Bar Staff have Doubled Risk of Death from Alcohol Consumption

A new report from the Office of National Statistics
reveals that bar staff and publicans are approximately twice as likely
to die from alcohol consumption as the general population. Other people
working in catering, hospitality and entertainment also have an
increased risk, as do hairdressers and barbers. Amongst male workers,
seafarers, butchers and labourers (building, woodworking, flooring and
tiling) have a similarly increased risk. Men with the lowest rates of
death from alcohol are farmers, managers, and professional drivers,
whilst women who work with children, as well as managers have the lowest
risk of dying from alcohol consumption. Overall, those in full-time
employment have a lower risk of alcohol-related death.

Doctors have traditionally had high rates of heavy drinking and alcohol-related death, but according to the new report, they
are now amongst the professions with the lowest rates of
alcohol-related death. The authors of the report comment on two factors
that may be important here. The first is that doctors are more aware of
the health implications of heavy drinking, and so have reduced their
drinking before the rest of the population. A similar trend was
previously observed with smoking. A second factor is that many doctors
may have cultural reasons for not drinking, as a relatively high
proportion of doctors are Asian (21% of doctors, relative to 4% of the
general population). Professor Martin Plant (quoted on the BBC news
website) suggests a third possibility, that it is the increased number
of women in the profession that has had “a civilising effect.” 

The figures for one occupation have been widely
misreported; a number of newspapers and magazines have reported that
“female office juniors” are twice as likely to die from alcohol as the
rest of the population. In fact, it would be more accurate to say they
are half as likely to die. The reason for this dramatic error is that
the report gives two measures, the Population Mortality Ratio (PMR) and
the Standardized Mortality Ratio (SMR). Neither is perfect, which is why
both are reported. PMR is based on the proportion of deaths (before the
age of 64) within each occupation, so is affected by how many people
die of other things. SMR looks at the number of people in each
occupation who died, relative to the number that would be expected to
die if the alcohol-related mortality rate was the same for all
occupations.

The PMR indicates that, of those “elementary office
occupations” who’d died, the proportion who had died of alcohol related
problems was twice that found in the general population. However, this
figure is misleading, as we can see when we look at the actual numbers
who died. Based on the rate of alcohol-related deaths in the general
population, we would expect 36 deaths from alcohol amongst female office
staff. In fact, there were only 17, less than half as many as would be
expected (SMR). The reason for the high proportion of alcohol related
deaths (PMR) is that whilst secretaries are only half as likely to die
from alcohol related problems as other people, they are even less likely
to die from anything else.

The full report is published in Health Statistics Quarterly 35, Autumn 2007, pp 6-12.