OUP user menu

Contribution of lifelong adverse experiences to social health inequalities: findings from a population survey in France

Emmanuelle Cambois, Florence Jusot
DOI: http://dx.doi.org/10.1093/eurpub/ckq119 667-673 First published online: 7 September 2010


Background: Recent research shows that adverse experiences, such as economic hardships or exclusion, contribute to deterioration of health status. However, individuals currently experiencing adverse experiences are excluded from conventional health surveys, which, in addition, often focus on current social situation but rarely address past adverse experiences. This research explores the role of such experiences on health and related social inequalities based on a new set of ad hoc questions included in a regular health survey. Methods: In 2004, the National Health, Health Care and Insurance Survey included three questions on lifelong adverse experiences (LAE): financial difficulties, housing difficulties due to financial hardship, isolation. Logistic regressions were used to analyse associations between LAE, current socio-economic status (SES) (education, occupation, income) and health status (self-perceived health, activity limitation, chronic morbidity), on a sample of 4308 men and women aged ≥35 years. Results: LAE were reported by 20% of the sample. They were more frequent in low SES groups but concerned >10% of the highest income group. LAE increased the risk of poor self-perceived health, diseases and activity limitations, even after controlling for current SES [odds ratio (OR) > 2]. LAE experienced only during childhood are also linked to health. LAE account for up to 32% of the OR of activity limitations associated with the lowest quintile among women and 26% among men. Conclusions: LAE contribute to the social health gradient and explain variability within social groups. It is useful to take lifetime social factors into account when monitoring health inequalities.

  • France
  • health inequalities
  • lifelong adverse experiences
  • routine health surveys


In France, social inequalities in mortality are large and persistent over time.1–3 Education, occupation and income, which reflect current social and material context, are predictors of mortality and are significantly associated with the risk of diseases or disability.4–7 However, life-course epidemiology points out that past trajectories are an important health determinant in addition to current socio-economic status (SES). The accumulation of exposure to risk factors over the life course (childhood deprivation, damaging work conditions, etc.) and their impact at critical periods of life contribute to health deterioration.8–14

Regarding the impact of social context during childhood, studies show clear associations with mortality risk,10,15 as well as various health problems such as chronic diseases and mental health problems.16–18 In France, childhood conditions are associated with mortality, poor functional health, poor self-assessed health or obesity.19–22 The health impact of childhood circumstances not only has to do with material deprivation, but also isolation, social support and attachment.23,24

In later life, adverse experiences such as hardship, downward occupational mobility or family disruptions are also significantly associated with poor health or high mortality risks.25–28 In France, mortality is associated with occupational careers or experience of isolation.22,29–31

Such adverse experiences can lead to social marginalization and can contribute, together with the current context of hardship and deprivation, to excess health risks for groups of the population considered as excluded. For Shaw and colleagues, social exclusion ‘refers not only to the economic hardship of relative economic poverty but also incorporates the notion of the process of marginalization’.32 The triggers of a marginalization process, such as job loss, migration, isolation or conflict could indeed be health damaging through a disruption in social networks, habits and support. Research in the field of exclusion and health in France has shown that the poor health status of specific groups such as homeless people or free health-care centre users, can not only be explained by the lack of material resources, poor living conditions, inadequate access to health care and health-damaging behaviours, but may also be due to psychosocial factors such as lack of emotional and social support, poor self-esteem and life control.33–35 For these groups of people, the combination of current material deprivation, psychosocial disadvantage and past experiences of disruptions and failures that led to exclusion might explain their poor health. Moreover, whilst these circumstances may be only temporary, having undergone them over the life course might still be health damaging.

Therefore, lifelong adverse experiences (LAE), leading or not to exclusion, can increase health risks beyond the current social context. They are important social factors to be considered in monitoring population health, but routine statistics lack accurate tools. The link between adverse experiences and health is generally analysed through ad hoc surveys or cohorts, which are not fully representative of the general population. Indeed, regular survey samples tend to exclude people currently experiencing them (not living in households, hard to reach, not willing to participate, etc.). And when population-based surveys incorporate information on past adverse experiences, collected through biographic tools, they have limited data on health outcomes. Finally, biographic tools are generally too large to be included on a regular basis in population health surveys. In order to analyse the link between adverse experiences and health in general population, a short set of questions on LAE was introduced in a population health survey in France in 2004. The questions relate to selected situations of material and social difficulties that might have occurred during the whole life. Therefore, this survey allows considering adverse experiences that may be temporary and do not systematically result in permanent social exclusion. This study firstly explores whether LAE are associated with health status beyond the current social situation. Second, it aims to assess how far LAE contribute to social health inequalities.


The National Health, Health Care and Insurance Survey

The National Health, Health Care and Insurance Survey (ESPS: ‘Enquête sur la Santé et la Protection Sociale’) is a biennial health interview survey coordinated by the Institute for Research and Information on Health Economics (IRDES), with a sample based on an ongoing random sample of French major health insurance beneficiaries (covering >95% of the population of private households). In 2004, ~40% of households sampled could not be reached (mostly due to incomplete or wrong addresses); 70% of the contacted households agreed to participate.36 Initially, households were contacted by telephone to obtain a key respondent to answer the core questionnaire eliciting the demographics of the household members and a selection of questions including, in the 2004 wave, the set of questions on LAE. As a second step, a self-completion health questionnaire was sent to each household member for return by mail. In 2004, 75% of the initial sample returned the questionnaire.


LAE in the general private household population were assessed through three questions aimed at identifying lifetime experiences of deprivation and hardship in terms of financial and housing difficulties, and social disruption through experiences of isolation. The question wording was as follows: has the person ever, during his/her life, (i) ‘experienced serious financial difficulties so that he/she could not meet basic needs or that he/she did not cope with these difficulties’; (ii) ‘needed to move in with relatives or friends or to move into sheltered housing as a result of financial difficulties’; (iii) ‘experienced a long-term period of isolation following an event such as a break-up, conflict or a move to another area or country’. In order to assess the long-term impact of LAE, for the last two questions, individuals were asked whether these experiences had occurred during childhood only, during adulthood only or both (financial difficulties might be less obvious to children and therefore less reliably reported as childhood experience by the surveyed persons).

Health indicators

We used the Eurostat Minimum European Health Module37 incorporated into this survey and which contains three questions covering complementary health dimensions: chronic morbidity (‘Do you have any chronic or long-lasting illness or health problem?’); self-perceived health (‘How is your health in general?’); long-term activity limitations (‘Because of health problems, to what extent have you been limited, for at least 6 months, in activities people usually do?’). Three binary indicators were built based on the three questions: reporting ‘chronic illness’ vs. ‘no chronic health illness’, reporting ‘being limited’ vs. ‘not limited’ and reporting ‘fair to very poor health’ vs. ‘good or very good’.

Indicators of current social status

To control for current SES, three indicators were used: education, income and occupation. Income was measured as household income, divided by the Organisation for Economic Co-operation and Development equivalence scale (1 for the first household member, 0.5 for the second one and 0.3 for the additional ones). Five quintiles were defined and an additional category was added corresponding to missing information (~12%). We considered four educational levels and the occupational status was measured by current occupation or the previous occupation for those retired or unemployed. We used the French occupational and social status classification: highly qualified occupations (professionals, managers and intellectual professions); skilled white-collar workers (nurses, elementary school teachers, technicians, etc.); farm owners; other self-employed (trade and craft business owner); trade and craft employees; clerical employees; skilled manual workers; unskilled manual workers and farm workers; without occupation (other than retired and unemployed).

Statistical method

Several logistic regressions adjusted for age were conducted separately for men and women.38

First, the association between LAE and each current SES indicator was analysed separately (Model 1) and simultaneously (Model 2) to assess the unequal risk of LAE.

Second, regression models 3 to 5 explore the association between LEA and the health dimensions. In Model 3, we analyse the health risks associated with each SES indicator and with LAE (univariate with control for age only). In order to confirm the association between LAE and health in all social groups, the analyses were separately reproduced in the lowest and the highest income groups. In Model 4, we analyse the health risks associated with SES indicators when they are simultaneously included in the model. Then, we include LAE in the Model 5 to see if its association with health remains significant beyond the current SES. As suggested by Van de Mheen and colleagues,39 the contribution of LAE to social health inequalities can be assessed by an index (Δ) being the percentage of decrease in the excess risk of poor health [odds ratio (OR) > 1] associated to SES between Models 4 and 5. Embedded Image (Δ computed on OR accurate to three decimal points).

Third, we analyse the association between each health indicator and the period in which the LAE occurred, without and with adjustment for SES indicators (respectively Models 6 and 7).

The LAE are based on self-reported retrospective information, and therefore respondents may be more likely to report LAE that have resulted in health problems and to omit others. Furthermore, respondents currently experiencing health problems and poor psychological well-being may be more likely to ‘darken’ their past (reconstruction phenomena).23 In order to control for this, we ran additional models adjusted for current psychological distress. As no reference mental health scale was available in this survey, we used the information on self-reported morbidity based on a list of diseases and symptoms that are coded and classified (i.e. reporting depression or symptoms of depression, such as anxiety) plus the information on self-reported medication classified a posteriori by physician (intake of medication coded as psychotropic).


This study is based on 1915 men and 2393 women, aged ≥35 years who responded to both the background and health questions. Table 1 provides the distribution of the sample by SES status, LEA and health indicators. In our sample, 20% of women and 18% of men reported one or more LAE, mostly long-term periods of isolation and financial difficulties. One third of those who reported housing difficulties and/or isolation (12% of women and 14% of men) experienced them in childhood only. Adverse experiences in both childhood and adulthood were rare.

View this table:
Table 1

Descriptive analysis of the study sample (2004 ESPS survey)

Age and SESN (%)N (%)
Age group, years
    35–44518 (27.1)691 (29.0)
    45–54509 (26.6)650 (27.2)
    55–64392 (20.5)442 (18.5)
    65–74291 (15.2)330 (13.8)
    ≥75205 (10.7)280 (11.7)
Level of education
    Primary (Educ 1)*425 (22.2)6230 (26.3)
    Lower secondary (Educ 2)788 (41.2)843 (35.2)
    Higher secondary (Educ 3)236 (12.3)359 (15.0)
    Post-secondary (Educ 4)466 (24.3)561 (23.4)
Occupational class
    Highly qualified occupations (High qual.)*390 (20.4)220 (9.2)
    Skilled white collar occupations (White col.)376 (19.6)486 (20.3)
    Farmers118 (6.2)101 (4.2)
    Self-employed (Self-empl.)185 (9.7)110 (4.6)
    Clerical employees (Adm. empl.)114 (6.0)603 (25.2)
    Trade and craft employees (Tr. empl.)33 (1.7)435 (18.2)
    Skilled manual workers (Sk. MW)555 (29.0)163 (6.8)
    Unskilled manual workers (Unsk. MW)144 (7.5)185 (7.7)
    Inactive (No occ.)– (–)90 (3.8)
Income group
    1st quintile (Quintile 1)*255 (13.3)373 (15.6)
    2nd quintile (Quintile 2)284 (14.8)413 (17.6)
    3rd quintile (Quintile 3)346 (18.1)417 (17.4)
    4th quintile (Quintile 4)340 (17.8)422 (17.6)
    5th quintile (Quintile 5)470 (24.5)465 (19.4)
    Unknown220 (11.5)303 (12.7)
Health indicators
Self-perceived health
    Very good or good1379 (72.0)1599 (66.8)
    Fair, poor or very poor536 (28.0)794 (33.2)
Chronic diseases
    None1271 (66.4)1562 (65.3)
    At least one644 (33.6)831 (34.7)
Activity limitation
    Not limited1514 (79.1)1855 (77.5)
    Limited401 (20.9)538 (22.5)
LAE (At least one)
    Have ever experienced339 (17.7)485 (20.3)
Financial problem
    Have ever experienced155 (8.1)246 (10.3)
Housing problem
    Have ever experienced94 (4.9)138 (5.8)
        in childhood only27 (1.4)53 (2.2)
        in adulthood only65 (3.4)85 (3.6)
        in childhood and adulthood2 (0.1)– (–)
Period of isolation
    Have ever experienced176 (9.2)259 (10.8)
        in childhood only83 (4.3)121 (5.1)
        in adulthood only82 (4.3)114 (4.8)
        in childhood & adulthood11 (0.6)24 (1.0)
Housing or/and isolation
    Have ever experienced231 (12.1)332 (13.9)
        in childhood only91 (4.8)131 (5.5)
        in adulthood only121 (6.3)167 (7.0)
        in childhood and adulthood19 (1.0)34 (1.4)
Number of LAE
    Only one type of experiences268 (14.0)349 (14.6)
    Only two types of experiences56 (2.9)114 (4.8)
    Only three types of experiences15 (0.8)22 (0.9)
Total1915 (100)2393 (100)
  • *Abbreviations used in following tables

LAE were reported in all SES groups with a decreasing gradient with increasing income, educational level or qualified occupation (table 2). However, LAE are reported by >10% of our population in the highest income quintiles or qualified occupations; more frequently for women than for men. Note that farmers reported less LAE than any other occupational groups. LAE remained strongly associated with current SES after controlling for age (Model 1). Only income remained strongly associated with LAE when the other SES indicators were simultaneously controlled for (Model 2). Farmer is the only occupation with a remaining (negative) relationship, while much of the association with education disappeared.

View this table:
Table 2

Associations between LAE and level of education, occupation and income

LAE frequencyModel 1aModel 2bLAE frequencyModel 1aModel 2b
N (%)OR [95% CI]OR [95% CI]N (%)OR [95% CI]OR [95% CI]
Educ 458 (12.5)1.01.094 (16.8)1.01.0
Educ 341 (17.4)1.5 [1.0–2.3]1.2 [0.8–2.0]66 (18.4)1.2 [0.8–1.6]1.0 [0.7–1.4]
Educ 2150 (19.0)1.7 [1.2–2.3]1.1 [0.7–1.6]180 (21.4)1.4 [1.1–1.9]1.1 [0.7–1.5]
Educ 190 (21.2)2.2 [1.5–3.3]1.4 [0.8–2.2]145 (23.0)2.3 [1.6–3.1]1.5 [0.9–2.3]
High qual.48 (12.3)1.01.037 (16.8)1.01.0
White col.55 (14.6)1.2 [0.8–1.9]1.0 [0.6–1.6]76 (15.6)0.9 [0.6–1.4]0.8 [0.5–1.3]
Farmers10 (8.5)0.7 [0.3–1.4]0.4 [0.2–0.8]6 (5.9)0.4 [0.2–1.0]0.2 [0.1–0.4]
Self-empl.38 (20.5)1.8 [1.2–2.9]1.4 [0.8–2.3]19 (17.3)1.2 [0.7–2.3]0.8 [0.4–1.5]
Ad. empl.20 (17.5)1.5 [0.9–2.7]1.0 [0.6–1.9]127 (21.1)1.3 [0.9–2.0]0.9 [0.6–1.5]
Tr. empl.9 (27.3)2.7 [1.2–6.1]1.7 [0.7–4.2]107 (24.6)1.7 [1.1–2.6]0.9 [0.6–1.6]
Sk. MW127 (22.9)2.1 [1.5–3.0]1.4 [0.9–2.2]32 (19.6)1.3 [0.8–2.2]0.7 [0.4–1.4]
Unsk MW32 (22.7)2.0 [1.2–3.3]1.1 [0.6–1.9]52 (28.1)2.1 [1.3–3.4]1.0 [0.6–1.8]
No occ.29 (32.2)3.0 [1.7–5.4]1.4 [0.7–2.7]
Quintile 550 (10.6)1.01.062 (13.3)1.01.0
Quintile 445 (13.2)1.3 [0.8–2.0]1.2 [0.7–1.8]62 (14.7)1.1 [0.8–1.6]1.1 [0.7–1.6]
Quintile 368 (19.7)2.0 [1.4–3.0]1.8 [1.2–2.7]82 (19.7)1.6 [1.1–2.3]1.4 [1.0–2.2]
Quintile 263 (22.2)2.4 [1.6–3.6]2.1 [1.4–3.3]90 (21.8)1.9 [1.3–2.7]1.7 [1.1–2.5]
Quintile 171 (27.8)3.2 [2.2–4.8]3.0 [1.9–4.7]138 (37.0)4.0 [2.9–5.7]3.8 [2.6–5.6]
Unknown42 (19.1)2.0 [1.3–3.2]1.8 [1.2–3.0]51 (16.8)1.4 [0.9–2.1]1.3 [0.9–2.0]
  • Men and women aged ≥35 years (Abbreviations = see full legends and associated abbreviations in Table 1); Italic = the difference is not statistically significant (95%)/Bold = OR statistically differs from 1 (95%)

  • a: Univariate logistic regression, adjusted on age only

  • b: Multivariate logistic regression, adjusted on age, education level, occupation and income

In our study population, 34% of men and women reported chronic disease or health problems, 28% of men and 33% of women reported fair-to-poor self-perceived health and 21% of men and 22% of women reported long-term activity limitations (table 1).

Model 3 shows that current low SES was significantly associated with poor self-perceived health and activity limitations for both men and women. LAE was also linked to poor health, for the three health dimensions, with a more than doubling of the ORs compared with those who did not report LAE (table 3).

View this table:
Table 3

ORs of poor health associated with education, occupation, income and LAE

Poor self-perceived healthAt least one chronic diseaseActivity limitations
Model 3aModel 4bModel 5b(c)Model 3aModel 4bModel 5b(c)Model 3 (a)Model 4 (b)Model 5 (b)(c)
OR [95% CI]OR [95% CI]OR [95% CI]Δ %OR [95% CI]OR [95% CI]OR [95% CI]Δ %OR [95% CI]OR [95% CI]OR [95% CI]Δ %
Educ 31.3 [0.9–2.0]1.0 [0.7–1.6]1.0 [0.7–1.6]1.0 [0.7–1.4]1.0 [0.7–1.4]1.0 [0.7–1.4]1.1 [0.7–1.7]1.0 [0.6–1.5]0.9 [0.6–1.5]
Educ 21.9 [1.4–2.6]1.2 [0.8–1.7]1.2 [0.8–1.8]0.8 [0.6–1.0]0.8 [0.6–1.1]0.8 [0.5–1.1]1.4 [1.0–1.9]1.0 [0.7–1.6]1.0 [0.7–1.6]
Educ 12.5 [1.8–3.5]1.4 [0.9–2.1]1.4 [0.9–2.1]0.9 [0.7–1.3]0.9 [0.6–1.4]0.9 [0.6–1.4]1.8 [1.3–2.6]1.2 [0.8–2.0]1.2 [0.8–1.9]
High qual.
White col.1.4 [0.9–2.0]1.2 [0.8–1.8]1.2 [0.8–1.8]0.9 [0.6–1.2]1.0 [0.6–1.3]0.9 [0.7–1.3]1.3 [0.9–1.9]1.2 [0.8–1.8]1.2 [0.8–1.8]
Farmers1.3 [0.8–2.1]0.8 [0.4–1.4]0.8 [0.5–1.5]0.6 [0.4–1.0]0.6 [0.4–1.0]0.7 [0.4–1.1]1.4 [0.8–2.3]0.9 [0.5–1.6]1.0 [0.6–1.8]
Self-empl.1.6 [1.1–2.5]1.3 [0.8–2.0]1.2 [0.8–2.0]0.8 [0.5–1.2]0.9 [0.6–1.3]0.8 [0.5–1.3]1.2 [0.7–1.9]0.9 [0.6–1.6]0.9 [0.5–1.5]
Ad. empl.2.8 [1.7–4.6]2.2 [1.3–3.9]2.2 [1.3–3.9]11.5 [0.9–2.3]1.6 [1.0–2.7]1.6 [1.0–2.7]–12.0 [1.2–3.5]1.7 [0.9–3.0]1.6 [0.9–3.0]
Tr. empl.2.1 [0.8–5.2]1.4 [0.6–3.8]1.4 [0.5–3.5]1.0 [0.5–2.4]1.2 [0.5–2.8]1.1 [0.5–2.6]1.6 [0.6–4.4]1.2 [0.4–3.5]1.1 [0.4–3.3]
Sk. MW2.9 [2.0–4.0]2.0 [1.3–3.1]2.0 [1.3–3.0]60.9 [0.7–1.2]1.0 [0.7–1.5]1.0 [0.7–1.4]1.7 [1.2–2.5]1.4 [0.9–2.1]1.3 [0.8–2.0]
Unsk MW2.9 [1.8–4.6]1.7 [1.0–2.9]1.7 [0.9–2.9]00.9 [0.6–1.4]0.9 [0.5–1.5]0.9 [0.5–1.5]2.4 [1.4–3.9]1.5 [0.9–2.8]1.5 [0.9–2.8]
Quintile 41.3 [0.9–1.9]1.0 [0.7–1.4]1.0 [0.7–1.4]0.9 [0.6–1.2]0.9 [0.7–1.3]0.9 [0.6–1.3]1.2 [0.8–1.8]1.0 [0.7–1.6]1.0 [0.7–1.5]
Quintile 31.6 [1.1–2.2]1.1 [0.7–1.6]1.0 [0.7–1.5]0.8 [0.6–1.2]0.9 [0.6–1.2]0.8 [0.6–1.2]1.4 [0.9–2.0]1.1 [0.7–1.7]1.0 [0.7–1.6]
Quintile 22.3 [1.6–3.2]1.6 [1.1–2.3]1.5 [1.0–2.2]200.9 [0.6–1.3]1.0 [0.7–1.4]0.9 [0.6–1.3]1.7 [1.2–2.5]1.4 [0.9–2.2]1.3 [0.8–2.0]
Quintile 13.1 [2.2–4.5]2.3 [1.5–3.5]2.0 [1.4–3.1]191.2 [0.9–1.7]1.4 [0.9–2.0]1.2 [0.8–1.8]2.6 [1.8–3.9]2.2 [1.4–3.4]1.9 [1.2–2.9]26
Unknown1.7 [1.2–2.6]1.4 [0.9–2.1]1.3 [0.9–2.0]0.8 [0.6–1.2]0.9 [0.6–1.3]0.8 [0.6–1.2]1.4 [1.0–2.2]1.3 [0.8–2.0]1.2 [0.8–1.8]
No LAE1.
LEA2.3 [1.8–3.0]2.0 [1.5–2.6]2.0 [1.6–2.6]2.0 [1.6–2.6]2.7 [2.0–3.6]2.4 [1.8–3.2]
Quintile 1
No LAE1.
LEA1.0 [0.6–1.8]1.0 [0.5–1.8]2.1 [1.2–3.8]2.1 [1.1–3.9]2.1 [1.1–3.9]2.1 [1.1–4.0]
Quintile 5
No LAE1.
LEA1.6 [0.8–3.3]1.6 [0.8–3.4]2.3 [1.2–4.3]2.4 [1.3–4.6]1.4 [0.6–3.2]1.5 [0.7–3.4]
Educ 31.3 [0.9–1.8]1.0 [0.7–1.4]1.0 [0.7–1.4]0.8 [0.6–1.1]0.8 [0.6–1.1]0.8 [0.6–1.1]1.3 [0.9–1.8]1.1 [0.7–1.7]1.1 [0.7–1.7]
Educ 21.8 [1.4–2.4]1.1 [0.8–1.6]1.1 [0.8–1.6]0.8 [0.7–1.0]0.8 [0.6–1.1]0.8 [0.6–1.1]1.5 [1.1–2.0]1.2 [0.8–1.7]1.2 [0.8–1.8]
Educ 12.8 [2.1–3.8]1.5 [1.0–2.2]1.4 [1.0–2.1]100.7 [0.5–0.9]0.7 [0.5–1.0]0.7 [0.5–1.0]1.7 [1.2–2.4]1.3 [0.8–2.0]1.2 [0.8–1.9]
High qual.
White col.1.0 [0.7–1.5]0.9 [0.6–1.3]0.9 [0.6–1.3]1.0 [0.7–1.4]1.0 [0.7–1.5]1.0 [0.7–1.4]1.0 [0.6–1.5]0.8 [0.5–1.3]0.8 [0.5–1.3]
Farmers1.3 [0.7–2.2]0.6 [0.3–1.0]0.7 [0.4–1.3]0.4 [0.3–0.7]0.4 [0.2–0.8]0.5 [0.3–0.9]0.9 [0.5–1.6]0.5 [0.2–0.9]0.6 [0.3–1.1]
Self-empl.1.5 [0.9–2.5]0.9 [0.5–1.7]1.0 [0.5–1.7]0.7 [0.4–1.1]0.7 [0.4–1.2]0.7 [0.4–1.2]1.6 [0.9–2.8]1.1 [0.6–2.0]1.1 [0.6–2.0]
Ad. empl.1.8 [1.2–2.7]1.3 [0.8–2.0]1.3 [0.8–2.0]0.9 [0.6–1.2]1.0 [0.7–1.5]1.0 [0.7–1.5]1.3 [0.8–2.0]0.9 [0.6–1.5]0.9 [0.5–1.5]
Tr. empl.2.5 [1.7–3.8]1.4 [0.9–2.3]1.4 [0.9–2.3]0.9 [0.6–1.2]1.0 [0.6–1.5]1.0 [0.6–1.5]1.7 [1.1–2.6]1.1 [0.6–1.8]1.1 [0.6–1.8]
Sk. MW2.3 [1.4–3.8]1.3 [0.7–2.2]1.3 [0.8–2.3]0.8 [0.5–1.3]1.0 [0.6–1.6]1.0 [0.6–1.7]1.1 [0.6–1.9]0.7 [0.4–1.3]0.7 [0.4–1.3]
Unsk MW3.4 [2.2–5.5]1.7 [1.0–3.0]1.7 [1.0–3.0]21.1 [0.7–1.7]1.2 [0.7–2.0]1.2 [0.7–2.0]1.8 [1.1–2.9]1.0 [0.6–1.9]1.0 [0.6–1.9]
No Occ.2.3 [1.3–4.1]1.1 [0.6–2.1]1.1 [0.6–2.0]0.8 [0.5–1.4]0.8 [0.5–1.5]0.8 [0.5–1.5]1.8 [1.0–3.2]1.1 [0.5–2.1]1.0 [0.5–2.0]
Quintile 41.7 [1.2–2.3]1.4 [1.0–2.0]1.4 [1.0–2.0]51.3 [0.9–1.7]1.4 [1.0–1.8]1.4 [1.0–1.8]1.8 [1.2–2.6]1.7 [1.2–2.5]1.7 [1.1–2.5]2
Quintile 32.2 [1.6–3.0]1.6 [1.1–2.3]1.6 [1.1–2.2]90.7 [0.5–1.0]0.8 [0.6–1.1]0.8 [0.6–1.1]1.7 [1.2–2.5]1.6 [1.1–2.4]1.5 [1.0–2.3]10
Quintile 22.3 [1.7–3.2]1.7 [1.2–2.4]1.6 [1.1–2.2]151.0 [0.7–1.3]1.1 [0.8–1.6]1.1 [0.8–1.5]1.9 [1.3–2.7]1.7 [1.1–2.5]1.6 [1.1–2.4]14
Quintile 14.4 [3.1–6.0]3.3 [2.3–4.7]2.7 [1.9–3.9]251.3 [1.0–1.8]1.7 [1.2–2.4]1.4 [1.0–2.0]362.7 [1.9–3.9]2.5 [1.7–3.8]2.0 [1.3–3.1]32
Unknown1.9 [1.3–2.7]1.6 [1.1–2.3]1.5 [1.1–2.2]81.1 [0.8–1.4]1.2 [0.9–1.7]1.2 [0.8–1.7]1.9 [1.3–2.8]1.8 [1.2–2.7]1.8 [1.2–2.7]8
No LAE1.
LEA2.9 [2.3–3.6]2.3 [1.9–3.0]2.0 [1.6–2.5]1.9 [1.5–2.4]3.1 [2.4–3.9]2.8 [2.2–3.5]
Quintile 1
No LAE1.
LEA3.4 [2.1–5.6]3.1 [1.8–5.2]2.4 [1.5–4.0]2.3 [1.4–4.0]4.2 [2.4–7.3]4.2 [2.4–7.3]
Quintile 5
No LAE1.
LEA2.9 [1.6–5.5]3.1 [1.6–5.9]1.4 [0.8–2.5]1.3 [0.7–2.5]1.6 [0.8–3.5]1.6 [0.8–3.7]
  • Men and women, aged ≥35 years (Abbreviations = see full legends and associated abbreviations in Table 1); Italic = the difference is not statistically significant (95%)/Bold = OR statistically differs from 1 (95%)

  • a: Univariate logistic regression, adjusted on age only

  • b: Multivariate logistic regression, adjusted on age, education level, occupation and income

  • c: Percentage decrease in ORs significantly higher than 1: Δ = (OR model 4 – OR model 5)/(OR model 4 – 1). Calculations based on ORs accurate to three decimal points

In the multivariate Model 4, only low income and low qualified occupations remained significantly associated with self-perceived health and low income to activity limitation. The excess risks were significant for all the income quintiles for women. Including LAE, in Model 5, induces only a slight attenuation of the occupational and income group’s OR. LAE was still strongly associated with deteriorated health for both sexes and for each health indicator while controlling for other SES.

Finally, focusing on the lowest income group, the association between LAE and poor health status remains significant for most health indicators (except self-perceived health among men). In the highest income group, LAE still impacted significantly on poor perceived health among women and chronic diseases in among men (table 3).

The comparison of the OR associated with SES indicators in Models 4 and 5 shows to what extent LAE contributed to the excess risk associated with the SES. It actually explained a small part of the inequalities for male clerical employees and unskilled manual workers (respectively 1 and 6% of the OR for poor self-perceived health), but up to a quarter of the excess risks associated with the lowest income quintile (20% of the OR for poor self-perceived health and 26% of the OR for activity limitations). For women, the contribution of LAE to excess risk in the lowest income group is much larger, explaining up to 25% of the OR for poor self-perceived health, 32% for activity limitations and 36% for chronic diseases.

More detailed analysis (available upon request) showed that each of the three types of LAE contributed to the overall association with health (borderline significant effects for perceived health and chronic disease for men). Moreover, incorporating indicators of current psychological distress led to a slight (not significant) reduction in the link between LAE and health, suggesting this may contribute to the association.

Finally, Models 6 and 7 provided evidence of the long-lasting influence of LAE (table 4): after controlling for age and current social status, both LAE occurring during adulthood only and LAE occurring during childhood only were significantly associated with the risk of poor health (except poor self-perceived health for men reporting LAE in adulthood only). The results suggest a cumulative risk associated with having experienced LAE both in childhood and in adulthood.

View this table:
Table 4

ORs of poor health associated with period of housing difficulties or long period of isolation

Poor self-perceived healthAt least one chronic diseaseActivity limitations
Model 6aModel 7bModel 6aModel 7bModel 6aModel 7b
OR [95% CI]OR [95% CI]OR [95% CI]OR [95% CI]OR [95% CI]OR [95% CI]
LAE never1.
    in childhood only2.4 [1.5–3.8]2.3 [1.5–3.7]1.9 [1.2–3.0]2.0 [1.3–3.1]2.4 [1.5–3.9]2.4 [1.5–3.9]
    in adulthood only1.6 [1.1–2.5]1.4 [0.9–2.2]1.7 [1.2–2.6]1.7 [1.1–2.6]2.3 [1.5–3.6]2.1 [1.4–3.3]
    in childhood and adulthood16 [5.1–53]15 [4.5–51]6.1 [2.2–17]6.0 [2.1–17]8.1 [3.0–22]6.7 [2.4–19]
LAE never1.
    in childhood only2.4 [1.6–3.5]2.3 [1.6–3.4]2.0 [1.4–2.9]1.9 [1.3–2.8]2.1 [1.4–3.1]2.0 [1.3–3.0]
    in adulthood only2.2 [1.5–3.0]1.7 [1.2–2.5]1.9 [1.3–2.6]1.8 [1.3–2.6]3.4 [2.4–4.8]3.0 [2.1–4.3]
    in childhood and adulthood7.1 [3.3–15]6.1 [2.8–13]3.4 [1.7–6.8]3.3 [1.6–6.9]6.1 [3.0–13]5.9 [2.9–12]
  • Men and women, aged ≥35 years; Italic = the difference is not statistically significant (95%)/Bold = OR statistically differs from 1 (95%)

  • a: Univariate logistic regression, adjusted on age only

  • b: Multivariate logistic regression, adjusted on age, education level, occupation and income


This study shows that 20% of the population over age 35 years reported LAE and provide evidence of a strong and long-lasting association between LAE and deteriorated health for a number of health dimensions. LAE are more frequent in the most disadvantaged groups. They largely contribute to the income gradient in health and also explain a part of the excess risks associated with low qualified occupations. However, we also found >10% LAE in the highest income quintile and they are significantly linked to poor perceived health among female advantaged groups and chronic diseases among male advantaged groups.

Although 20% reporting LAE seems high, it appears to be consistent with estimates from the late 1990s for France.40 Furthermore, this 20% may even be an underestimate due to the survey methodology. Since the ESPS is a household survey which, like most conventional population health surveys, was conducted on a selected population who could be contacted and agreed to participate, it misses those who are currently experiencing adverse circumstances, specifically people not living in a household, and those who did not respond due to social and/or health problems. Second, the study sample excludes persons who did not return the health questionnaire, which may be similarly related to social and/or health problems. However, as we had data from the background questionnaire, we could test the magnitude of bias due to non-response to the health questionnaire by considering that non-respondents were (i) all in poor health status and (ii) all in good health status. Neither of these scenarios significantly changed our conclusions: the effect of LAE on health was slightly increased with the ‘missing in good health’ assumption and decreased with the ‘missing in poor health’ assumption. Our questions may also overestimate LAE related health risks due to their retrospective nature and the possible a posteriori reconstruction effect, with those in poor health and distress being more likely to darken their past experiences. However, our results were not significantly modified by incorporating information on current psychological distress.

Despite these limitations, LAE were found to be strongly associated with poor health status for a number of health dimensions. The higher risk of poor health could be due to various possible determinants: deleterious effect of economic hardship,11,12,23,41 stressful events,11,13,14 job loss,31 disruption or isolation.24,26,28,30 The increased health risks may also be explained by a reverse causation process: long-term health problems may have been responsible for adverse experiences such as job loss,42 decreasing earnings,43,44 isolation, family breakups, etc. However, our results show that LAE occurring during childhood only were significantly associated with the risk of poor health status to the same extent or more so than LAE occurring in adulthood only. The findings support the hypothesis of a causal influence of LAE on health status given that LAE reported only in childhood may be less suspected of reverse causality. Furthermore, the results suggest a cumulative impact of having experienced LAE at several periods of the life course.

Being much more frequent in lower social groups, LAE are a risk factor that contributes to the social health gradient, consistent with previous research.45 With regard to current SES, LAE are strongly associated with low income; first, because current economic hardship is one of the dimensions directly included in our LAE measurement. Secondly, past periods of economic hardship also accounted for in the LAE measurement, are known to be predictors of subsequent economic hardships.44 Given the strong association between LAE and income and the strong association between income and health status, our findings show a major contribution of adverse life events to the income health inequalities, especially among women, for whom they account for about a third of the OR of chronic diseases and activity limitations associated with the lowest income quintile.

Interestingly, the adjustment for SES induced only a slight decrease in the association between LAE and health status. This finding stresses the fact that current SES and LAE do not correspond to the same social health determinants. LAE bring different insights, including past experiences that might not have conducted to permanent hardship but still impact health, independently of current social situation. This study highlights that LAE are an important health determinant above and beyond the current socio-economic situation. It pleads for the inclusion of this life social factor in the set of social dimensions regularly used for monitoring social health inequalities as well as inequalities within social groups. Further analyses on health-care use, health-related behaviours and health trajectories in relation to LAE are also needed to properly design public health policies on equity.

Key points

  • We explore the association between LAE and health status in the French general population and analyse their contribution to social health inequalities.

  • The use of a short set of LAE questions included in a French population health survey in 2004 identifies 20% of adults having an increased risk of poor health.

  • LAE is significantly associated with poor health status, beyond current social status

  • LAE experienced in childhood only are still a determinant of health in adulthood

  • LAE are more frequently reported by low socio-economic groups and contribute to social health inequalities.


View Abstract