There is a health gap between poor and well off workers during their working lives and this continues as they eventually down tools.
Workers at the lowest rung of the career ladder suffer from more ill health and a shorter life and this is magnified as they age.
It had been thought in retirement this gap would shrink.
But a new study of British civil servants found those in the lowest grades still suffered from stress after retirement which their higher grades colleagues do not.
Professor of Medical Sociology Tarani Chandola at the University of Manchester said: “The socio-economic gradient in health is well known, although the magnitude of the gradient varies over the lifecourse.
“This socio-economic-health gradient peaks around retirement in the US and a number of European countries with some evidence of widening even after retirement in the UK.
“This widening in health inequalities could be a reflection of the accumulation of socio-economic disadvantages over the lifecourse with early life inequalities in health becoming magnified over the life cycle.
“Retirement could potentially moderate this pattern of widening health inequalities with age if the health effects of retirement differ between socioeconomic groups.
“Involuntary retirement, which is more common among disadvantaged workers, is associated with poorer health, however, evidence for the effect of overall retirement on health is ambiguous.”
The lead author added past studies on retirement and health suggested retirement had a beneficial effect on mental health but no overall effect on perceived general health and physical health.
Other studies suggested the health benefits mainly occurs among employees working in low occupational grades and poor quality work who report improved health upon retirement, maybe as a consequence of a reduction in work related stressors.
But this appeared to contradict other observed patterns of widening social inequalities in health in early old age and retirement.
Cortisol is a stress hormone that follows a diurnal profile, peaking around 30 minutes after awakening, and returning to very low levels by bedtime.
Stressors disrupt the cortisol profile resulting in a higher level with a flatter slope from waking until bedtime.
These shallower slopes are a key biomarker associated with higher levels of stress.
They are also linked to cardiovascular mortality - a one standard deviation increase in cortisol at bedtime was associated with a doubling of the relative risk of cardiovascular mortality within six to eight years.
The study investigated whether workers who had recently retired had lower biological stress levels as indicated by steeper, so more advantageous, diurnal cortisol slopes compared to those still working in later life.
It followed 1,143 civil servants taking part in the London-based Whitehall II civil servants study.
They had an average age of 60 and stress hormones were measured from five samples collected across the day while employment grade was used to categorise people into high, middle or low grades.
Retirement was associated with lower stress levels with those who had recently retired had steeper diurnal slopes compared to those who remained in work.
But this apparent benefit of retirement on lowering biological stress response levels was only confined to those in high status jobs.
Workers in the lowest status jobs had flatter diurnal cortisol slopes compared to those in the top jobs.
And retirement increased, rather than decreased these differences in biological stress levels.
British civil servants in the lowest status jobs had the highest levels of stress as indicated by flatter and more adverse diurnal cortisol slopes compared to those in the highest status jobs.
Prof Chandola concluded: “Socio-economic differences in a biomarker associated with stress increase, rather than decrease, around the retirement period.
“These biological differences associated with transitions into retirement for different occupational groups may partly explain the pattern of widening social inequalities in health in early old age.
“It may seem counter-intuitive that stopping low status work which may be stressful does not reduce biological levels of stress.
“This may be because workers who retire from low status jobs often face financial and other pressures in retirement.
“This study suggests that people’s stress levels are not just determined by immediate circumstances, but by long run factors over the course of their lives.”
The study was published in the Journal of Gerontology.