental health. But the causality could also be reversed. Instead of being a cause for social isolation and loneliness and subsequently for poor mental health, overindebtedness just as well could be a consequence of social disconnectedness and poor mental health. Moreover, reciprocity or bidirectional causality in the relationship between overindebtedness and social disconnectedness or poor mental health likewise appear plausible. Used cross-sectional data do not allow to draw any conclusions on causation or reverse causation. It is further important or desirable in epidemiological research that the reliability of an exposure assessment is preferably high, which means that the observed exposure status conforms largely to the true exposure status, i.e., that all or most of the study participants are assigned correctly to the exposed and the unexposed. If this is not the case, a misclassification bias is likely to occur.An exposure assessment is usually qualified and evaluated by the two quality criteria of sensitivity and specificity, which are high and unproblematic when those classified as exposed are truly and predominantly exposed (sensitivity) and those classified as unexposed are truly and (almost) completely unexposed (specificity). In the present study, however, misclassification of exposure and reduced specificity must be assumed. Some overindebted and therefore exposed individuals may have been incorrectly assigned to the unexposed group of not overindebted individuals, because a representative random sample of the general or resident population was used, which, in all probability must include a representative proportion of overindebted individuals but in which these individuals could not have been identified and therefore were classified as not overindebted. If such a misclassification of the dichotomous exposure is not systematically related to or dependent on the outcome, i.e., does not systematically differ between the “cases” and the “controls”, the bias is nondifferential, unsystematic and toward the null value (no association) [43]. This means that nondifferential exposure misclassification always leads to an underestimation of the true strength of the association between exposure and outcome. In other words, the strong association observed between overindebtedness and social disconnectedness (or poor mental health) would have been even stronger if the overindebted individuals in the representative sample of the general population could have been identified and correctly assigned to the exposed category. It is further important or desirable in epidemiological research that the reliability of an exposure assessment is preferably high, which means that the observed exposure status conforms largely to the true exposure status, i.e., that all or most of the study participants are assigned correctly to the exposed and the unexposed. If this is not the case, a misclassification bias is likely to occur. An exposure assessment is usually qualified and evaluated by the two quality criteria of sensitivity and specificity, which are high and unproblematic when those classified as exposed are truly and predominantly exposed (sensitivity) and those classified as unexposed are truly and (almost) completely unexposed (specificity). In the present study, however, misclassification of exposure and reduced specificity must be assumed. Some overindebted and therefore exposed individuals may have been incorrectly assigned to the unexposed group of not overindebted individuals, because a representative random sample of the general or resident population was used, which, in all probability must include a representative proportion of overindebted individuals but in which these individuals could not have been identified and therefore were classified as not overindebted. If such a misclassification of the dichotomous exposure is not systematically related to or dependent on the outcome, i.e., does not systematically differ between the “cases” and the “controls”, the bias is nondifferential, unsystematic and toward the null value (no association) [43]. This means that nondifferential exposure misclassification always leads to an underestimation of the true strength of the association between exposure and outcome. In other words, the strong association observed between overindebtedness and social disconnectedness (or poor mental health) would have been even stronger if the overindebted individuals in the representative sample of the general population could have been identified and correctly assigned to the exposed category. There is a possibility of a selection bias. Overindebted individuals who seek professional debt advice may systematically differ in their social or mental health status from other overindebted individuals who do not register as clients of a debt advisory center. The reasons of some overindebted individuals not to get help from debt advisory centres or to indirectly self-exclude from the study or survey participation might be systematically related to their social connectedness or mental health condition. Lonely and socially isolated or strongly depressed people in financial distress may be less likely to register in a debt advisory center. This might somewhat limit the generalisability of the findings.However, such self-selection of less lonely and socially isolated or less severely distressed and depressed people would presumably lead to an underestimation rather than an overestimation of the found negative associations between overindebtedness and social connectedness on the one hand and overindebtedness and mental health on the other. There is a possibility of a selection bias. Overindebted individuals who seek professional debt advice may systematically differ in their social or mental health status from other overindebted individuals who do not register as clients of a debt advisory center. The reasons of some overindebted individuals not to get help from debt advisory centres or to indirectly self-exclude from the study or survey participation might be systematically related to their social connectedness or mental health condition. Lonely and socially isolated or strongly depressed people in financial distress may be less likely to register in a debt advisory center. This might somewhat limit the generalisability of the findings. However, such self-selection of less lonely and socially isolated or less severely distressed and depressed people would presumably lead to an underestimation rather than an overestimation of the found negative associations between overindebtedness and social connectedness on the one hand and overindebtedness and mental health on the other. Finally it should be noted that all data used in the study were collected before the COVID-19 pandemic. Study findings cannot blindly and unconditionally be transferred to the present and post-COVID time. Things have changed a lot since then. Overindebtedness, loneliness and mental health problems appear to have increased due to or at least in the follow-up of the pandemic although loneliness was rising and an epidemic of overindebtedness was observed even before the COVID-19 pandemic. But the mental health crisis that is recently reported and particularly for Switzerland’s youth is strongly suspected to have occurred afterwards and as a direct consequence of the COVID-19 pandemic. However, this incident may have changed the relevance and prevalence of certain phenomena but not necessarily associations between them on an individual level. Conclusion Self-reported reasons of overindebted individuals for their financial situation have already indicated and suggested that overindebtedness is usually or at least frequently the consequence of a combination and cumulation of critical life events and resulting loss experiences (loss of job and social status, loss of health, loss of social relations, and loss or lack of financial autonomy). It might be reasonably assumed, therefore, that it is partly the relative lack or loss of social connectedness or integration and support, as suggested in this study, and partly the loss of one’s sense of control, as explored and shown in previous studies [10, 26], and as a result of cumulative life events that are occurring ahead of and along with overindebtedness and that make overindebted individuals feel unsupported, isolated, lonely and helpless and thus mentally ill sooner or later. To prevent people in severe financial hardship and relative poverty from becoming increasingly mentally ill, which further aggravates their misery, they should not only receive professional debt advice and a real chance and realistic prospect of debt relief in the not too distant future but also, for example, psychological help and support from other persons affected and/or from professionals and experts from social welfare and/or health care systems. From a societal and public health point of view, it is important to acknowledge that overindebted individuals are or at least feel strongly socially isolated and excluded, that they are growing in numbers and by far not (anymore) a marginal phenomenon and that overindebtedness is a severe risk factor for social well-being and connectedness and for mental health. Overindebted individuals therefore need not only instrumental and practical help to deal with arrears and outstanding payments but also emotional support and professional psychological or psychiatric help to cope with strong and frequent negative feelings of loneliness, hopelessness, discouragement and depression. Moreover, as poor or overindebted people typically feel socially isolated and excluded and therefore are at risk of withdrawing socially, they will not search for or ask for such support or professional help. This also has to be taken into account in regard to prevention measures. Acknowledgements Special thank must go to Joanna Herzig who initiated and conducted the survey among the clients of the four official debt advisory centers, developed the questionnaire together with the author and collected a small but important part of the data upon which this study is based. Author’s contributions OH conceptualized the study, merged and prepared the data, and subsequently analyzed the pooled data, and finally drafted and revised the manuscript. Funding The study received no funds from any third-party donor or funding institution in the public, commercial or non-profit sectors. Data availability The collection of the survey data among the 219 overindebted clients of the debt advisory centers was not supported by any third-party funds. Therefore these self-collected data were not intended for passing on to third parties, and its use is basically limited to the responsible scientists and the institutions involved in the data collection (Debt Prevention Zurich, University of Zurich). Nevertheless, data can be provided as raw data in Excel-file format on request. The secondary data from the Swiss Health Survey additionally used in this study are periodically collected and provided by the Swiss Federal Statistical Office. They are publicly funded and in principle freely accessible. However, they are made available for research purposes only, upon request and on a contractual basis (to ensure data protection and prevent unauthorized data transfer) and by disclosing the research aims. Declarations Ethics approval and consent to participate The study did not require ethics approval because all self-reported survey data used were not from patients, did not include medical records and were voluntarily and completely anonymously collected and therefore do not fall within the scope of Switzerland’s Federal Act on Research involving Human Beings. This so-called Swiss Human Research Act does explicitly not apply to anonymously collected or anonymized health-related data (HRA Art. 2 / https://www.fedlex.admin.ch/eli/cc/2013/617/en). Self-reported survey data collected among the 219 overindebted adult residents of the Canton of Zurich therefore were granted exemption from ethics approval (Kanton Zürich, Kantonale Ethikkommission. BASEC-Nr. Req-2019-00173), and informed consent was obtained from all survey participants and overindebted individuals. No approval was required likewise for the use of the secondary and publicly accessible data of the Swiss Health Survey among randomly selected inhabitants of Switzerland which are periodically collected and provided by the Federal Statistical Office. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The collection of the survey data among the 219 overindebted clients of the debt advisory centers was not supported by any third-party funds. Therefore these self-collected data were not intended for passing on to third parties, and its use is basically limited to the responsible scientists and the institutions involved in the data collection (Debt Prevention Zurich, University of Zurich). Nevertheless, data can be provided as raw data in Excel-file format on request. The secondary data from the Swiss Health Survey additionally used in this study are periodically collected and provided by the Swiss Federal Statistical Office. They are publicly funded and in principle freely accessible. However, they are made available for research purposes only, upon request and on a contractual basis (to ensure data protection and prevent unauthorized data transfer) and by disclosing the research aims.