Oh, the lure of a glass of chilled white wine at the end of a hard day. Or a goblet of luscious red by a leaping fire on a winter's evening. Or a gin and tonic, poured over cracking ice and lemon. Can't you hear it calling as you tramp home, tired, head buzzing with the day? Well, maybe not any more - if you heeded the recent study which warned that even a small glass of wine a day increases a woman's risk of breast cancer.
In all likelihood, most women just shrugged and reached for the corkscrew. There are, after all, so many conflicting stories about what is good and bad for you these days. Unfortunately, this is not some easily dismissed, pie-in-the-sky trial involving a couple of hundred people.
It is the Million Women Study, run by some very senior scientists at Oxford University. In research, size really does matter - and this is the biggest project of its kind on the planet.
Well over a million women - 1.3 million to be accurate - were recruited across the UK, some as far back as 1996 and the last in 2001. Most were in their 50s, invited with the help of the NHS breast-screening service the first time they went for a mammogram. The information they have shared about their lifestyle, history, health and habits, together with the data that is being gathered over the years about the diseases they get, is a goldmine for researchers.
The Million Women Study began as an attempt to get to the truth about hormone replacement therapy (HRT): would hormones rejuvenate women who were hitting the menopause or give them cancer? But now it has become a painstaking effort to write a blueprint for women's health.
The survey was started by Professor Valerie Beral, head of Oxford University's cancer epidemiology unit. When she began planning it in 1993, she quickly realised the study would have to be massive to answer the thorny and still controversial questions over HRT and possible links to cancers and other diseases. But enrolling a million women meant they would be able to tackle a lot of other issues too, and the answers are slowly coming in.
And yes, she says, it is intended to provide a definitive blueprint for women's health, spotlighting all the issues from the pill, to alcohol, to diet, to childbearing, to the menopause.
"That's what we plan to do, slowly and reliably over time," says Beral. "We're interested in not creating false stories, so it is totally reliable information - the sort of information that women want to know about their health."
This is work very much in progress, but parts of the picture have emerged. In broad terms, the implications for women who want to minimise their cancer (and heart disease) risks are as follows:
• Avoid HRT. Women run twice their normal risk of breast cancer while taking the combined form of oestrogen and progestagen (though the extra risk goes away once they stop). Oestrogen alone is not so risky. Both forms of HRT increase the risk of ovarian cancer, although not by as much as they do breast cancer. Oestrogen alone increases the risk of endometrial cancer, which affects the lining of the womb.
• Do not smoke.
• Don't let yourself get overweight. Interestingly, you shouldn't be too thin either. A body mass index (BMI, your weight in kilograms divided by the square of your height in metres) of around 24 is good.
• Ideally, don't drink at all.
• Don't worry about taking the pill when you are young. It protects against ovarian cancer.
• Take some exercise. Doing exercise helps you to avoid heart disease, but it doesn't have to be too strenuous.
• Breastfeed for as long as possible. Doing what the breasts were intended for protects against breast cancer. In an ideal, risk-avoiding world, women would also start having babies young and have many more of them, both of which protect against breast cancer.
Those are the headlines so far, but there is much, much more to come. The million women enrolled in the study have filled in three detailed questionnaires about their lives, their habits and their health, and the Oxford team can track the cancers to which some will succumb through the NHS central registries.
In the last two years, the researchers have also been able to get information on any hospital admissions. Education, income and social class all have a bearing on the diseases we get and how long we live, but a study like this can make allowances for all these factors to get a clear and unbiased picture of the effects of particular aspects of one's lifestyle.
The women have been asked everything from whether they cook with olive oil or butter, to how often they eat biscuits and drink tea, to how much they weighed when they were born, whether they were bigger or smaller than average at the age of 10, and what size clothes they wore at the age of 20. They have been asked how often they use a mobile phone, whether they belong to a religious or dancing group, how many hours they sleep and how often they feel happy.
With this much data, the scientists will be able to investigate a whole range of issues. Maybe those who enjoy painting or music in their spare time or who go to church are less stressed and happier than those with little social activity to report. And are those who say they eat five portions of fruit and vegetables a day really healthier?
The researchers have also focused on levels of obesity among women. Beral is keen to know whether those who were overweight as babies are doomed to a disease-ridden life. "There is so much talk now about how much worse it is to be obese as a child. We need to look at changes in body size over time and see how relevant it is," she says. "My guess is that it is more what you are now than what you have been in the past."
And this is probably the most upbeat message of the entire study: the effects of our bad habits don't seem to last. Nowhere - so far - has the team found lifestyles, habits or behaviours among the million women in their wayward youth that have caused cancers, heart disease or other illnesses later on.
Once you stop drinking, your increased cancer risk starts to melt away. It's the same with HRT. And even with smoking. "If you give up smoking, within five years your risk of lung cancer has halved," Beral says. "It's about what you are now. You can change it. The only exposure that I'm sure does persist is radiation. It mutates the DNA directly. But most of what we see seems to
be reversible."
So you may be what you eat - the work on diet is not complete - but you are not necessarily what you have eaten. In a world suddenly made that little bit gloomier by the absence of a relaxing daily drink, it's a cheery message.
The Million Women Study was only a glint in Beral's eye until computer scanning technology made it possible to process data from so many questionnaires. This meant that single-digit numbers - the number of alcoholic drinks consumed each day, for instance - could be scanned automatically, "instead of having somebody copy everything out".
The survey began on a shoestring, Beral says, and still has a tight budget for its scale, but is now funded by Cancer Research UK and the Medical Research Council (MRC) - probably the most respectable scientific backing possible in this country. The Oxford University unit also has a global reputation.
Epidemiological studies such as this one, investigating how often diseases occur in a specific population and why, have to be carried out on a large scale, but this is bigger than anything tried before. Dr Lesley Walker of Cancer Research UK gets irritated by people who suggest it is "just another study", comparable with all those investigations of red meat or vegetables in the diet of only 300 women.
"It is not just one study. It is huge - and huge equals power," Walker says. In a study of this size, it is possible to find effects that are too small to be seen in a population of a few thousand, but that are still significant.
Even so, Professor Sheila Bird of the MRC's biostatistics unit and the Royal Statistical Society, sounds a cautious note. Big in itself does not necessarily mean free of bias. One in four of the population in the age group signed up - but, she asks, "How different are the quarter who volunteered from the rest?" First they accepted breast screening and then they agreed to be part of the Million Women Study. Does that mean, for instance, that they are middle-class women more concerned about breast cancer?
Avoiding and then adjusting for any such bias, however, is at the heart of the work of the Oxford cancer epidemiology unit. And Walker points out that the Million Women Study is under constant scrutiny from the best scientists in this field - facing rigorous peer review, first to get funding from Cancer Research UK and the MRC, and then to get each scientific paper published as they are completed. The findings have all appeared in leading medical journals, such as the Lancet.
So what do the study's findings really mean for each of us as individuals? Should we all be following Beral's developing blueprint to the letter? The answer is, not necessarily. We may all run an increased risk of cancer if we drink, but how serious that is depends on how high or low a risk we had to start with.
Walker puts it well: "What they are coming out with is risks across the population that they are studying. Each of us as individuals will have been born with a greater or lesser risk of a cancer because of the genes we have inherited and, on top of that, our lifestyle has to be factored in.
"We may have only one vice and it may be drinking three glasses a night, but if we had 10 children before the age of 30, it is not going to make much difference. One needs the whole picture."
And each of us may have a different perspective on acceptable risk. As a herd, we have a 9.5% chance of getting breast cancer before we are 75. Drinking every day raises that risk to 10.6%. If we think we have no other major risk factors lurking, such as a mother or sister who died of the disease, but we believe a drink a day significantly improves our life, we might choose to go with the extra risk. The great thing about the Million Women Study is that it is giving us the information with which to make an informed choice.
Almost all of the criticism of the study has been from the HRT lobby - usually gynaecologists and GPs who prescribe the drugs. They have argued long and loudly that the findings are based on misinterpretation of the data and that the risks of cancer are outweighed by the benefits of hormones in alleviating menopausal symptoms and reducing bone fractures - an acknowledged benefit of HRT.
Some mudslingers in this lobby have even tried to write off the entire study as unreliable. But Dr Tim Hillard, a consultant obstetrician and gynaecologist in Poole and the chairman of the British Menopause Society, which has been unflaggingly critical over HRT, accepts its general scientific validity - although he does say that such studies show you trends, rather than proving anything.
"They have a huge amount of data," he says. "You can't dispute the results, but it is the interpretation that can be the problem.
"Epidemiology is a very interesting science which helps you understand what is going on, but it doesn't prove it. It is an association. If you drink regularly, you appear to have a higher risk. It doesn't mean you will get cancer."
And that's true. The Million Women Study can't tell you whether you will get cancer. It can nail down how many women will die because they are obese or they drink every night, but it can't tell you who they will be.
That's where Beral's blueprint comes in. Once we know what the risks are, we can choose how many of them we want to take. It's up to us. One thing we can be sure of in a world full of confusing headlines: a million women can't be completely wrong.
