Alcohol causes cancer, and here’s the evidence

Alcohol causes at least seven types of cancer

This week, the Department of Health is launching a new advertising campaign to raise awareness of how alcohol can affect your health in the long-term.

Many of us are aware of the short-term effects of drinking too much – feeling sick, a hangover, a spot of embarrassment, a vague but hard-to-pin-down sense of guilt – but the long-term effects often slip under the radar. These include a higher risk of many cancers, heart disease, stroke and more.

Cancer Research UK is supporting the new campaign so we wanted to use this blog post to discuss some of the science around alcohol and cancer.

So is alcohol really linked to cancer?

Yes. Surveys tell us that only about a third of people realise that alcohol can increase the risk of cancer but actually the evidence in this area has been very strong for a number of decades.

The International Agency for Research on Cancer produces reports that are widely seen as the gold standard for working out what causes cancer and what doesn’t. They first said that there is “sufficient evidence” that “alcoholic beverages are carcinogenic to humans” way back in 1988. Since then, many more studies have been published. There are too many to list completely here, but this is a good review for the scientifically minded among you. IARC, incidentally, confirmed their ruling in 2007, and again last year.

Which cancers are affected? And how many?

Alcohol causes at least seven types of cancer, including cancers of the mouth, oesophagus (food pipe), pharynx (upper throat), larynx (voice box), breast, bowel and liver. There is also increasing evidence linking alcohol to pancreatic cancer.

Estimating the number of cancers that are linked to alcohol is always going to be a rough business. One analysis by IARC estimated that in Western European countries like the UK, around 5 per cent of cancers are linked to alcohol, which works out to around 15,000 cases a year. In the Oxford Textbook of Medicine, Professors Richard Doll and Richard Peto estimated that 6 per cent of cancer deaths in the UK are caused by alcohol, adding up to around 9,000 a year.

What level of drinking affects the risk of cancer?

There is a clear linear relationship between the amount of alcohol someone drinks, and their cancer risk. In other words, the more people drink, the higher their risk.

But cutting back drinking so you don’t feel drunk doesn’t mean you avoid risks to your health. Alcohol can increase the risk of cancer at levels too low to make an average person drunk. Studies have consistently shown that as little as three units a day – the amount in a pint of strong lager or a large glass of wine – can significantly increase the risk of mouth, oesophageal, laryngeal, breast and bowel cancers. At this level, the risks are fairly small but they get bigger the more you drink.

However, it seems that alcohol only increases the risk of liver or pancreatic cancers if people drink large amounts. This is because alcohol affects the risk of these cancers by causing cirrhosis or pancreatitis, conditions that are linked to heavy drinking.

Which is worse: binge drinking or spreading my drinking across the week?

We don’t actually know, and this is one of the big holes in the current evidence. Imagine someone who generally drinks nothing during the week but then knocks them back at weekends. So far, there’s not been enough research to tell if they have a higher or lower risk of cancer compared to someone who drinks the same total amount, but spread throughout the week.

There are hardly any published studies looking at the effects of different drinking patterns. One study suggested that weekend drinking had particularly strong effects on the risk of breast cancer, but it’s not conclusive in itself. As we said above, it’s the total amount that matters.

How high are the risks?

There’s a good meta-analysis (an overview of existing studies) that compares the effects of different levels of drinking on different cancer types. By collecting the results of previous studies, this analysis concluded, for example, that drinking 6 units a day (around 2 pints of strong lager):

• increases the risk of mouth cancer by 3 times (200 per cent)
• increases the risk of oesophageal cancer by 2 times (100 per cent)
• increases the risk of breast cancer by 55 per cent
• increases the risk of bowel cancer by 10-19 per cent

These figures are “relative risks”. They show how a person’s odds of developing cancer change as they drink, but they don’t tell you what those odds were in the first place. Those are called “absolute risks”.

For example, for a woman, the lifetime risk of breast cancer (the odds of developing the disease at some point in one’s life) is 11 per cent or 1 in 9. If that goes up by 55 per cent, the new absolute risk becomes 17 per cent or 1 in 6.

For a man, the lifetime risk of oesophageal cancer is 1.3 per cent or 1 in 75. If that doubles, the new absolute risk is 2.6 per cent or 1 in 38.

It is worth noting that breast and bowel cancer are far more common than mouth or oesophageal cancers so the absolute risk of developing these cancers is much higher. Even if that risk goes up by a relatively small amount, that translates to a large number of actual cases. For example, the Million Women Study concluded that if 1,000 UK women under the age of 75 drank an extra unit a day, they would develop 15 extra cancers, 11 of which would be breast cancer.

How do these studies actually work?

There are two main types. “Case-control studies” compare people with cancer to healthy people to see if differences in their drinking habits are linked to their disease. “Cohort studies” are generally stronger. They follow large groups of healthy people, collect detailed information about their lifestyle, medical history and more, and see what happens to their health over the course of years or even decades.

In both cases, it is important to collect information on other aspects of a person’s lifestyle that could also affect their risk of cancer. For example, if you were studying the link between alcohol and mouth cancer, it’s important to account for whether people smoked or not, since smoking is a major cause of mouth cancer and smoking and drinking often go hand-in-hand. Likewise, being overweight also causes cancer, and drinking alcohol can make people put on weight. These are called “confounding factors” and scientists use statistical methods to adjust for them. In this way, they can consider the effects of alcohol alone.

Typically, scientists measure alcohol consumption with questionnaires that ask people to report how much they drink. These questionnaires have an obvious drawback in that they rely on people being honest rather than playing down how much they drink – and indeed this is a criticism frequently levelled at studies of alcohol consumption.

But actually, it turns out that alcohol drinking is measured very well by questionnaires, at least for those used by the best and biggest studies.  For example, the Million Women Study validated their questionnaire by comparing it to a 7-day food diary where participants write down everything they eat/drink on a daily basis for a week. They found a good level of agreement between the two measures. The EPIC study validated its questionnaire against actual urine and blood samples and found that alcohol was actually one of the parts of people’s diets most accurately measured by the questionnaires.

How does alcohol actually cause cancer?

There are probably many answers to this question because alcohol does a lot of things in our bodies. First and foremost, your body converts alcohol into a toxic chemical called acetaldehyde – responsible for many of the symptoms of a hangover. But acetaldehyde can also damage DNA – it sticks bulky molecules onto the famous double-helix and prevents our cells from repairing this damage.

Genetic studies support this idea. Some people in East Asian countries, like China and Japan, have genetic faults that either make them better at converting alcohol to acetaldehyde, or worse at getting rid of acetaldehyde. Either way, they build up unusually high levels of this chemical when they drink. And when they drink, they have a higher-than-usual risk of cancer.

As well as producing acetaldehyde, alcohol can also boost levels of oestrogen in the body, which could explain the link with breast cancer. And it increases the odds of developing cirrhosis, which, in turn, causes liver cancer. Finally, it can also make it easier for the tissues of the mouth or throat to absorb other cancer-causing chemicals, such as those found in cigarette smoke.

Do all types of alcohol affect the risk of cancer? Even wine? What about red wine?

You’ll see from the section above that, as far as cancer goes, the harmful effects of alcohol are common to all drinks, rather than any specific type. All alcoholic drinks, for example, produce acetaldehyde in the body.

There are some disagreements. Take wine, for example. It is difficult to untangle the possibility that wine consumption could simply be linked to generally healthier lifestyles. Some studies have found that wine increases the risk of cancer to a lesser degree than beer or spirits, others have said that it has the same effect, and yet others have concluded that it’s particularly harmful when it comes to cancers of the mouth or throat.  For example, the Million Women Study found that women who only drank wine have similarly higher risks of cancer than those who drank all types of alcohol.

Red wine contains a chemical called resveratrol, which has some anti-cancer effects in laboratory cells. Many studies are looking at resveratrol as a possible drug for treating or preventing cancer, but as we’ve discussed elsewhere on this blog, this is a far cry from saying that red wine could protect people from cancer. A purified form of a chemical is not the same as the food or drink that contains it, and work in laboratory cells doesn’t automatically translate to effects in living people.

Isn’t some alcohol good for you?

There is evidence that drinking small amounts of alcohol can reduce the risk of heart disease in certain age groups. However, heavy drinking increases the risk of heart disease. Interestingly, a recent review of the global effects of alcohol estimated that alcohol causes twice as many cases of heart disease as it prevents.

We have to weigh up the heart disease effect against the links between alcohol and cancer, high blood pressure, some types of stroke, cirrhosis, liver disease, pancreatic disease and more. One analysis of 34 studies found that people who drink less than a unit a day have around 17-18 per cent lower risks of “total mortality”, which means that at any given age, they are less likely to die of any cause. These benefits disappear at roughly the level of alcohol drinking that the Government guidelines are set at.

It’s also important to realise that the benefits of light-drinking only applies to older age groups. According to one study, if you look at overall mortality, there is no beneficial level of drinking for women under 55 or men under 35.

The balance between the risks of cancer, heart disease and other conditions is why we are not suggesting that anyone avoids alcohol altogether. Instead, Cancer Research UK’s advice is to limit one’s drinking to one small drink a day for women (which is about two units a day) and two small drinks a day for men (about three to four units a day).

However, the important point is that, as we said above, there is a linear relationship between the amount you drink and your risk of cancer. This means that whatever you drink already, cutting down by some amount will help to reduce your risk. And, of course, doing so is entirely down to individual choice.

– Ed