IPM Take
Cardiovascular prevention has become increasingly sophisticated at describing the individual while remaining comparatively crude about the world in which that individual works.
Clinicians ask about smoking, blood pressure, cholesterol, diabetes and family history, yet occupation is often left in the medical record as demographic background rather than treated as information that might help explain where risk accumulates and where prevention fails.
A new Chinese study involving more than 1.4 million working adults suggests that may be a missed opportunity.
Cardiovascular risk varied substantially across occupational groups, but so did something arguably more actionable: whether hypertension, diabetes and high cholesterol were actually being treated and controlled. The highest burden appeared in parts of the secondary sector, particularly electricity, heat, gas and water production and supply, while modern service workers generally had lower disease prevalence but unexpectedly poor treatment rates.
That mismatch is the policy signal.
Occupation may not need to enter a risk calculator as though a job title were a biological risk factor. Instead, it could function as a population-level map showing where screening, treatment and prevention resources need to be concentrated.
The study is cross-sectional, and employment sector can capture many things at once: geography, socioeconomic position, physical demands, work schedules, health behaviours and healthcare access. It cannot tell policymakers that changing occupation would reduce cardiovascular risk.
It can tell them something more practical.
Prevention does not fail evenly across the workforce.
Executive Summary
A nationwide cross-sectional study published in JACC analysed cardiovascular risk among 1,421,390 employees aged 20 to 65 years who underwent health examinations at 167 centres across 24 Chinese provinces between 2021 and 2023.
Participants had a mean age of 33.8 years, and 53.4% were men. Researchers classified employment into 18 occupational groups spanning primary, secondary and tertiary economic sectors and calculated age- and sex-standardised prevalence of coronary heart disease and major cardiometabolic risk factors.
Overall coronary heart disease prevalence was 1.33%, with a pronounced geographic gradient in which disease and cardiometabolic risk were generally highest in northern China and lowest in the south.
Occupational differences were also substantial. Modern service sectors had the lowest overall burden, particularly information transmission, software and information technology, while the highest prevalence was concentrated in parts of the secondary sector, especially electricity, heat, gas and water production and supply.
Yet lower disease burden did not automatically mean better prevention.
Across the full cohort, only 29.5% of people with hypertension were receiving treatment and 20.8% had controlled blood pressure. Among people with diabetes, treatment and control rates were 36.6% and 21.9%, respectively. Hypercholesterolaemia treatment stood at 17.0%, while only 9.5% achieved LDL cholesterol control.
Modern service workers, despite their relatively favourable disease profile, had some of the lowest treatment rates for hypertension, diabetes and hypercholesterolaemia.
Among adults aged at least 35 without established atherosclerotic cardiovascular disease, 23.3% were classified as having moderate-to-high 10-year risk using the China-PAR model. That proportion ranged from 30.4% among agriculture, forestry, animal husbandry and fishery workers to 14.7% among workers servicing advanced manufacturing.
Modelling suggested that bringing elevated blood pressure and cholesterol to target and eliminating current smoking could produce different reductions in predicted cardiovascular risk across occupations.
Those simulations are not evidence that such interventions have already reduced events. They illustrate how much potentially modifiable risk is distributed differently across occupational groups.
Why it matters
- HTA bodies: The study highlights a dimension of prevention that conventional technology assessment often overlooks: where an intervention is deployed may influence its value almost as much as what the intervention is. Workplace blood-pressure screening, lipid detection and targeted prevention programmes may produce different benefits across occupational settings, but implementation studies are needed to establish effectiveness and cost-effectiveness.
- Payers: Population-wide cardiovascular prevention can be expensive when delivered indiscriminately. Occupational segmentation could help identify groups with high underlying risk or particularly poor treatment and control, supporting more targeted workplace screening, referral and follow-up while avoiding the assumption that employees in apparently healthier sectors need little preventive attention.
- Industry / innovation partners: Digital screening, workplace health platforms and remote risk-management tools could use occupational context to improve targeting, but job category should not become a crude proxy for individual disease risk. The opportunity lies in combining employment setting with validated clinical risk factors, access information and longitudinal outcomes rather than building another simplistic score.
Cardiovascular risk is usually described as something a person carries: high blood pressure, cholesterol, diabetes, smoking history, excess weight or a family predisposition that follows them from home to clinic.
The new Chinese data suggest that risk may also leave a pattern across workplaces.
Researchers analysed more than 1.4 million employed adults who underwent health examinations across 167 centres in 24 provinces, creating an unusually large picture of how cardiovascular disease and its major risk factors are distributed across occupations in a rapidly changing economy.
The first striking feature was geographic rather than occupational. Coronary heart disease prevalence was highest in northern China and lowest in the south, with hypertension, obesity, diabetes and lipid abnormalities showing broadly similar regional patterns.
That matters because occupation and geography are rarely independent. Industries concentrate in particular regions, working conditions vary with local economic structures, and access to healthcare can differ considerably across provinces.
The researchers therefore standardised occupational estimates for age and sex, but the broader socioeconomic and geographic context remains part of the story.
Within that landscape, however, employment patterns were difficult to ignore.
Workers in modern service industries generally had the lowest prevalence of coronary heart disease and cardiometabolic risk factors, with particularly favourable profiles among people employed in information transmission, software and information technology.
At the opposite end, the highest burden clustered in the secondary sector, particularly among workers involved in the production and supply of electricity, heat power, gas and water.
It would be easy to turn that result into a story about “dangerous jobs”.
The data do not support that conclusion.
The study was cross-sectional, which means exposure and health status were measured at essentially the same point. It cannot determine whether a particular occupation caused hypertension, dyslipidaemia or coronary disease.
Occupational categories were also assigned according to employer registration and industry affiliation rather than detailed measurements of each worker’s physical activity, stress, shift patterns, pollution exposure or socioeconomic circumstances.
A job title in this study is therefore better understood as a marker for a bundle of exposures and social conditions than as an isolated cardiovascular cause.
That may actually make occupation more useful for public policy.
Public-health systems do not always need a variable to be causal before it helps identify where resources are needed.
A postcode can reveal geographic inequalities without causing disease. A hospital’s catchment area can identify unmet need without producing that need. Occupation may function similarly by showing where cardiovascular burden and prevention gaps cluster across the working population.
The treatment data make that argument much harder to dismiss.
Less than one third of workers with hypertension were being treated, and only around one in five had achieved control. Diabetes treatment reached little more than one third of affected workers, while control remained similarly poor.
The cholesterol figures were more striking still.
Only 17% of workers with hypercholesterolaemia were receiving treatment, and fewer than one in ten met the study’s LDL cholesterol control criteria.
Those numbers show that the prevention problem is not simply finding people at high risk.
It is finding them and then successfully connecting them to sustained care.
The occupational pattern complicates the story further because the sectors with the healthiest overall profiles were not necessarily the sectors with the best management.
Modern service workers had comparatively low prevalence of many cardiovascular risk factors, yet they also showed some of the lowest treatment rates for hypertension, diabetes and hypercholesterolaemia, along with poor control of blood pressure and LDL cholesterol.
That is precisely why crude assumptions based on workforce appearance can be dangerous.
A younger, more educated, office-based workforce may look healthier at population level and still contain substantial untreated risk because employees perceive themselves as low risk, spend less time in regular healthcare, or encounter other barriers to long-term treatment.
The study cannot tell us which explanation is correct.
It does show that low disease prevalence and high-quality prevention are not the same thing.
That distinction becomes even more important when the researchers move from diagnosed disease to predicted future risk.
Among workers aged at least 35 who had no previous atherosclerotic cardiovascular disease, nearly one quarter were classified as having moderate-to-high 10-year risk under the China-PAR model.
But the average conceals considerable occupational variation.
The proportion reached just over 30% among workers in agriculture, forestry, animal husbandry and fishery, compared with fewer than 15% among employees providing services to advanced manufacturing.
Those differences could matter greatly when public-health authorities decide where to prioritise prevention.
The conventional model would invite everyone to a broadly similar screening programme and then wait for individual test results to identify who needs intervention.
An occupation-informed model could work in the opposite direction.
Health authorities could identify sectors in which hypertension, smoking or dyslipidaemia are particularly concentrated and deploy screening, follow-up and education more intensively in those workplaces.
That does not require treating occupation as destiny.
It means using population data to put preventive infrastructure where the probability of finding untreated risk is highest.
The researchers explored that possibility through simulation.
They modelled what would happen to predicted cardiovascular risk if elevated blood pressure and cholesterol were brought to target levels and current smokers were instead classified as non-smokers.
Potential reductions in predicted risk varied meaningfully across occupational groups.
Those findings should not be confused with intervention results. Nobody in the simulation was actually treated, nobody stopped smoking as part of an experiment, and no cardiovascular events were prevented.
But the exercise demonstrates something important for policy.
Different workforces have different prevention opportunities.
In one sector, blood-pressure control may offer the largest theoretical gain. In another, smoking reduction or lipid treatment may matter more. A single national cardiovascular prevention campaign can therefore be evidence-based while still being badly targeted.
This is particularly relevant in a country as large and economically heterogeneous as China.
Rapid industrialisation, urbanisation and expansion of service industries have transformed the structure of work over the past several decades. Cardiovascular prevention strategies developed for one economic landscape may not map neatly onto another.
The findings also reinforce a larger international debate about occupational health.
Work can influence cardiovascular risk through multiple pathways, including physical demands, sedentary time, shift work, psychosocial stress, working hours, socioeconomic position and exposure to heat, noise or pollutants.
Some of those relationships are well established, while others remain difficult to separate from the socioeconomic circumstances surrounding particular jobs.
Yet occupational health programmes have historically focused heavily on preventing injuries and exposures directly attributable to work.
Chronic cardiovascular disease sits awkwardly between occupational medicine and general healthcare because it is rarely caused by one identifiable workplace exposure.
That institutional division may need to soften.
If millions of adults spend most of their waking hours at work, the workplace is not simply a possible source of cardiovascular risk. It is also an unusually powerful place to deliver prevention.
Blood pressure can be measured there. Smoking support can be offered there. Employees with high glucose or cholesterol can be connected to primary care there. Digital follow-up can continue beyond an annual health examination.
The central challenge is ensuring that workplace screening does not become a substitute for treatment.
China already has extensive health-examination infrastructure, and the present study itself was possible because so many workers underwent examinations. Yet the low treatment and control rates demonstrate the limits of detection without continuity of care.
Finding hypertension is useful only if somebody owns what happens next.
That is the point made in the accompanying JACC editorial by Yuan Lu: surveillance has to become implementation.
This distinction should resonate beyond China.
Many health systems have accumulated enormous quantities of cardiovascular risk data without producing proportionate success in risk-factor control. Electronic records know who has hypertension. Pharmacies know who stops filling medication. Employers know where large populations of working-age adults spend their days.
The failure is often not lack of information.
It is failure to connect information with an intervention pathway.
Occupation could provide one of those connections because it turns an abstract population into a reachable group.
But using employment data in this way also requires political caution.
Occupational risk profiling should not become a mechanism for employers or insurers to discriminate against workers in industries associated with higher cardiovascular burden. Nor should a high-risk occupation allow policymakers to imply that disease is primarily an individual worker’s responsibility when working conditions, healthcare access and socioeconomic inequalities may contribute substantially.
The purpose of occupational segmentation should be to move resources toward need, not risk away from institutions and onto workers.
The study’s limitations reinforce that point.
Participants were employed adults who attended organised health examinations, which may not represent informal workers, unemployed adults or populations with less access to occupational health services. The cohort was also relatively young, with an average age below 34, so the observed prevalence of established coronary disease was low.
The analysis measured occupational categories at one point rather than reconstructing lifetime employment histories, and it could not capture the many individual exposures hidden inside broad industry labels.
The findings therefore do not justify adding occupation to cardiovascular risk calculators tomorrow.
They make a different case.
Risk calculators answer which individual is more likely to develop disease. Public-health policy has to answer another question: where are those individuals concentrated, and how do we reach them?
A dataset of 1.4 million workers suggests that employment may help answer the second question.
That is particularly important when the largest gaps lie not only in risk, but in treatment.
A workforce can have relatively low average disease prevalence and still be poorly served by prevention. Another can carry high predicted risk but require a very different mix of blood-pressure, lipid and smoking interventions.
Once those differences are visible, a universal prevention strategy starts to look less equitable rather than more.
Cardiovascular risk may still belong to the individual, but prevention policy has to operate at population scale. Occupation could help tell health systems where to look, where treatment is failing and where the next preventive euro, yuan or dollar is likely to matter most.
