Abstract
Purpose:
Relatively little is known about how working-age life course cumulative exposure to employment intensity and job complexity informs older-age cognitive function. We investigate these associations, separately for men and women, and net of known confounders.
Methods:
Using retrospective lifetime employment histories of Europeans born 1923–1959 (2004–2009, N=22 266), we calculate cumulative working-age exposure to non-employment, full-time and part-time employment, and a professional occupation. In gender-stratified linear regression models, these indicators predict cognitive function score based on the DemTect scale and Mini Mental State Exam.
Results:
Non-employment ≥ 25% of the working life course was associated with poorer cognitive function for men by -.43 points (95% CI = -.79, -.06) on a 19-point scale. Women’s full-time employment, even if ≤ 25% of the working lifetime, was associated with a cognitive advantage over never-employment by .60 points (95% CI = .17, 1.02). Compared to predominantly non-professionally employed men, those working professionally for ≥ 75% of the life course had better cognition by .38 points (95% CI = .16, .60).
Conclusions:
This paper provides novel evidence that older-age cognitive functioning is associated with cumulative exposure to both employment intensity and complexity, but that these relationships vary by sex.
Keywords: cognitive reserve, life history, employment, women’s health, men’s health
INTRODUCTION
Cognitive stimulation at work increasingly is viewed as a mid-life determinant of cognitive health at older ages.1,2,3 Employment is a major time use for most adults for a large fraction of their lives, with considerable heterogeneity across employment situations in their mental demands and incentives to continue developing “cognitive reserve”,2,4,5 or the brain’s capacity to optimize performance amidst aging, likely due to prior, repeated utilization of specific brain regions—e.g. in response to stimulation.6,7 Past research suggests consequences of the duration of time spent employed, here labeled employment intensity, and duration exposed to complex work. However, past studies have not considered the confluence of varying intensity of exposure obtained via full versus part time work, cumulative time out of the labor force, cumulative exposure to complex employment, and potential sex differences in their distribution or association with cognitive function.
Non-employment limits exposure to cognitively stimulating employment activities and social interaction with co-workers, potentially negatively impacting cognitive health. Moreover, non-employment due to illness may have different implications for cognitive reserve than homemaking or other activities. Past research linking the intensity of employment to later cognitive function almost exclusively studies women, because many born before 1960 were not employed. Research finds that European women who spent little time working for pay had poorer cognitive function than women who worked more consistently.3,8 Research using retrospective reports of any earlier-life employment gaps - but not accounting for their duration - finds for older men and women that unemployment is negatively associated with cross-sectional measures of older age cognition but with slower aging-related cognitive decline over 2 years, while illness and homemaking have indirect, negative effects, working through poorer older-age health and socio-economic status.9 Moreover, gender differences in associations are not observed9. Research on men finds that mid-life unemployment and its duration (measured up to 12 months) are negatively associated with contemporaneous self-reported cognitive performance.10 Several studies show that leaving employment via retirement at the statutory age is associated with faster cognitive decline than remaining employed.11,12
Regarding the intensity of labor market attachment, full-time employment may increase cognitive reserve by providing cognitive stimulation in employment tasks. However, part time employment could have different implications; women’s part-time employment may comparatively strengthen cognitive reserve by providing employment stimulation with less work-family strain. For European women, one study finds that mostly working part-time throughout their lives is cognitively advantageous compared to mostly working full-time, but further research is needed to understand the importance of duration spent in full versus part-time work6. Overall, little is known about the potential cumulative consequences of non-employment across a lifetime, nor about cumulative intensity of paid employment or potential gender differences in associations.
More cognitive stimulation in complex work should comparatively increase cognitive reserve among the employed. Some studies have used fine-grained measures of job complexity, capturing dimensions of work with data, people, and things.2,4,5,13 They find higher cognitive functioning at older ages among European men and women with greater complexity in main lifetime occupation.4,5,13 A recent study pooling older American men and women found that cognitive health was better for those consistently employed in high skill jobs, compared to individuals employed in low-skill jobs.1 Research has also shown that duration of employment in a complex job is negatively linked with dementia onset2, with no significant differences by sex. Moreover, Europeans who retired as early as possible, largely men in low-skill jobs, showed lower cognitive losses after retirement than those who retired later from more complex jobs, suggesting differential benefits of employment based on complexity.11
Importantly, selection processes may shape life course employment histories and later cognitive function. Illness early in life may lead to lowered employment intensity or complexity, and might also have direct effects on later cognition. Individuals with poor innate cognitive abilities may select into less complex or less secure jobs, while individuals with strong innate cognitive abilities may select into more stimulating jobs with fewer interruptions 14, 15. One study found that non-employment predicted two year change in older-age cognition partially due to selection into non-employment.9
Historical and demographic factors contextualize these theoretical expectations. Employment histories of contemporary older men and women are highly gendered but also vary in intensity and complexity.16 The 1920s to 1950s birth cohorts led marked changes in women’s and men’s employment patterns across the 20th century. Men17,18 and a growing fraction of women23 experienced occupational expansion into professional or high-skilled jobs. Most men experienced steady, full-time employment across the life course with modest amounts of lifetime non-employment.19 Women from these cohorts were employed more years of their lives than previous cohorts and increasingly varied in their post-childbearing levels of attachment to the labor market and full-time versus part-time employment.20,21,22 Nonetheless, sizable percentages of women in these cohorts did not work for pay for much of their life course.19,21 These changing patterns mean that for contemporary older men in wealthy economies, life course exposure to cognitively relevant exposures at work varied by complexity of their work and experiences of non-employment. Contemporary older women experienced variation in duration of attachment across the life course, intensity of full- versus part-time employment over those attached years, and complexity of their work.
However, prior studies capturing employment intensity and complexity across midlife have either categorized employment pattern histories by type—masking duration of time spent in particular employment conditions—or examined narrow windows of the life course, rather than the total cumulative exposure. We consider whether cognitive function at older ages varies by the percentage of a working-age lifetime spent employed, whether the association depends on how much was full-time versus part-time employment, and whether any association depends on the fraction that was worked in a professional occupation, our marker of complexity. We conduct analyses separately for men and women 50 and older, providing insight for an important set of birth cohorts entering later life.
METHODS
Data
This study used nationally representative data from release 7.0 of The Survey of Health, Ageing and Retirement in Europe (SHARE) — a longitudinal panel study of individuals ages 50 and older from 27 European countries and Israel. Wave 1 data were collected in 2004–05 and Wave 2 in 2006–07. Wave 3, known as SHARELIFE, was collected in 2008–09 as a retrospective life history that includes detailed employment information. We analyze data from countries that participated in the first three waves: Austria, Belgium, Czech Republic, Denmark, France, Germany, Greece, Ireland, Italy, Netherlands, Poland, Spain, Sweden, and Switzerland. Measures of cognitive function and all other predictors were drawn from Wave 2 unless they were only available in Wave 1. The sample is restricted to individuals ages 50–85 at time of interview who - if they ever married and/or worked for pay - did so for the first time before age 50. We selected this age range to increase potential variation in cognitive function while minimizing the effects of selective survival. From 25 236 eligible respondents, we dropped respondents missing on employment characteristics for more than 1% of their working-age lifetime (N = 1575) or on childhood socioeconomic status (N = 825). The analytic sample contains 22 266 respondents with complete data from Wave 3 and Wave 1 or 2.
Measures
Our measure of cognitive function is based on the DemTect scale24 and the mini-mental-state-exam (MMSE),25 adapted for SHARE.26 Summing five scaled responses for delayed word recall (0–4), immediate word recall (0–4), numeracy (0–4), orientation (0–4), and verbal fluency (0–3), this score ranges from 0–19. Lower scores indicate worse cognitive functioning.
Measures of cumulative lifetime employment characteristics were constructed from the Wave 3 retrospective work history panel created using SHARE guidelines.27 For each year from age 15 to year of interview, information is provided on part-time or full-time employment status and occupation type. For each year, individuals are categorized as an “employee or self-employed” in a job that lasted 6 months or longer (=1) versus 17 other non-employment options including “unemployed,” “sick or disabled,” and “looking after home or family” (=0). For each year of a job spell, respondents report whether their working hours are “part-time” or “full-time.” Individuals who note multiple changes between part-time and full-time in a given job spell are coded as part-time for the entire spell. For each year of employment we created an indicator of professional job = 1 for the SHARE-aggregated categories of ‘professional’ and ‘legislative’ jobs, and = 0 for aggregated categories of ‘manual worker / elementary occupation,’ ‘clerical’ ‘technician,’ ‘sales job,’ ‘skilled agricultural,’ ‘crafts worker,’ and ‘armed forces.’
From the year of full-time educational completion to age 50, each respondent’s sum of years occupying each of these statuses (i.e., not employed, employed full time, employed part time, and professional) is divided by the number of observation years, producing cumulative percentage measures. We chose to use a relative measure of time in employment since educational completion, in contrast to an absolute measure of years because those who stayed in formal education longer likely worked fewer years between the ages of 18 and 50. For these birth cohorts, only a negligible percentage of men but relatively high percentages of women were not employed or worked part-time more than temporarily.16 Accordingly, measures of employment intensity are sex specific.
For men, we created an indicator of relatively weak labor market attachment distinguishing those who were not employed ≥ 25% of lifetime (yes = 1, no = 0). Men who were non-employed were usually unemployed or in military service. We created two variables capturing women’s intensity of labor force attachment: percentage of lifetime worked full-time (never worked for pay, worked full-time 0–24.9% of lifetime, 25–74.9%, and 75–100%) and worked part-time ≥ 25% of lifetime (yes = 1, no = 0). When considered together, these two indicators leave as the reference category women who never worked for pay, or worked part time for less than 25% of the observed period and never worked full time. Women who were not employed were typically homemaking or unemployed. To capture employment complexity for men and women, we created an indicator for those who worked a professional job ≥ 75% of lifetime (yes = 1, no = 0). Cutpoints of 25 and 75% were chosen for these indicators to capture reasonable points in the population distribution and ensure large enough groups for comparison.
We adjust for several early life characteristics that influence both employment characteristics and cognitive function. Highest educational degree has three categories, collapsed from the original seven categories of the 1997 International Standard Classification of Education (ISCED): less than secondary school, completed secondary school/vocational training, and some university or more. A scale of childhood advantage at age 10 ranges from 0 to 4, created by summing four binary variables: had any advantaged household feature (fixed bath, running cold water, running hot water, inside toilet, or central heating), had at least one bookcase in the household, no overcrowding (a bedroom for every person), and occupation of the primary breadwinner was managerial/professional. These four items scaled unidimensionally using the Mokken scaling procedure (Loevinger H coefficient = 0.49, p<.001). We also adjust for the respondent’s age (centered at the mean), square of centered age, nativity status (1 = non-native born, 0 = native-born), country of residence, and whether ever married (yes = 1, no = 0), and ever had children (yes = 1, no = 0).
Analytic Strategy
We present descriptive statistics by sex, assessing the significance of the difference for each variable with univariate regressions. We estimate sex-stratified linear regression models predicting cognitive score, adjusting for centered age, centered age-squared and the interactions between these terms and country of residence to account for documented cross-national variation in age-related declines in cognition.28 Other adjustments include educational attainment, early life advantage, nativity and marital and parental status. To account for survey design, we use SHARE-provided individual-level probability weights and cluster standard errors by household and by country using Stata Version 15 SVY commands. After regression model estimation, we estimate average marginal effects and assess pairwise contrasts.
RESULTS
Table 1 shows that European men ages 50–85 have slightly higher average cognitive function (15.2) than women (14.6). About 8% of men spent 25% or more of their working age lifetime not employed and the remainder worked full time for 75–100%. Among women, 14.3% never worked for pay and 15.0% worked part-time for 25% or more of their working-age lifetime. The modal experience for women was working full-time 75%-100% of their lifetime (35.3%), followed by having worked full-time 25%-74.9% of their lifetime (29.8%), and then by having worked full-time 0–25% of a lifetime (20.7%)—the latter two categories potentially including women who worked part-time at some points. About one in ten men worked a professional job for 75% or more of their lifetime (10.7%), compared to only about one in twenty women (5.8%). These differences motivate stratification of the multivariable analysis.
Table 1.
Weighted Characteristics of Respondents Ages 50–85 by Gender: Survey of Health, Aging, and Retirement in Europe (SHARE), Waves 1–3, 2004–2009
| Men (n=9788) | Women (n=12 478) | ||||
|---|---|---|---|---|---|
| Characteristic | Mean or % (SE) | Mean or % (SE) | p | ||
|
| |||||
| Cognition Score (0–19) | 15.2 (0.5) | 14.6(0.6) | 0.00 | ||
| Not Employed 25% or More of Working Lifetime | 8.2 (0.0) | N/A | |||
| Worked Part-Time 25% or More of Working Lifetime | N/A | 15 (0.0) | |||
| Percentage of Working Lifetime Worked Full-Time | N/A | ||||
| Never Worked | 14.3 (0.0) | ||||
| Worked Full-Time 0–24.9% | 20.7 (0.0) | ||||
| Worked Full-Time 25.0–74.9% | 29.8 (0.0) | ||||
| Worked Full-Time 75.0–100% | 35.3 (0.1) | ||||
| Worked a Professional Job 75% or More of Working Lifetime | 10.7 | (0.0) | 5.8 | (0.0) | 0.00 |
| Age at Interview | 64.2 | (0.3) | 65.5 | (0.2) | 0.00 |
| Non-Native Born | 6.7 | (0.0) | 6.9 | (0.0) | 0.59 |
| Childhood Advantage Score (0–4)a | 1.4 | (0.2) | 1.3 | (0.1) | 0.01 |
| Education | |||||
| Middle School or Less | 37.8 | (0.1) | 51.2 | (0.1) | 0.00 |
| Some High School to Completed High School / Vocational School | 39.2 | (0.1) | 33.9 | (0.1) | 0.00 |
| Some College or More | 23.3 | 14.9 | (0.0) | 0.00 | |
| Ever Married | 91.9 | (0,0) | 94.2 | (0.0) | 0.01 |
| Ever Had Children | 85.7 | lo.O) | 89.2 | (0.0) | 0.00 |
| Country | |||||
| Austria | 2.1 | (0.0) | 2.1 | (0.0) | 0.77 |
| Belgium | 3 | (0.0) | 3 0 | (0.0) | 0.47 |
| Czech Republic | 2.8 | (0.0) | 3.0 | (0.0) | 0.39 |
| Denmark | 1.6 | (0.0) | 1.5 | (0.0) | 0.37 |
| East Germany | 5.6 | (0.1) | 5.1 | (0.1) | 0.34 |
| France | 15.5 | (0.1) | 16.4 | (0.2) | 0.25 |
| Greece | 3.2 | (0.0) | 3.0 | (0.0) | 0.39 |
| Ireland | 0.8 | (0.0) | 0.7 | (0.0) | 0.62 |
| Italy | 18.3 | (0.2) | 18.2 | (0.2) | 0.95 |
| Netherlands | 4.4 | (0.1) | 4.1 | (0.0) | 0.36 |
| Poland | 8.7 | (0.1) | 9.8 | (0:1) | 0.29 |
| Spain | 10.6 | (0.1) | 11.2 | (0.1) | 0.31 |
| Sweden | 2.5 | (0.0) | 2.3 | (0.0) | 0.37 |
| Switzerland | 2 | (0.0) | 2.0 | (0.0) | 0.47 |
| West Germany | 18.8 | (0.2) | 17.6 | (0.2) | 0.19 |
Note. Working Lifetime is considered from educational completion to age 50.
Higher value indicates higher childhood socioeconomic status
Table 2 presents the focal linear regression results for men with adjustments for all covariates, and Table 3 presents these results for women. The full set of coefficients is available in Appendix Table 1 and 2. Table 2 shows that net of known determinants including educational degree, men who were not employed for 25% or more of their working-age lifetime have poorer cognitive function (b = −0.43; 95% confidence interval [CI] = −0.79, −0.06) than men employed 75 –100%. Men who worked a professional job for 75% or more of their lifetimes have better cognitive function (b = 0.38; 95% CI = 0.16, 0.60).
Table 2.
Association between Cognitive Score and Intensity and Complexity of Lifetime Labor Market Attachment for Men Ages 50–85 (n=9788), Survey of Health, Aging, and Retirement in Europe (SHARE), Waves 1–3, 2004–2009
| Variable | b | (95% CI) |
|---|---|---|
| Not Employed 25% or More of Lifetimea | −0.43 | (−0.79, −0.06) |
| Worked a Professional Job 75% or More of Lifetimeb | 0.38 | (0.16, 0.60) |
| Non-Native Born (Ref: Native Born) | −0.70 | (−1.14, −0.25) |
| Ever Married | 0.69 | (0.34, 1.04) |
| Ever Had Children | 0.14 | (−0.12, 0.40) |
| Education (Ref: Middle School or Less) | ||
| Some High School to Completed High School / Vocational School | 1.36 | (1.05, 1.67) |
| Some College or More | 2.11 | (1.80, 2.43) |
| Childhood Advantage | 0.33 | (0.22, 0.44) |
| Constant | 13.32 | (12.83, 13.81) |
Note.CI = confidence interval. Model also includes a three-way interaction between centered age, centered age squared, and country. Full regression output is available in Appendix Table 1.
Reference Category: Worked 75.1% or More of Lifetime (Not Employed 24.9% or Less of Lifetime)
Reference Category: Only Worked a Non-Professional Job for 75.1% or More of Lifetime or Worked a Professional Job for 74.9% or Less of Lifetime, Combined with a Non-Professional Job for .2% to 25.1% of Lifetime
Table 3.
Association between Cognitive Score and Intensity and Complexity of Lifetime Labor Market Attachment for Women Ages 50–85 (n=12,478), Survey of Health, Aging, and Retirement in Europe (SHARE), Waves 1–3, 2004–2009
| Variable | b | (95% CI) |
|---|---|---|
| Percentage of Lifetime Worked Full-Time (Ref: Never Worked) | ||
| Worked Full-Time 0–24.9% | 0.60 | (0.17, 1.02) |
| Worked Full-Time 25.0–74.9% | 0.82 | (0.48, 1.17) |
| Worked Full-Time 75.0–100% | 0.85 | (0.65, 1.06) |
| Worked Part-Time 25% or More of Lifetimea | 0.14 | (−0.09, 0.38) |
| Worked a Professional Job 75% or More of Lifetimeb | 0.17 | (−0.03, 0.37) |
| Non-Native Born (Ref: Native Born) | −0.19 | (−0.83, 0.45) |
| Ever Married | 0.10 | (−0.19, 0.39) |
| Ever Had Children | 0.18 | (−0.27, 0.63) |
| Education (Ref: Middle School or Less) | ||
| Some High School to Completed High School / Vocational School | 1.39 | (0.97, 1.81) |
| Some College or More | 1.91 | (1.54, 2.28) |
| Childhood Advantage | 0.43 | (0.25, 0.62) |
| Constant | 13.64 | (13.11, 14.17) |
Note.CI = confidence interval. Model also includes a three-way interaction between centered age, centered age squared, and country. Full regression output is shown in Appendix Table 2.
Reference Category: Worked Part-Time for 24.9% or Less of Lifetime
Reference Category: Only Worked a Non-Professional Job for Any Percentage of Lifetime or Worked a Professional Job for 74.9% or Less of Lifetime, Combined with a Non-Professional Job for 0% to 25.1% of Lifetime
Table 3 shows that employment intensity is associated with cognitive score for women as well. Women who worked full-time 0–24.9% of their working life course (and possibly part-time for a short period) have higher cognitive scores (b = 0.60; 95% CI = 0.17, 0.12) than women who never worked, and this is also true for women who worked full-time 25.0–74.9% (b = 0.82; 95% CI = 0.48, 1.17) or 75.0–100% (b = 0.85; 95% CI = 0.65, 1.06). Linear combination tests (not shown) demonstrate that the important distinction is between those who were never employed and all other levels of employment intensity. Women who worked part-time for 25% or more of their lifetime (and who did not work a substantial fraction of their lives full-time) do not have different cognitive scores (b = 0.14; 95% CI = −0.09, 0.38) than women who worked part-time for less than 25% of their life course or who did not work for pay. However, net of all else, women who worked a professional job for 75% or more of their lifetime do not have better cognitive function (b = 0.17; 95% CI = −0.03, 0.37).
To enhance interpretation, we predicted cognitive scores for individuals with common combinations of employment intensity and complexity, shown in Table 4, and made pairwise comparisons. Women who never worked for pay have the lowest predicted cognitive scores— compared to women who worked any combination of part-time or full-time professional or non-professional employment. Among women working for pay, there are only a few differences, but we observe the joint importance of employment intensity and complexity: professional women who worked full-time for at least one quarter of their life course have higher predicted scores relative to non-professional women who worked full-time for less than a quarter of their life course. We observe substantially higher predicted cognitive scores for men who worked professionally for 75% or more of their lifetime, compared to predominantly non-professional men with little time non-employed, who themselves have substantially higher scores than non-professional men who were not employed for 25% or more of their working lifetimes.
Table 4.
Predicted Cognitive Function Score for Men and Women Ages 50–85 across Combined Cumulative Lifetime Employment Characteristics, Survey of Health, Aging and Retirement in Europe (SHARE), Waves 1–3, 2004–2009
| WOMEN | |||
|
| |||
| Never Worked | 14.37 (14.13, 14.61) N=1834 | ||
| Worked Professionally ≥75% | Worked Non-Professionally | ||
| Worked Full-Time 0–24.9% | 15.14(14.81, 15.48) N=98 | 14.97 (14.74, 15.21) N=2472**,* | |
| Worked Full-Time 25–74.9% | 15.37 (15.17, 15.57) N=76** | 15.20(15.07, 15.32) N=3511 | |
| Worked Full-Time 75–100% | 15.40(15.19, 15.60) N=642* | 15.23 (15.07, 15.38) N=3845 | |
| Worked Part-Time ≥ 25% | 15.33 (15.06, 15.61) N=169 | 15.16(14.96, 15.36) N=1808 | |
| MEN | |||
|
| |||
| Worked Professionally ≥75% | Worked Non-Professionally | ||
| Non-Employed < 25% | 15.82(15.61, 16.02) N=1213 | 15.44(15.41, 15.47) N=7838 | |
| Non-Employed ≥ 25% | N/Aa | 15.01 (14.66, 15.36) N=736 | |
Note. Estimation at the sample means for all other variables in the model is utilized.
p = .03 between ‘Worked Professionally, Full-Time 75–100%’ and ‘Worked Non-Professionally, Full-Time 0–24.9%’
p = .00 between ‘Worked Professionally, Full-Time 25–74.9%‘ and ‘Worked Non-Professionally, Full-Time 0–24.9%’
N=1, thus excluded from estimation.
As a robustness check to rule out potential poor health selection into life course non-employment, we estimated regression models that excluded individuals with a non-employment gap characterized by sickness in the early-career. Results were not substantively different (details and results available on request).
DISCUSSION
We assessed the association between older age cognitive function and cumulative lifetime employment intensity and job complexity among European women and men. Prior research demonstrates the importance of employment-related stimulation for older age cognitive health, yet the roles of cumulative employment intensity and complexity over an entire working-age lifetime are less clear. Moreover, prior research does not account for sex differences in employment histories for pre-1960 birth cohorts or how their association with cognitive function may vary by sex.
We found that substantial non-employment accumulated over the working life course, predominantly spent in unemployment, was associated with poorer cognitive function for men. The little past research including men used a pooled sample of men and women and found that experiencing at least one earlier-life employment gap due to unemployment or illness was negatively associated with older-age cognition.9 Considering duration more explicitly, we found a strong cognitive disadvantage—close to half a point on a 19 point scale—for men who spent a quarter or more of their lifetime non-employed.
Prior research also demonstrated that older women with very little to no work experience have an older-age cognitive health disadvantage compared to steadily working women.3,8 We found supportive evidence that compared to any full-time work experience, non-employment, predominantly spent in homemaking, was associated with substantially poorer cognition at older ages. Prior research also found that mothers who did not work for pay have a cognitive disadvantage compared to those who primarily worked part-time, but did not account for exposure duration.8 Our predictions consider employment intensity and complexity and demonstrated that women who worked part-time for at least a quarter of the life course, whether professionally or non-professionally, were advantaged relative to women who did not work for pay. These findings for men and women highlight the value of examining cumulative non-employment over an entire working lifetime. They also suggest that for women, even moderate amounts of employment across the life course, part or full-time, may be cognitively protective.
Prior research demonstrated that longer exposure to a complex job is associated with better older-age cognition1 and negatively associated with dementia onset;2 we considered differences between men and women and specified more clearly the duration of exposure. Predicted values showed that women’s cognitive advantage in job complexity is tied to their employment intensity—professional women who worked full-time for at least one quarter of their lifetime were advantaged compared to non-professional women who worked full-time for less than one quarter of their lifetime. Predicted values also showed even greater gaps between men who worked in professional jobs at least 75% of their working life course, non-professional men, and men who were not in the labor force for 25% or more, and demonstrated the value of stratifying by gender.
While these findings advance our understanding of links between employment and cognition, future research can make further contributions. To capture a range of cognitive functioning among older persons, we included a range of birth cohorts (1923 to 1959). Sensitivity analyses demonstrated birth cohort heterogeneity in exposure to employment characteristics for women, and in their associations with cognition for men and women (Appendix Tables B.1, B.2, and B.3). The nature of lifetime employment characteristics and their associations with cognitive health will continue to change over historical time, meriting theoretical explanations for underlying mechanisms. Thus, it is important for future research to sort out period versus aging effects by studying larger cohorts and tighter bands of calendar time. Our measures of occupational type limit examination of heterogeneity within large categories of professional or non-professional employment. Other studies2,29 with access to detailed occupational titles have begun exploring specific aspects of work tasks, but given the low percentages of men and women who worked professional jobs in these cohorts, we pooled Europeans in this analysis. Europe’s regional diversity may be associated with demonstrated heterogeneity in women’s employment histories across European countries,16,19 and in the work and family policies that influence the impacts of employment on health, pointing toward the importance for future research to examine cross-national variation in the relationship between cumulative employment characteristics and cognitive health and develop supporting theories. Future research should also develop measures to capture innate cognitive capacity and early life health that may predict non-employment or complex employment. Lastly, future research should explore each component of the cognition measure. Particular employment characteristics may be beneficial for specific components of cognitive health, producing different intervention suggestions.
Highlights.
Accumulating substantial time in non-employment across the life course is associated with poorer cognitive function for older men and women.
Having worked for even a small portion of the life course—whether full-time or part-time—is cognitively advantageous for older women compared to never having worked.
There are gender differences in the older-age cognitive advantage of predominantly working a professional versus non-professional job across the life course; non-professional women’s cognitive disadvantage is tied to little time spent in full-time employment.
The associations between older-age cognitive function and cumulative lifetime employment intensity and job complexity are independent of educational attainment, childhood socioeconomic status, age, country, and marital and parental status.
Acknowledgements and Funding Disclosure:
This work was supported by the National Institute on Aging of the National Institutes of Health under Award Number 5T32AG033533 and T32AG000221. The authors gratefully acknowledge use of the services and facilities of the Population Studies Center at the University of Michigan, funded by the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the National Institutes of Health under Award Number P2CHD041028.
Abbreviations:
- CI
Confidence Interval
- ISCED
International Standard Classification of Education
- SHARE
Survey of Health, Aging, and Retirement in Europe
APPENDIX TABLES
Appendix Table A.1.
Association between Cognitive Score and Intensity and Complexity of Lifetime Labor Market Attachment for Men Ages 50–85 (n=9788), Survey of Health, Aging, and Retirement in Europe (SHARE), Waves 1–3, 2004–2009
| Variable | b | (95% CI) | |
|---|---|---|---|
| Not Employed 25% or More of Lifetime | −0.43 | (−0.79, −0.06) | |
| Worked a Professional Job 75% or More of Lifetime | 0.38 | (0.16, 0.60) | |
| Non-Native Born (Ref: Native Born) | −0.70 | (−1.14, −0.25) | |
| Ever Married | 0.69 | (0.34, 1.04) | |
| Ever Had Children | 0.14 | (−0.12, 0.40) | |
| Education (Ref: Middle School or Less) | |||
| Some High School to Completed High School / Vocational School | 1.36 | (1.05, 1.67) | |
| Some College or More | 2.11 | (1.80, 2.43) | |
| Childhood Advantage | 0.33 | (0.22, 0.44) | |
| Age at Interview (Centered) | −0.11 | (−0.11, −0.10) | |
| Age at Interview (Centered) Squared | 0.00 | (0.00, 0.00) | |
| Country (Ref: Austria) | |||
| Belgium | 0.15 | (0.05, 0.25) | |
| Czech Republic | −0.12 | (−0.20, −0.04) | |
| Denmark | 0.56 | (0.47, 0.66) | |
| East Germany | 0.10 | (0.03, 0.17) | |
| France | −0.19 | (−0.27, −0.12) | |
| Greece | 0.34 | (0.27, 0.41) | |
| Ireland | 0.05 | (−0.09, 0.19) | |
| Italy | −0.81 | (−0.93, −0.69) | |
| Netherlands | 0.53 | (0.40, 0.66) | |
| Poland | −1.42 | (−1.50, −1.35) | |
| Spain | −1.49 | (−1.64, −1.33) | |
| Sweden | 1.53 | (1.36, 1.70) | |
| Switzerland | 0.91 | (0.82, 1.00) | |
| West Germany | 0.49 | (0.45, 0.54) | |
| CountryXAge W3 (Centered) (Ref: Austria*Age 0 at Interview) | |||
| Belgium*Age W3 (Centered) | 0.00 | (−0.00, 0.01) | |
| Czech Republic*Age W3 (Centered) | 0.01 | (0.00, 0.01) | |
| Denmark*Age W3 (Centered) | 0.03 | (0.03, 0.04) | |
| East Germany*Age W3 (Centered) | 0.08 | (0.07, 0.08) | |
| France*Age W3 (Centered) | 0.02 | (0.01, 0.02) | |
| Greece* Age W3 (Centered) | 0.02 | (0.02, 0.03) | |
| Ireland* Age W3 (Centered) | 0.03 | (0.02, 0.03) | |
| Italy*Age W3 (Centered) | −0.02 | (−0.02, −0.01) | |
| Netherlands*Age W3 (Centered) | 0.04 | (0.04, 0.04) | |
| Poland*Age W3 (Centered) | −0.02 | (−0.02, −0.02) | |
| Spain*Age W3 (Centered) | −0.03 | (−0.03, −0.02) | |
| Sweden*Age W3 (Centered) | 0.08 | (0.08, 0.09) | |
| Switzerland*Age W3 (Centered) | 0.03 | (0.03, 0.04) | |
| West Germany*Age W3 (Centered) | 0.04 | (0.04, 0.05) | |
| CountryXAge W3 (Centered) Squared (Ref: Austria*Age 0 at Interview Squared) | |||
| Belgium*Age W3 (Centered) Squared | −0.01 | (−0.01, −0.01) | |
| Czech Republic*Age W3 (Centered) Squared | −0.00 | (−0.00, −0.00) | |
| Denmark*Age W3 (Centered) Squared | −0.01 | (−0.01, −0.01) | |
| East Germany*Age W3 (Centered) Squared | −0.00 | (−0.01, −0.00) | |
| France*Age W3 (Centered) Squared | −0.01 | (−0.01, −0.01) | |
| Greece* Age W3 (Centered) Squared | −0.01 | (−0.01, −0.01) | |
| Ireland* Age W3 (Centered) Squared | −0.01 | (−0.01, −0.01) | |
| Italy*Age W3 (Centered) Squared | −0.01 | (−0.01, −0.01) | |
| Netherlands*Age W3 (Centered) Squared | −0.00 | (−0.00, −0.00) | |
| Poland*Age W3 (Centered) Squared | −0.01 | (−0.01, −0.01) | |
| Spain*Age W3 (Centered) Squared | −0.01 | (−0.01, −0.01) | |
| Sweden*Age W3 (Centered) Squared | −0.01 | (−0.01, −0.01) | |
| Switzerland*Age W3 (Centered) Squared | −0.00 | (−0.00, −0.00) | |
| West Germany*Age W3 (Centered) Squared | −0.00 | (−0.00, −0.00) | |
| Constant | 13.32 | (12.83, 13.81) | |
Note. CI = confidence interval.
Appendix Table A.2.
Association between Cognitive Score and Intensity and Complexity of Lifetime Labor Market Attachment for Women Ages 50–85 (n=12 478), Survey of Health, Aging, and Retirement in Europe (SHARE), Waves 1–3, 2004–2009
| Variable | b | (95% CI) |
|---|---|---|
| Percentage of Lifetime Worked Full-Time (Ref: Never Worked) | ||
| Worked Full-Time 0–24.9% | 0.60 | (0.17, 1.02) |
| Worked Full-Time 25.0–74.9% | 0.82 | (0.48, 1.17) |
| Worked Full-Time 75.0–100% | 0.85 | (0.65, 1.06) |
| Worked Part-Time 25% or More of Lifetime | 0.14 | (−0.09, 0.38) |
| Worked a Professional Job 75% or More of Lifetime | 0.17 | (−0.03, 0.37) |
| Non-Native Born (Ref: Native Born) | −0.19 | (−0.83, 0.45) |
| Ever Married | 0.10 | (−0.19, 0.39) |
| Ever Had Children | 0.18 | (−0.27, 0.63) |
| Education (Ref: Middle School or Less) | ||
| Some High School to Completed High School / Vocational School | 1.39 | (0.97, 1.81) |
| Some College or More | 1.91 | (1.54, 2.28) |
| Childhood Advantage | 0.43 | (0.25, 0.62) |
| Age at Interview (Centered) | −0.05 | (−0.06, −0.05) |
| Age at Interview (Cenetered) Squared | −0.00 | (−0.00, −0.00) |
| Country (Ref: Austria) | ||
| Belgium | −0.51 | (−0.55, −0.47) |
| Czech Republic | −0.32 | (−0.50, −0.13) |
| Denmark | −0.20 | (−0.33, −0.07) |
| East Germany | −0.72 | (−0.90, −0.55) |
| France | −0.49 | (−0.54, −0.44) |
| Greece | −0.76 | (−0.91, −0.61) |
| Ireland | −1.30 | (−1.38, −1.22) |
| Italy | −1.68 | (−1.87, −1.49) |
| Netherlands | 0.35 | (−0.18, −0.51) |
| Poland | −2.02 | (−2.20, −1.84) |
| Spain | −2.83 | (−3.04, −2.61) |
| Sweden | 0.53 | (−0.43, −0.63) |
| Switzerland | 0.39 | (−0.17, −0.61) |
| West Germany | 0.20 | (−0.09, −0.31) |
| CountryXAge W3 (Centered) (Ref: Austria*Age 0 at Interview) | ||
| Belgium*Age W3 (Centered) | −0.04 | (−0.04, −0.03) |
| Czech Republic*Age W3 (Centered) | −0.06 | (−0.06, −0.05) |
| Denmark*Age W3 (Centered) | −0.02 | (−0.02, −0.01) |
| East Germany*Age W3 (Centered) | −0.01 | (−0.02, −0.01) |
| France*Age W3 (Centered) | −0.02 | (−0.03, −0.02) |
| Greece* Age W3 (Centered) | −0.07 | (−0.08, −0.07) |
| Ireland* Age W3 (Centered) | 0.00 | (−0.00, −0.01) |
| Italy*Age W3 (Centered) | −0.12 | (−0.12, −0.11) |
| Netherlands*Age W3 (Centered) | 0.00 | (−0.01, −0.00) |
| Poland*Age W3 (Centered) | −0.10 | (−0.10, −0.09) |
| Spain*Age W3 (Centered) | −0.13 | (−0.13, −0.12) |
| Sweden*Age W3 (Centered) | 0.03 | (0.02, 0.03) |
| Switzerland*Age W3 (Centered) | −0.01 | (−0.01, −0.00) |
| West Germany*Age W3 (Centered) | −0.02 | (−0.02, −0.01) |
| CountryXAge W3 (Centered) Squared (Ref: Austria*Age 0 at Interview Squared) | ||
| Belgium*Age W3 (Centered) Squared | −0.00 | (−0.00, −0.00) |
| Czech Republic*Age W3 (Centered) Squared | −0.00 | (−0.01, −0.00) |
| Denmark*Age W3 (Centered) Squared | 0.00 | (0.00, 0.00) |
| East Germany*Age W3 (Centered) Squared | 0.00 | (0.00, 0.00) |
| France*Age W3 (Centered) Squared | −0.00 | (−0.00, −0.00) |
| Greece* Age W3 (Centered) Squared | −0.00 | (−0.00, −0.00) |
| Ireland* Age W3 (Centered) Squared | 0.00 | (0.00, 0.00) |
| Italy*Age W3 (Centered) Squared | −0.00 | (−0.00, −0.00) |
| Netherlands*Age W3 (Centered) Squared | −0.00 | (−0.00, −0.00) |
| Poland*Age W3 (Centered) Squared | −0.00 | (−0.00, −0.00) |
| Spain*Age W3 (Centered) Squared | −0.00 | (−0.00, −0.00) |
| Sweden*Age W3 (Centered) Squared | 0.00 | (0.00, 0.00) |
| Switzerland*Age W3 (Centered) Squared | 0.00 | (0.00, 0.00) |
| West Germany*Age W3 (Centered) Squared | 0.00 | (−0.00, 0.00) |
| Constant | 13.64 | (13.11, 14.17) |
Note. CI = confidence interval.
Appendix Table B.1.
Weighted Employment Characteristics of Respondents Ages 50–85 by Birth Cohort and Gender: Survey of Health, Aging, and Retirement in Europe (SHARE), Waves 1–3, 2004–2009
| Characteristic | % (SE) | % (SE) | p | ||
|---|---|---|---|---|---|
|
| |||||
| MEN |
|||||
| < Age 65 (n=4823) | ≥ Age 65 (n=4965) | ||||
|
|
|||||
| Not Employed 25% or More of Working Lifetime | 7.6 | (0.0) | 9.0 | (0.0) | 0.50 |
| Worked a Professional Job 75% or More of Working Lifetime | 12.0 | (0.0) | 9.2 | (0.0) | 0.04 |
| WOMEN |
|||||
| < Age 64 (n=5999) | ≥ Age 64 (n=6479) | ||||
|
|
|||||
| Percentage of Working Lifetime Worked Full-Time | |||||
| Never Worked | 8.5 | (0.0) | 19.6 | (0.1) | 0.00 |
| Worked Full-Time 0–24.9% | 19.9 | (0.0) | 21.3 | (0.0) | 0.24 |
| Worked Full-Time 25.0–74.9% | 29.8 | (0.0) | 29.7 | (0.0) | 0.95 |
| Worked Full-Time 75.0–100% | 41.7 | (0.0) | 29.4 | (0.0) | 0.00 |
| Worked Part-Time 25% or More of Working Lifetime | 18.2 | (0.0) | 12.1 | (0.0) | 0.04 |
| Worked a Professional Job 75% or More of Working Lifetime | 7.7 | (0.0) | 4.1 | (0.0) | 0.00 |
Note. Working Lifetime is considered from educational completion to age 50.
Appendix Table B.2.
Association between Cognitive Score and Intensity and Complexity of Lifetime Labor Market Attachment for Men Ages 50–85, Stratified by Birth Cohort, Survey of Health, Aging, and Retirement in Europe (SHARE), Waves 1–3, 2004–2009
| < Age 65 (n=4823) | ≥ Age 65 (n=4965) | |||
|---|---|---|---|---|
|
| ||||
| Variable | b | (95% CI) | b | (95% CI) |
| Not Employed 25% or More of Lifetime | −0.49 | (−0.94, −0.03) | −0.31 | (−0.82, 0.20) |
| Worked a Professional Job 75% or More of Lifetime | 0.29 | (−0.09, 0.66) | 0.53 | (0.20, 0.85) |
| Non-Native Born (Ref: Native Born) | −1.33 | (−2.02, −0.65) | −0.04 | (−0.25, 0.18) |
| Ever Married | 0.76 | (0.55, 0.97) | 0.62 | (−0.09, 1.34) |
| Ever Had Children | 0.16 | (−0.07, 0.39) | 0.11 | (−0.52, 0.73) |
| Education (Ref: Middle School or Less) | ||||
| Some High School to Completed High School / Vocational School | 1.19 | (0.74, 1.64) | 1.58 | (1.32, 1.85) |
| Some College or More | 1.93 | (1.54, 2.32) | 2.38 | (1.95, 2.81) |
| Childhood Advantage | 0.28 | (0.19, 0.38) | 0.37 | (0.19, 0.54) |
| Age at Interview (Centered) | 0.26 | (0.22, 0.30) | −0.03 | (−0.05, −0.01) |
| Age at Interview (Centered) Squared | 0.03 | (0.03, 0.04) | −0.01 | (−0.01, −0.00) |
| Country (Ref: Austria) | ||||
| Belgium | 0.05 | (−0.10, 0.20) | 0.08 | (−0.04, 0.21) |
| Czech Republic | 1.04 | (0.89, 1.19) | −0.01 | (−0.10, 0.08) |
| Denmark | −0.08 | (−0.27, 0.11) | 0.08 | (−0.03, 0.19) |
| East Germany | 2.51 | (2.23, 2.78) | −0.02 | (−0.10, 0.06) |
| France | −0.71 | (−0.88, −0.53) | −0.54 | (−0.64, −0.45) |
| Greece | −1.24 | (−1.46, −1.02) | 0.64 | (0.51, 0.76) |
| Ireland | −2.27 | (−2.47, −2.07) | 0.12 | (0.04, 0.21) |
| Italy | −2.98 | (−3.22, −2.73) | −0.58 | (−0.74, −0.42) |
| Netherlands | 0.76 | (0.52, 1.00) | 0.31 | (0.18, 0.44) |
| Poland | −2.68 | (−2.80, −2.57) | −1.64 | (−1.87, −1.41) |
| Spain | −2.34 | (−2.55, −2.13) | −1.37 | (−1.54, −1.19) |
| Sweden | 0.34 | (0.08, 0.59) | 1.26 | (1.07, 1.44) |
| Switzerland | 1.03 | (0.91, 1.15) | 0.84 | (0.67, 1.02) |
| West Germany | 0.08 | (−0.02, 0.18) | 0.33 | (0.26, 0.41) |
| CountryXAge W3 (Centered) (Ref: Austria*Age 0 at Interview) | ||||
| Belgium*Age W3 (Centered) | −0.25 | (−0.28, −0.21) | −0.12 | (−0.14, −0.10) |
| Czech Republic*Age W3 (Centered) | 0.16 | (0.12, 0.20) | −0.11 | (−0.14, −0.09) |
| Denmark*Age W3 (Centered) | −0.34 | (−0.39, −0.29) | 0.17 | (0.14, 0.19) |
| East Germany*Age W3 (Centered) | 0.49 | (0.43, 0.56) | −0.04 | (−0.07, −0.01) |
| France*Age W3 (Centered) | −0.36 | (−0.41, −0.31) | 0.02 | (0.01, 0.04) |
| Greece*Age W3 (Centered) | −0.48 | (−0.53, −0.43) | −0.06 | (−0.08, −0.05) |
| Ireland*Age W3 (Centered) | −0.76 | (−0.80, −0.72) | −0.05 | (−0.07, −0.03) |
| Italy*Age W3 (Centered) | −0.76 | (−0.81, −0.71) | −0.17 | (−0.19, −0.15) |
| Netherlands*Age W3 (Centered) | −0.12 | (−0.16, −0.07) | 0.03 | (0.02, 0.05) |
| Poland*Age W3 (Centered) | −0.54 | (−0.58, −0.51) | −0.01 | (−0.04, 0.02) |
| Spain*Age W3 (Centered) | −0.33 | (−0.37, −0.30) | −0.03 | (−0.05, −0.02) |
| Sweden*Age W3 (Centered) | −0.42 | (−0.46, −0.38) | 0.11 | (0.10, 0.12) |
| Switzerland*Age W3 (Centered) | −0.17 | (−0.22, −0.11) | −0.09 | (−0.11, −0.07) |
| West Germany*Age W3 (Centered) | −0.27 | (−0.29, −0.24) | −0.04 | (−0.07, −0.02) |
| CountryXAge W3 (Centered) Squared (Ref: Austria*Age 0 at Interview Squared) | ||||
| Belgium*Age W3 (Centered) Squared | −0.03 | (−0.03, −0.03) | 0.01 | (0.00, 0.01) |
| Czech Republic*Age W3 (Centered) Squared | −0.00 | −(0.01, −0.00) | 0.01 | (0.00, 0.01) |
| Denmark*Age W3 (Centered) Squared | −0.04 | (−0.04, −0.03) | −0.01 | (−0.01, −0.01) |
| East Germany*Age W3 (Centered) Squared | 0.01 | (0.00, 0.01) | 0.00 | (0.00, 0.01) |
| France*Age W3 (Centered) Squared | −0.04 | (−0.04, −0.03) | −0.00 | (−0.01, −0.00) |
| Greece*Age W3 (Centered) Squared | −0.04 | (−0.04, −0.04) | −0.00 | (−0.00, 0.00) |
| Ireland*Age W3 (Centered) Squared | −0.06 | (−0.07, −0.06) | −0.00 | (−0.01, −0.00) |
| Italy*Age W3 (Centered) Squared | −0.06 | (−0.06, −0.05) | 0.01 | (0.00, 0.01) |
| Netherlands*Age W3 (Centered) Squared | −0.02 | (−0.02, −0.02) | 0.00 | (−0.00, 0.00) |
| Poland*Age W3 (Centered) Squared | −0.05 | (−0.05, −0.05) | −0.01 | (−0.01, −0.00) |
| Spain*Age W3 (Centered) Squared | −0.03 | (−0.04, −0.03) | −0.01 | (−0.01, −0.01) |
| Sweden*Age W3 (Centered) Squared | −0.05 | (−0.05, −0.04) | −0.01 | (−0.01, −0.01) |
| Switzerland*Age W3 (Centered) Squared | −0.03 | (−0.03, −0.02) | 0.01 | (0.01, 0.01) |
| West Germany*Age W3 (Centered) Squared | −0.03 | (−0.03, −0.03) | 0.00 | (0.00, 0.01) |
| Constant | 14.08 | (13.51, 14.65) | 13.23 | (12.73, 13.72) |
Note. CI = confidence interval.
Appendix Table B.3.
Association between Cognitive Score and Intensity and Complexity of Lifetime Labor Market Attachment for Women Ages 50–85, Stratified by Birth Cohort, Survey of Health, Aging, and Retirement in Europe (SHARE), Waves 1–3, 2004–2009
| < Age 64 (n=5999) | ≥ Age 64 (n=6479) | |||
|---|---|---|---|---|
|
| ||||
| Variable | b | (95% CI) | b | (95% CI) |
| Percentage of Lifetime Worked Full-Time (Ref: Never Worked) | ||||
| Worked Full-Time 0–24.9% | 1.01 | (0.55, 1.46) | 0.42 | (−0.13, 0.98) |
| Worked Full-Time 25.0–74.9% | 1.10 | (0.59. 1.60) | 0.76 | (0.27, 1, 26) |
| Worked Full-Time 75.0–100% | 1.44 | (0.97. 1.92) | 0.48 | (0.30, 0.65) |
| Worked Part-Time 25% or More of Lifetime | 0.40 | (0.17, 0.62) | −0.10 | (−0.61, 0.41) |
| Worked a Professional Job 75% or More of Lifetime | 0.16 | (−0.25, 0.56) | 0.23 | (−0.21, 0.68) |
| Non-Native Born (Ref: Native Born) | −0.79 | (−1.13, −0.45) | 0.28 | (−0.38, 0.94) |
| Ever Married | 0.19 | (−0.28, 0.65) | −0.03 | (−0.47, 0.42) |
| Ever Had Children | 0.22 | (−0.38, 0.83) | 0.17 | (−0.33, 0.66) |
| Education (Ref: Middle School or Less) | ||||
| Some High School to Completed High School / Vocational School | 1.48 | (1.15. 1.81) | 1.25 | (0.72, 1.78) |
| Some College or More | 1.00 | (1.62, 2.18) | 1.97 | (1.37, 2.57) |
| Childhood Advantage | 0.28 | (0.14, 0.42) | 0.57 | (0.35, 0.79) |
| Age at Interview (Centered) | 0.13 | (0.10, 0.16) | −0.24 | (−0.27, −0.22) |
| Age at Interview (Centered) Squared | 0.01 | (0.00, 0.00) | 0.01 | (0.01, 0.01) |
| Country (Ref: Austria) | ||||
| Belgium | −1.98 | (−2.15, −1.81) | −0.86 | (−0.96, −0.77) |
| Czech Republic | −0.96 | (−1.21, −0.70) | −0.35 | (−0.64, −0.05) |
| Denmark | −0.08 | (−0.29, 0.13) | −0.52 | (−0.65, − 0.39) |
| East Germany | −1.99 | (−2.32, −1.66) | −0.80 | (−1.01, −0.59) |
| France | −2.44 | (−2.57, −2.32) | −0.80 | (−0.89, −0.71) |
| Greece | −1.68 | (−2.00, −1.36) | −1.19 | (−1.35, −1.04) |
| Ireland | −0.95 | (−1.10, −0.80) | −1.24 | (−1.41, −1.07) |
| Italy | −1.01 | (−1.43, −0.58) | −1.77 | (−1.97, −1.58) |
| Netherlands | −0.69 | (−0.88, −0.50) | −0.13 | (−0.38, 0.13) |
| Poland | −3.17 | (−3.42, −2.93) | −1.70 | (−1.93, −1.46) |
| Spain | −4.04 | (−4.27, −3.82) | −3.27 | (−3.52, −3.03) |
| Sweden | −0.04 | (−0.27, 0.19) | 0.00 | (−0.13, 0.14) |
| Switzerland | 0.16 | (−0.18, 0.50) | −0.27 | (−0.58, 0.03) |
| West Germany | 0.01 | (−0.15, 0.17) | 0.01 | (−0.11, 0.13) |
| CountryXAge W3 (Centered) (Ref: Austria*Age 0 at Interview) | ||||
| Belgium*Age W3 (Centered) | −0.32 | (−0.35, −0.28) | 0.20 | (0.18, 0.22) |
| Czech Republic*Age W3 (Centered) | −0.08 | (−0.13, −0.03) | 0.16 | (0.14, 0.17) |
| Denmark*Age W3 (Centered) | 0.03 | (−0.01, 0.06) | 0.18 | (0.17, 0.19) |
| East Germany*Age W3 (Centered) | −0.28 | (−0.34, −0.23) | 0.09 | (0.09, 0.10) |
| France*Age W3 (Centered) | −0.53 | (−0.56, −0.49) | 0.11 | (0.10, 0.13) |
| Greece*Age W3 (Centered) | −0.3 | (−0.36, −0.25) | 0.15 | (0.13, 0.16) |
| Ireland*Age W3 (Centered) | 0.11 | (0.06, 0.15) | −0.09 | (−0.12, −0.05) |
| Italy*Age W3 (Centered) | 0.09 | (0.03, 0.16) | −0.00 | (−0.02, 0.01) |
| Netherlands*Age W3 (Centered) | −0.32 | (−0.35, −0.28) | 0.15 | (0.14, 0.17) |
| Poland*Age W3 (Centered) | −0.23 | (−0.26, −0.20) | −0.02 | (−0.04, −0.01) |
| Spain*Age W3 (Centered) | −0.47 | (−0.50, −0.44) | 0.04 | (0.02, 0.06) |
| Sweden*Age W3 (Centered) | −0.13 | (−0.18, −0.08) | 0.28 | (0.26, 0.30) |
| Switzerland*Age W3 (Centered) | −0.11 | (−0.15, −0.07) | 0.30 | (0.28, 0.31) |
| West Germany*Age W3 (Centered) | 0.00 | (−0.05, 0.06) | 0.14 | (0.11, 0.17) |
| CountryXAge W3 (Centered) Squared (Ref: Austria*Age 0 at Interview Squared) | ||||
| Belgium*Age W3 (Centered) Squared | −0.01 | (−0.01, −0.01) | −0.02 | (−0.02, −0.02) |
| Czech Republic*Age W3 (Centered) Squared | 0.00 | −(0.00, 0.00) | −0.02 | (−0.02, −0.02) |
| Denmark*Age W3 (Centered) Squared | 0.01 | (0.00, 0.01) | −0.01 | (−0.01, −0.01) |
| East Germany*Age W3 (Centered) Squared | −0.01 | (−0.02, −0.01) | −0.01 | (−0.01, −0.01) |
| France*Age W3 (Centered) Squared | −0.03 | (−0.03, −0.03) | −0.01 | (−0.01, −0.01) |
| Greece*Age W3 (Centered) Squared | −0.01 | (−0.01, −0.01) | −0.01 | (−0.02, −0.01) |
| Ireland*Age W3 (Centered) Squared | 0.01 | (0.01, 0.01) | 0.01 | (0.00, 0.01) |
| Italy*Age W3 (Centered) Squared | 0.01 | (0.01, 0.01) | −0.01 | (−0.01, −0.01) |
| Netherlands*Age W3 (Centered) Squared | −0.02 | (−0.02, −0.02) | −0.01 | (−0.01, −0.01) |
| Poland*Age W3 (Centered) Squared | −0.00 | (−0.01, −0.00) | −0.01 | (−0.01, −0.01) |
| Spain*Age W3 (Centered) Squared | −0.02 | (−0.02, −0.02) | −0.01 | (−0.01, −0.01) |
| Sweden*Age W3 (Centered) Squared | −0.01 | (−0.01, −0.00) | −0.02 | (−0.02, −0.01) |
| Switzerland*Age W3 (Centered) Squared | −0.01 | (−0.01, −0.00) | −0.02 | (−0.02, −0.02) |
| West Germany*Age W3 (Centered) Squared | 0.01 | (0.00, 0.01) | −0.01 | (−0.01, −0.01) |
| Constant | 13.97 | (13.33, 14.61) | 14.21 | (13.70, 14.73) |
Note. CI = confidence interval.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Disclosure of Potential and Actual Conflicts of Interest:
There is no declared conflict of interest.
REFERENCES
- 1.Kobayashi LC, Feldman JM. Employment trajectories in midlife and cognitive performance in later life: longitudinal study of older American men and women. J Epidemiol Community Health. 2019; 73(3):232–238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kröger E, Andel R, Lindsay J, Benounissa Z, Verreault R, Laurin D. Is complexity of work associated with risk of dementia? The Canadian Study of Health and Aging. American Journal of Epidemiology. 2008; 167(7):820–830. [DOI] [PubMed] [Google Scholar]
- 3.Mosca I, Wright RE. Effect of retirement on cognition: evidence from the Irish marriage bar. Demography. 2018; 55(4):1317–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Andel R, Kåreholt I, Parker MG, Thorslund M, Gatz M. Complexity of primary lifetime occupation and cognition in advanced old age. Journal of Aging and Health. 2007; 19(3):397–415. [DOI] [PubMed] [Google Scholar]
- 5.Smart EL, Gow AJ, Deary IJ. Occupational complexity and lifetime cognitive abilities. Neurology. 2014; 83(24):2285–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Stern Y. Cognitive reserve. Neuropsychologia. 2009; 47(10):2015–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Stern Y. Cognitive reserve in ageing and Alzheimer’s disease. The Lancet Neurology. 2012; 11(11):1006–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ice E, Ang S, Greenberg K, Burgard S. Women’s work-family histories and cognitive performance in later life.” American Journal of Epidemiology; 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Leist AK, Glymour MM, Mackenbach JP, van Lenthe FJ, Avendano M. Time away from work predicts later cognitive function: differences by activity during leave. Annals of Epidemiology. 2013; 23(8):455–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fryer D, Warr P. Unemployment and cognitive difficulties. British Journal of Clinical Psychology. 1984; 23(1):67–8. [DOI] [PubMed] [Google Scholar]
- 11.Celidoni M, Dal Bianco C, Weber G. Retirement and cognitive decline: A longitudinal analysis using SHARE data. Journal of Health Economics. 2017; 56:113–125. [DOI] [PubMed] [Google Scholar]
- 12.Clouston SA, Denier N. Mental retirement and health selection: Analyses from the US Health and Retirement Study. Social Science & Medicine. 2017; 178:78–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Finkel D, Andel R, Gatz M, Pedersen NL. The role of occupational complexity in trajectories of cognitive aging before and after retirement. Psychology and aging. 2009; 24(3):563. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kokko K, Pulkkinen L, Puustinen M. Selection into long-term unemployment and its psychological consequences. International Journal of Behavioral Development. 2000; 24(3):310–20. [Google Scholar]
- 15.Schmidt FL, Hunter J. General mental ability in the world of work: occupational attainment and job performance. Journal of personality and social psychology. 2004. January;86(1):162. [DOI] [PubMed] [Google Scholar]
- 16.Möhring K. Life course regimes in Europe: Individual employment histories in comparative and historical perspective. Journal of European Social Policy. 2016; 26(2):124–39. [Google Scholar]
- 17.Goos M, Manning A, Salomons A. Job polarization in Europe. American economic review. 2009; 99(2):58–63. [Google Scholar]
- 18.Hoftijzer M, Gortazar L. Skills and europe’s labor market: How technological change and other drivers of skill demand and supply are shaping Europe’s labor market. World Bank. 2018.
- 19.Möhring K. Employment histories and pension incomes in Europe: A multilevel analysis of the role of institutional factors. European Societies. 2015; 17(1):3–26. [Google Scholar]
- 20.Deutermann WV Jr, Brown SC. Voluntary part-time workers: a growing part of the labor force. Monthly Lab. Rev. 1978;101:3. [PubMed] [Google Scholar]
- 21.Goldin C, Mitchell J. The new life cycle of women’s employment: Disappearing humps, sagging middles, expanding tops. Journal of Economic Perspectives. 2017; 31(1):161–82. [Google Scholar]
- 22.Sundötrom M. Part-time work in Sweden: Trends and equality effects. Journal of Economic issues. 1991; 25(1):167–78. [Google Scholar]
- 23.Goldin C. The quiet revolution that transformed women’s employment, education, and family. American economic review. 2006; 96(2):1–21. [Google Scholar]
- 24.Kalbe E, Kessler J, Calabrese P, Smith R, Passmore AP, Brand MA, Bullock R. DemTect: a new, sensitive cognitive screening test to support the diagnosis of mild cognitive impairment and early dementia. International journal of geriatric psychiatry. 2004; 19(2):136–43. [DOI] [PubMed] [Google Scholar]
- 25.Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician. Journal of psychiatric research. 1975; 12(3):189–98. [DOI] [PubMed] [Google Scholar]
- 26.Ziegler U. Dementia in Germany: past trends and future developments. Südwestdeutscher Verlag für Hochschulschriften; 2011.
- 27.Brugiavini A, Cavapozzi D, Pasini G, Trevisan E. Working life histories from SHARELIFE: A retrospective panel; SHARE WP 2013:11. Available at: http://www.share-project.org/fileadmin/pdf_documentation/Working_Paper_Series/WP_Series_11_2013_Brugiavini_Cavapozzi_Pasini_Trevisan.pdf. Accessed March, 2015. [Google Scholar]
- 28.Fratiglioni L, Wang HX. Brain reserve hypothesis in dementia. Journal of Alzheimer’s disease. 2007; 12(1):11–22. [DOI] [PubMed] [Google Scholar]
- 29.Pool LR, Weuve J, Wilson RS, Bültmann U, Evans DA, De Leon CF. Occupational cognitive requirements and late-life cognitive aging. Neurology. 2016; 86(15):1386–92. [DOI] [PMC free article] [PubMed] [Google Scholar]