Lifestyle and health factors accounted for 84% of men's excess cancer rate in Shanghai but 26% in the UK

Men get more cancer than women. The pattern holds for most cancers that both sexes can develop, in most countries, and has done for decades. What has never been settled is why: whether men are built differently or simply live differently.

A study published in JAMA Oncology in September suggests the answer depends on where the question is asked. In a large group of adults in Shanghai, smoking, drinking and nine other measurable factors accounted for most of the difference between men and women. In an even larger group in the UK, the same factors accounted for about a quarter of it, and the rest had no explanation in the data.

The gap in cancer incidence between men and women

The paper begins with the basic fact. Cancer risk is higher in males than females globally. An analysis of cancer registries on five continents found rates higher in men than women at 32 of 35 sites, after setting aside cancers of the breast and reproductive organs.

Two kinds of explanation compete. One is behavior. Men in many societies smoke and drink more, and both cause cancer. The other is biology. A review in Nature Reviews Cancer lists the candidates, noting that differences in behavior add to underlying genetic predisposition and influences of sex steroid hormones.

The puzzle is that the gap survives where behavior is similar. The registry analysis found the excess in men consistent across rich and poor regions. Its authors judged that the sex disparity for 13 cancer sites was considered to be entirely unexplained by known risk factors, and titled their paper with the word enigmatic.

The new study asks the question the authors frame as whether this sex disparity reflects a biological inevitability or is potentially modifiable. Its method is to measure, person by person, how much of the gap the known risk factors can account for, in two very different places.

How the study divided up men’s excess cancer rate

The authors, Zhuo-Ying Li, Yong-Bing Xiang and colleagues, combined three long-running studies that follow volunteers over time. Two are in Shanghai, one of men and one of women. The third is the UK Biobank. All enrolled adults aged 40 to 74 years at baseline with no prior cancer. Baseline data were collected from 1997 to 2010, with follow-up through 2021.

In total there were 589,766 participants, 134,742 in Shanghai and 455,024 in the UK, and 70,300 of them developed a cancer of a kind that either sex can get. The count covers all such cancers, excluding breast and genital-organ cancers.

Eleven factors were examined. Three were behaviors: smoking, heavy drinking and physical inactivity. Four were dietary: low intake of vegetables, low intake of fruit, and high intake of red meat and of processed meat. Four were health conditions: gallstones, chronic liver disease, diabetes and obesity.

For each population the researchers first measured the gap itself, then estimated how much smaller it would be if nobody had the risk factor in question. The result is expressed as a share of the gap. The authors attach a warning to the numbers. The shares are estimates that hold under the assumptions of their statistical models, and do not represent definitive causal partitions. They also do not add up, because the factors overlap: a man who smokes and drinks is counted under both.

In Shanghai, smoking and alcohol accounted for most of it

	Shanghai cohorts	UK Biobank
Participants	134,742	455,024
Cancer rate in men compared with women	66% higher	49% higher
Extra cases per 100,000 men a year, compared with women	216	200
Accounted for by all 11 factors together	183 cases, or 84%	52 cases, or 26%
Largest single contributor	Smoking, 52%	Obesity, 8%
Second largest	Heavy drinking, 15%	Smoking, 6%

The shares for single factors overlap, so they can sum to more than the combined figure.

In Shanghai the gap was wide and most of it could be accounted for. The 11 factors together were responsible for about 183 of the 216 extra cases per 100,000 men each year, accounting for 84.44% of the male-female incidence rate difference.

Smoking towered over everything else. Cigarette smoking alone was associated with 51.75% of the sex gap. Heavy drinking came next at 15%, then chronic liver disease and low fruit intake at about 10% each. Obesity and diabetes barely registered. Metabolic factors, including obesity and diabetes, contributed minimally to the gap.

The reason is visible in who smokes. In the study’s own cohorts, the authors estimate that smoking lay behind 21.67% of all the shared cancers in Shanghai men and 57.61% of their lung cancers. For Shanghai women the figures were 0.79% and 4.04%. A national study of tobacco in China a decade ago found the same divide: two-thirds of the men smoked, and Chinese women now smoke far less than in previous generations. Of the women in that study born since 1960, 1% had ever smoked, against 10% of those born in the 1930s.

Alcohol follows the same pattern, with a biological twist. In East Asian populations, aldehyde dehydrogenase 2 (ALDH2) deficiency is common and limits acetaldehyde detoxification, amplifying cancer risk. In plainer terms, many people of East Asian descent break down alcohol’s toxic by-product slowly, which makes drinking more dangerous for them. Since it is mostly men who drink heavily, that extra danger falls mostly on men.

In the UK, most of the excess cancer rate was unexplained

The British gap was nearly as large in absolute terms, about 200 extra cases per 100,000 men each year. The same 11 factors were behind 52 of them, accounting for 26.21% of the sex disparity.

The ranking was different too. For all shared cancers combined, obesity was the leading contributor, at about 8% of the gap. Smoking accounted for 6% and diabetes for under 4%.

That does not mean smoking is harmless in Britain. The authors estimate that it lay behind 75.51% of lung cancers in British men and 66.63% in British women. The point is that it does so in both sexes. A cause that affects men and women almost equally cannot explain a difference between them. The authors put it this way: smoking and heavy alcohol drinking are strongly concentrated among males in the Chinese population but are more similar between the sexes in the UK population.

In the UK, then, roughly three-quarters of men’s excess had no explanation in the data. The authors describe the result as leaving much of the disparity unexplained by the measured factors.

What else could lie behind the gap between men and women

The British result matches earlier work in Western countries. A study of 294,100 older Americans, led by researchers at the US National Cancer Institute, tested a similar list of behaviors and measurements. It found that rates were lower in men than in women only for thyroid and gallbladder cancers, and that for the rest only a modest proportion of the male excess was explained by risk factors (ranging from 50% for lung cancer to 11% for esophageal adenocarcinoma). Its authors concluded that such male predominance is largely unexplained by risk factors, and pointed to biological differences between the sexes.

A separate analysis of the UK Biobank last year took a different route, adjusting for smoking, alcohol, body size and other risks and seeing what remained. Some differences vanished. But men remained at greater risk than women for cancers at eight sites, among them the esophagus, bladder, liver and kidney.

What the biology might be is an active field. One proposal concerns the X chromosome. Women have two and men one. Geneticists who examined more than 4,100 tumors began from the observation that there is a striking and unexplained male predominance across many cancer types. They found a small set of genes on the X chromosome that guard against tumors and that stay switched on in both of a woman’s copies. A man who loses his single copy has no spare. The researchers concluded that this explains a portion of the reduced cancer incidence in females as compared to males across a variety of tumor types.

The new study does not test any of this. It points to biology by elimination, and to other things it could not measure. The authors call for work on unmeasured exposures at work and in the environment alongside what they term intrinsic biological determinants.

Limits of comparing cancer in China and the UK

Habits were recorded once. By the authors’ account, risk factors were assessed only at baseline and were reported by participants. People who later quit smoking or gained weight were still counted as they began.

The Chinese men and women were not enrolled together. The women’s study began in 1997 and the men’s in 2002. As the paper puts it, sex is structurally confounded with cohort and calendar time in the Chinese analysis, confounded meaning that the effect of being male cannot be fully separated from the effect of joining a later study in a fast-changing country.

Important exposures are missing. The authors write: “we lacked data on occupational exposures, ambient air pollution, and specific infectious agents”. If men’s jobs expose them to more carcinogens, some of that effect may be hiding inside the smoking and drinking estimates, or inside the unexplained remainder.

Neither sample is the nation. Here the paper is direct: the Shanghai cohorts represent an urban Chinese population, whereas the UK Biobank is subject to a well-documented healthy-volunteer selection bias. Biobank volunteers are healthier than Britons in general, and the pattern in the wider population could differ.

Ages. Everyone was 40 to 74 at the start, so cancers in younger adults are not covered.

Attribution is not proof. Saying that smoking accounts for half the gap is an estimate from a model. It is not a demonstration that removing smoking would halve it.

For Shanghai the authors draw a practical conclusion, that the priority is tobacco and alcohol control among males. For the UK the message is that the known levers are not enough. Sex is one of many things that shape a person’s own risk of cancer, which is assessed by their doctor.

In 589,766 adults, eleven lifestyle and health factors accounted for 84% of men’s excess cancer rate over women in Shanghai, led by smoking, and for 26% in the UK, in an analysis that recorded habits only once and could not measure workplace, environmental or biological differences between the sexes.