Abstract
Background
Dementia is a major public health concern, especially in low- and middle-income countries. While structured exercise benefits cognition, occupational physical activity may not.
Objective
To investigate the association between lifetime occupational physical demand and cognitive impairment among older adults in Brazil.
Methods
Cross-sectional baseline data from the Brazilian Longitudinal Study of Aging (ELSI-Brazil) were analyzed (n = 9412). An algorithm was used to estimate the number of participants aged 60+ (n = 5432) with dementia and participants reported the physical demand of the job held for most of their life. Logistic regression models estimated the impact of physical demand on dementia status.
Results
After exclusions, 4924 participants were analyzed. Compared with ‘never worked’, intense physical effort was associated with higher odds of dementia (OR 2.52; p = 0.013). Other categories showed no associations. On sensitivity analyses using ‘intense physical effort’ as reference, lower demand categories were associated with less dementia, and this effect was only present among women.
Conclusion
Intense occupational physical demand was associated with greater odds of dementia. Given the cross-sectional design, reverse causation and residual confounding (such as job control) may explain these findings. Occupational physical activity differs from leisure-time exercise in psychosocial context, suggesting this association is unlikely to reflect physical effort alone and may instead be driven by unmeasured psychosocial job characteristics frequently linked with physical jobs. Therefore, solely reducing physical demands is insufficient without addressing job control and stress. Findings highlight the need for occupational health interventions addressing psychosocial risks in physically demanding jobs.
Introduction
Dementia is a major public health concern, ramping up faster than expected in low- and middle-income countries. 1 In Brazil, a middle-income country, a recent analysis has estimated a dementia prevalence of 5.8%, 2 while GBD estimates projected a 206% increase by 2050, resulting in 5.7 million people with the condition. 3 Conversely, the current scenario of Brazilian dementia care is far from reassuring.
While accessing proper diagnosis and care is essential to reduce caregiver burden and quality of life, prevention strategies are paramount from a societal perspective, and previous research has shown that eliminating twelve modifiable risk factors can reduce dementia prevalence in up to 54% in the poorest regions of Brazil. Among these, one of the most impactful for dementia development is physical inactivity. 4
Although it is clear that structured physical exercise is beneficial for cognition, 4 the effect of physical activities outside this setting is still an ongoing debate. Unlike voluntary exercise, occupational activity often involves repetitive movements, limited autonomy, and psychosocial stressors, potentially negating its health benefits. 5 Previous studies have reported multiple detrimental health effects of high levels occupational physical activity, including dementia and mortality risk.6–9 Meanwhile, some studies have reported a beneficial effect of occupational physical activity for cognition. 10
In occupational health and safety (OHS) frameworks, job strain models emphasize how psychosocial exposures—high demand and low control—affect long-term brain health. 11 Jobs involving intense physical demand are especially associated with Occupational Health and Safety (OHS) risk models, particularly the Job Demand-Control model, including factors such as low decision latitude, high work stress, and reduced cognitive stimulation, which have been shown to be linked to dementia risk.10,12,13 Moreover, in the work environment, more physically demanding jobs (“blue collar jobs”) are historically associated with less education in early life.
This study's significance derives from the necessity of a structured, societal effort to prevent dementia sprawling from early life, and including occupational interventions, especially in a diverse middle-income country such as Brazil, where treatment resources are scarce. Our main objective is to explore the association between different levels of physical effort in the work setting and dementia in late life in order to guide dementia prevention initiatives. We hypothesize that greater physical effort in the work setting may be beneficial for cognition, which contrasts with some previous findings in the OHS literature.
Methods
Sample
We analyzed cross-sectional baseline data from the Brazilian Longitudinal Study of Aging (ELSI-Brazil), a nationally representative study of individuals aged 50+ (n = 9412), which was collected throughout 2015–2016. Full study procedures can be found elsewhere. 14
Variables
We used an algorithm approach to estimate the number of people living with dementia within the cohort. First, a cognitive assessment including temporal orientation (day, month, year, and day of the week), semantic verbal fluency (number of animals within 1 min), a 10-word list test for immediate recall and late recall, prospective memory (delayed task on a cue), and 4 semantic memory questions (2 about common items, and 2 about political knowledge) was administered.
Subsequently, scores for each domain were created and standardized by subtracting the participants’ mean from the sample mean and dividing them by the sample standard deviation. The mean standardized z-score for all subdomains created a global cognitive score. A normative subsample was used to calculate the predicted scores for the entire sample. Standardized predicted scores were subtracted from each individual score to calculate residual scores, which were in turn divided by the standard deviation of the residuals from the normative subsample.
Finally, a resulting z-score of −1,5 or lower was defined as cognitive impairment. Participants who met this criterion and endorsed functional impairment were classified as people living with dementia. Full procedures for identifying dementia in the cohort are described elsewhere. 2
Work activities were classified by self-report using the question “How would you describe physical demand of the job you had most of your life?”. There were 5 answers: “never worked”, “mostly sitting” (e.g., office job), “mostly standing or walking” (e.g., security worker, hairdresser), “some physical effort” (carrying weights, such as nurse or carpenter), and “intense physical effort” (carrying heavy weights, such as miner or docker).
The following variables were used as control covariates in our multivariate analyses: age, sex, marital status (having or not having a partner), years of education, number of chronic conditions, sedentarism (defined as the absence of at least 10 min of moderate physical exercise per week), family income per capita and alcohol and tobacco consumption (any current use). The total number of chronic conditions was estimated from self-report of 14 conditions.
Analyses
Descriptive characteristics of the study sample were summarized using frequencies and percentages for categorical variables, and means with standard deviations for continuous variables. We compared categorical variables (sex, having a partner, alcohol use, tobacco use and sedentarism) using Pearson's chi square and continuous variables (age, school attainment, number of comorbidities and household income per capita) using t-test across dementia groups. These comparisons provided a descriptive assessment of between-group differences in variables that may influence the association between occupational exposures and dementia outcomes.
Multivariate logistic regression models were performed to estimate the association between physical demand at work and dementia status, adjusting for sex, age, formal education, having a partner, household income per capita, alcohol consumption, tobacco use, sedentarism and number of chronic conditions.
We also performed sensitivity analyses modifying the reference group, in order to account for non-linear effects and minimize selection bias that might arise when extreme categories contain disproportionate numbers of participants or differential risk profiles. Finally, to evaluate whether the association between occupational physical effort and dementia differed by sex, we conducted an interaction (effect-modification) analysis using logistic regression. We fitted a multivariable logistic regression model including the cross-product term between job stress and sex adjusting for the same covariates as the main model. The statistical significance of effect modification was assessed using a joint Wald test of all interaction terms.
All models were estimated using StataSE 17 and accounted for the complex sampling design using the ‘svy’ command to present accurate proportions according to the sampling weights. Statistical significance was defined a priori as a two-sided α level of 0.05 for all analyses.
Reporting
We used the STROBE reporting guideline 15 to draft this manuscript, and the STROBE reporting checklist 16 when editing, included in Supplement A.
Results
From the total sample (n = 9412), 3980 individuals were excluded due to an age younger than 60 years old, and 508 were excluded due to missing data. Data from 4924 respondents was included, and from these, 311 individuals (5.4%) were identified as having dementia. Regarding type of work, 100 participants never worked (2%), 745 reported a “mostly sitting” job (16%), 1565 reported a “mostly standing or walking” job (32%), 1535 reported a “some physical effort job” (30%) and 979 reported an “intense physical effort job” (20%). The mean age was 69.6 (69.1-70) and 2949 were women (55.6%).
Dementia was more common among older participants, with less education, women, without a partner, without alcohol use, with lower household income per capita and with a sedentary lifestyle (Table 1).
Characteristics of the total sample and according to dementia status.
SD = Standard deviation.
When compared to a “never worked” status, a “mostly sitting” (OR 2.21 [0.93-5.2], p = 0.071), a “mostly standing or walking” (OR 1.31 [0.57-2.99], p = 0.523) or a “some physical effort” (OR 1.76 [0.80-3.89], p = 0.161) type of work were not associated with the occurrence of dementia. Contrariwise, “intense physical effort” type of work was associated with a higher occurrence of dementia (OR 2.52 [1.21-5.24], p = 0.013) (Table 2).
Association of type of work with dementia status (n = 4924).
†Adjusted for sex, age, marital status, educational level, household income per capita, alcohol use, smoking, number of comorbidities and level of physical activity.
These findings may be subject to selection bias, as the “never worked” group may include, for example, people with mental disorders that are not able to work and also have a higher risk of dementia due to their condition. Therefore, we conducted a sensitivity analysis using the “intense physical effort” group as reference. In this analysis, compared to having an “intense physical effort” job, having a “mostly standing or walking” (OR 0.52 [0.34-0.78], p = 0.002), a “some physical effort” job (OR 0.7 [0.50-0.97], p = 0.034), or people who never worked (OR 0.40 [0.19-0.82], p = 0.013) were inversely associated with dementia. On the other hand, having a “mostly sitting” type of work did not differ from an “intense physical effort” job (OR 0.87 [0.51-1.49], p = 0.620).
We also conducted an interaction analyses for gender effects on the main predictor (occupational physical effort), which yielded significant results (p = 0.01). We then ran separate analyses for each group, and revealed a significant effect of physical effort on dementia status only for women. When compared with an intense physical effort, “moderate physical effort” (OR 0.43, p < 0.001), “mostly standing or walking” (OR 0.37, p < 0.001) and “never worked” (OR 0.3, p = 0.002) had less dementia, while “mostly sitting” was not associated with dementia. No associations of physical effort level with dementia were found for any category among men.
When using the physical effort variable as a continuous measure, each level of reduction (from intense effort to never worked) represented a reduction in the occurrence of dementia (OR 0.85 [0.75-0.97], p = 0.018).
Discussion
Our main hypothesis that occupational physical activity can benefit cognitive function was rejected. We found an association between the most intense level of physical effort in the work setting and a higher occurrence of dementia, which is independent of educational level and household income. It is hard to draw conclusions regarding causality, as this is a cross-sectional analysis. Nevertheless, physical demands at work do not appear to be beneficial, and may cause harm, despite consistent findings that point to the benefits of physical exercise in dementia prevention.
Previous studies have also revealed similar results. One retrospective case control study for Alzheimer's Disease Dementia including 357 participants showed that cases had higher occupational physical demands and lower mental demands than controls between the 3rd and 6th decades of life. 17 A recent article with data from the Health and Retirement Study (HRS), which partners Elsi-Brazil, focused on workers aged 55 years and older using a similar interview structure (rating physical demands from 1-4). In this study, cognitive performance in verbal episodic memory and reasoning was inversely associated with the level of physical demand at work. 9
It also appears that some variation found in the current literature may be associated to nuances of the type of physical effort exerted. An analysis from the Korean Longitudinal Study of Aging, which followed participants from 45 years of age, revealed that high physical strength demands were protective for cognition, while frequent heavy lifting or bending, kneeling or squatting increased the risk for cognitive impairment, but only among those with less formal education. 18
Mechanistically, this relationship may be explained by work related psychosocial stress and low job control accompanying these highly physically demanding jobs, consistent with the Job Demand–Control model.11,19 Other hypotheses point to the lack of mental demands, independence and decision-making usually associated with highly physical jobs. 9 Cognitive demands in the work setting have been identified as a protective factor for cognitive decline and dementia, which is consistent with the cognitive reserve model. 20 These complex interactions in the work setting are likely the main drivers of cognitive effects, overshadowing individual consequences of physical effort.
Two studies using data from a large Norwegian cohort of individuals aged 70 years or older evaluated the trajectories of occupational physical activities 21 and occupational cognitive demands 22 and their association with mild cognitive impairment and dementia risk. Individuals with a stable high occupational physical activity trajectory had a higher risk of dementia, while participants with a high Routine Task Intensity index (less cognitive demand) had a higher risk of mild cognitive impairment and dementia.
High physical strain combined with low cognitive stimulation can elevate allostatic load and chronic inflammation, pathways implicated in dementia, 23 a phenomenon depicted in previous studies that revealed that jobs with low job control (independent of intensity of demands) may be associated with more dementia 11 probably through increased levels of stress in these settings. 12
This highlights the importance of properly structured exercised associated with leisure activities in dementia prevention, as it appears these cannot be substituted for work associated physical activities. As shown in cohorts comparing leisure time physical activity and occupational physical acting as predictors for incident dementia, the latter is detrimental while the prior is protective.7,8 Occupational physical activity differs from leisure-time exercise in voluntariness, recovery, and psychosocial context, and even for individuals with high occupational physical activity, leisure time physical activity can be added with independent health benefits. 24
Furthermore, a study analyzing “job types” divided into blue-collar (industry with manual labor), white-collar (officials, business, managers) and pink-collar jobs (retail and service), found a worse cognitive performance for blue collar jobs when compared to white collar jobs, but only for women. These findings include mental and physical demands associated to each type. 13
In some cohorts, gender variations were observed. Liu et al. reported that low occupational status and blue collar jobs were associated with worse baseline cognitive scores and low occupational status was associated with faster cognitive decline in white women. 25 This pattern varies according to the population studied, with some cohorts pointing to an effect for both men and women, only for women or for neither. 26 In our study, we found and effect for the entire group and for women, but not for men individually.
The stronger association observed among women may reflect gendered occupational trajectories, labor market segregation, and differential exposure to physically demanding and low-control jobs. For instance, in Latin America, most jobs in the manufacturing sector (which are frequently high physical demand jobs) are dominated by males, which also hold most of the positions with better pay and quality of life. As such, women in these industries are pushed to the lower-pay (frequently lower-autonomy and higher-stress) positions. As a previous study revealed, even in the textile industry which employs more women than man, 49% of men but only 7% of women have “good jobs”, defined by better pay. 27
In our sensitivity analysis, we also found that intermediate levels of physical effort at work are associated with less dementia than high physical effort, but sitting jobs are not. This may lead to the conclusion that a total lack of physical activity or too much physical activity may be detrimental to cognitive function.
This study is especially limited by its cross-sectional design, which may limit the extent to which we can draw conclusions regarding the significance of our findings, as associations can be bidirectional. The findings should be interpreted cautiously and should not be used to infer causality. For instance, early cognitive decline may have influenced the participants occupational trajectories, leading to less cognitive demanding and more physically demanding roles. Conversely, it is much more likely that the type of work across the lifespan affects cognition in late life than the other way around. In the meanwhile, the relatively wide confidence intervals also indicate some statistical imprecision, which should be considered when interpreting the magnitude of this association. Furthermore, we have used a validated algorithm for dementia diagnosis, which is practical and comprehensive, but less accurate than a traditional clinical evaluation, allowing for misclassifications, which may bias our outcome. Further studies using gold-standard diagnostic procedures are required to draw definitive conclusions.
Our exposure assessment relied on a single self-reported item, lacking objective measures, limiting precision and conclusions draws, as its subject to recall bias and errors and constitutes a validity weakness. Furthermore, the type of work categories have multiple nuances and, although we controlled for income and education, they might still be, at some level, an indirect measure of socioeconomic status. We were also unable to control for workplace characteristics, such as grouping specific sectors of the industry, measures of job control, stress and workplace hazards, allowing for residual confounding. This absence of adjustment for psychosocial work factors seriously precludes causal inference. Future studies should integrate validated exposure indices (such as accelerometry) 28 and psychosocial metrics to disentangle physical and cognitive demands.
From an occupational health standpoint, findings highlight the need for workplace interventions that go beyond reducing physical demands, but rather aim to reduce psychosocial strain, enhance worker autonomy, and incorporate cognitive stimulation opportunities in physically demanding roles, with an especially attentive approach for women in the workplace.
From an OHS perspective, interventions should extend beyond reducing physical workload alone. Concrete strategies include psychosocial risk assessment using validated tools such as the Job Content Questionnaire 29 or the Copenhagen Psychosocial Questionnaire (COPSOQ), 30 which may be used to coordinate job redesign initiatives in order to increase autonomy, task variation, and integration of cognitively engaging activities within physically demanding roles.
In physically demanding jobs, frequent breaks should be offered and physical tasks should be alternated with mental tasks. Automation can also open space for more cognitively stimulating tasks such as operating machinery instead of direct manufacturing. Every worker should partake in the decision making process and be aware of their teams or sector goals in an unsegregated manner, while also developing a clear career path leading to less strenuous routines.
Conclusions
Although sedentarism has been shown to be associated with a worse cognitive performance and increased dementia risk, physically demanding jobs are not associated with better cognitive outcomes in late life. Meanwhile, work stress, low job control and lower cognitive demands associated with these activities might exert a negative effect on cognition independent of educational level and physical exercise unrelated to work. Despite this, reducing physical workload alone is unlikely to mitigate dementia risk if psychosocial job stressors persist.
Some data suggest mentally demanding jobs, on the contrary, can be beneficial. With development of new technologies, moving from physical work into mental work and improving work environments can potentially reduce dementia risk. OHS policies should prioritize integrated job design approaches that address both physical and psychosocial exposures, particularly in blue-collar and highly gender-segregated occupations.
Supplemental Material
sj-docx-1-wor-10.1177_10519815261427708 - Supplemental material for Physically demanding work and cognitive impairment in late life: Results from the Brazilian longitudinal study of aging
Supplemental material, sj-docx-1-wor-10.1177_10519815261427708 for Physically demanding work and cognitive impairment in late life: Results from the Brazilian longitudinal study of aging by Matheus Ghossain Barbosa, Wendell Lima Rabelo, Laiss Bertola, Maria Fernanda Lima-Costa and Cleusa Pinheiro Ferri in WORK
Footnotes
Acknowledgments
ELSI-Brazil was supported by the Brazilian Ministry of Health: DECIT/SCTIE – Department of Science and Technology from the Secretariat of Science, Technology and Strategic Inputs (Grants: 404965/2012-1 and TED 28/2017); COPID/DECIV/SAPS – Health Coordination of the Older Person in Primary Care, Department of Life Course from the Secretariat of Primary Health Care (Grants: 20836, 22566, 23700, 25560, 25552, and 27510).
MFLC and CPF are recipients of National Research Council (CNPq) research productivity fellowships. The study sponsors had no role in the design or management of the study, data collection and interpretation, analysis, the writing of the study or the decision to submit the study for publication, nor did they have ultimate authority over any of these activities.
Ethical Approval
The research ethics committee of the Fundação Oswaldo Cruz – Gerais approved the study (protocol numbers 34649814.3.0000.5091. Genotyping of the cohort population was approved by Brazil's national research ethics committee (Certificado de Apresentação para Apreciação Ética: 63725117.9.0000.5091). All research procedures have been conducted according to the principles stated in the Declaration of Helsinki.
Informed consent
Participants signed separate informed consent forms for the interviews, physical measurements, and the laboratory assays, authorized sample storages, and access to administrative records.
Funding
Brazilian Ministry of Health: DECIT/SCTIE (Grants: 404965/2012-1 and TED 28/2017); COPID/DECIV/SAPS (Grants: 20836, 22566, 23700, 25560, 25552, and 27510).
Ministério da Saúde, (grant number 04965/2012-1 and TED 28/2017, 20836, 22566, 23700, 25560, 25552, and 27510).
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
