Commentary

Oscar Wilde (amongst others) infamously said “everything in moderation, including moderation itself”. The response, in short, in relation to head and neck cancer risk from smoking and alcohol, is – “well, not quite!”.

Lubin et al. present the most comprehensive and in-depth analysis of the traditional risk factors for head and neck cancer – smoking and alcohol drinking. Uniquely, they investigate the differential risks for head and neck cancer (and for its subsites) associated with frequency (per day) and duration (per year) for these behaviours.

The methodology employed in this study is robust. This is the largest study of its kind, pooling data from 15 separate case-control studies from across Europe, North America, and South/Central America within the INHANCE (International Head And Neck Cancer Epidemiology) consortium.1 The smoking analysis contained around 2000 cases each of oral, pharyngeal and laryngeal cases; while the alcohol analysis had around 3000 cases of each. Comparisons were made with around 8000 controls.

The main findings reported were that fewer cigarettes smoked per day for many years gave a greater risk for head and neck cancer than many cigarettes smoked per day for fewer years. This was all the more interesting because the comparison between the two groups was for equal pack-years. (Where pack-years are defined as one pack of 20 cigarettes smoked each day for a year).

While for alcohol consumption, it was the other way round, heavy drinking per day for a few years gave a greater risk for head and neck cancer than fewer drinks per day for many years. This comparison was controlled for total alcohol exposure.

They also confirmed the relationship that smoking was more strongly associated with laryngeal cancers and alcohol consumption more strongly associated with pharyngeal and oral cavity cancers. However, this finding is reported in relative terms, the absolute risks are not clear from the data presented here, but one would expect the risk associated with oral and pharyngeal cancer to be greater from smoking than alcohol consumption – and this message may have got lost.

Furthermore, for both the smoking and alcohol analyses, there was no safe minimal limit at which either smoking or alcohol consumption was protective. Risk for head and neck cancer increased steadily through to 15 cigarettes per day; and through 10 alcoholic drinks per day.

Investigating the synergistic effect of magnifying the risks when both smoking and alcohol consumption, while beyond the scope of this analysis, could be interesting to throw into the mix in terms of the pattern analyses of these behaviours as examined here. And future analyses, assessing differences across the world and socioeconomic groups would also be interesting.

The findings cast further light on our understanding of the risk factors for head and neck cancer – it is not quite as simple an explanation as smoking and alcohol consumption per se. Both of these behaviours are highly complex. The findings also have subtle implications for policy and practice. While one would not wish these findings to be misrepresented as harm-reduction messages, better understanding these behaviours could potentially go some way towards developing behavioural risk assessment tools in the future.

And we all have to be careful with the old Scots saying “a little of what you fancy does you good”...

Declaration of interest: David Conway is a member of the INHANCE consortium, but was not involved in this study.