Latest / Common Sense Health / False Hope and Fuzzy Maths
Transcript
- Stephen Thomas BSc. Hons: Hello. Today we're diving into a headline grabbing statistic that's been making the rounds. It's the idea that consuming soy can slash the risk of breast cancer mortality by a whopping 30%. Sounds fantastic, doesn't it? Like a magic bean that not only tastes good but also fights cancer. But... A major player in this narrative is a study called the Shanghai Breast Cancer Study, a rather hefty piece of research that followed thousands of women. The initial takeaway often presented is that women who munched on more soy had a significantly reduced chance of dying from breast cancer, or even seeing it return for those that had it once before. For anyone touched by this dreadful disease, that 30 % figure can spark a real flicker of hope. It's presented as a powerful endorsement for including soy in your diet. But, as with all things that sound almost too good to be true, a bit of that good old joined up thinking and a dash of common sense are needed. So let's peel back the layers of this particular onion, shall we? So sit back and take some notes. That glorious 30 % reduction is what we call a relative risk reduction, RRR. Now, this isn't to say it's meaningless, but it's like saying a small shop is having a 50 % off all item sale, well if an item was originally 10 pence, you're only saving 5 pence. But if it was 100 pounds, you're saving 50. The percentage is the same, but the actual tangible benefit in pounds and pence, or in this case, lives saved, is vastly different. The problem is, without knowing the original price, or in our medical analogy, the baseline risk, that 30 % RRR doesn't tell us the absolute risk reduction, ARR. Imagine the baseline risk of recurrence for these women was, say, 50%. A 30 % relative reduction would mean the new risk is 35%, so an absolute 15 % reduction. Well, that's a pretty chunky saving. But what if the baseline risk was already rather low, perhaps 5%, and 30 % relative reduction only translate to an absolute reduction of 1.5 % from 5 to 3.5%. Suddenly that headline-grabbing 30 % isn't quite as earth-shattering in the real world. So it's still a benefit, mind you, but perhaps not the miracle many might infer. Now you might be thinking, well, just tell me the baseline risk. And here's where it gets a bit trickier, especially in a population like China, because the Shanghai breast cancer study is an observational study. This just simply means researchers observed what people were already doing rather than telling them what to do. They didn't have a control group of women who ate no soy at all and then compare them to women who ate some soy wasn't like that at all. Think of it this way. ⁓ Trying to find a non-soy baseline in China is like trying to find a no-tea baseline in Britain. It's practically impossible. Because soy has been a dietary staple in many Asian cultures for centuries, woven into their fabric actually, or ⁓ part of their cuisine and their lifestyle. So what the researchers did was compare women who ate the most soy to women who ate the least soy. Within that same soy-consuming population, the least soy group still likely consumed some amount. wasn't a true zero. Therefore, the 30 % reduction is really saying if you're already eating soy, eating more of it within the spectrum of the typical Chinese diet seems to offer an additional benefit compared to eating less of it. So it's not comparing soy eaters to non-soy eaters because that latter group hardly exists naturally in that context there in China. So this makes it difficult to ascertain the true meaningful impact for someone who might currently consume very little or no soy and is maybe contemplating adding it to their diet. Another crucial aspect of observational studies, especially in nutrition, is something called healthy user bias. Now this is a bit like joining a gym. So when you hear about people who go to the gym regularly being healthier, thinner and living longer, you might think well the gym is the magic bullet. But have you ever considered this? Who joins a gym? Typically it's people who are already more health conscious who might also eat better, sleep more, smoke less and generally lead a more active lifestyle? Is it just the gym membership that's making them healthier or is it the entire healthy user profile? Now similarly in the Shanghai study women who consumed more soy might also have had healthier overall diets. Perhaps they were eating more meat or less processed food or they might have been more attuned to health advice. These other healthy behaviors could be the thing contributing to the improved outcomes So it makes it harder to isolate the exact sole effect of soy the researchers try to adjust for these confounding factors statistically, but it's incredibly difficult to account for every single one of them, so This distinction between relative and absolute risk isn't just some sort of dry academic exercise for me. It is actually a deeply painfully personal memory actually because years ago I lost my girlfriend to breast cancer. She was just 23 years old and I met her under some pretty difficult circumstances. She was the nurse working on the intensive care ward where my mother was being treated. And she was everything you'd hope a nurse would be. You know, incredibly hard working. ⁓ doing endless hours, very compassionate and she was absolutely devoted to her vocation of looking after sick people. Now, she did work gruelling and long hours in a notoriously low-paid job, pouring her heart into the care of others and she was a trusting soul who believed in the system. And when the diagnosis came, she was exactly the kind of vulnerable patient who would look to the medical establishment and the mainstream media as well for some hope. ⁓ This is exactly why these clickbait headlines of 30 % reduced risk make my blood boil because I know exactly how a terrified 23 year old girl reacts to them. They grasp at them like a life raft. They completely overhaul their lives, their diets and their peace of mind based on these inflated promises changes. So to tell a trusting desperate patient that a specific food will cut their risk of dying by nearly a third without explaining that the absolute reduction might be less than 2%. It's not just bad science. I think it's actually cruel. The mainstream media outlets that churn out these figures without any context are entirely complicit in peddling false hope to the people who can least afford to listen to it. So to understand how misleading these numbers can be, only need to look at maybe, let's look at another infamous medical marketing triumph. Years ago, a massive advertising campaign for a popular particular statin drug, proudly proclaimed that it reduced the risk of heart attack by 36%. Now, if you're worried about your heart, a 36 % reduction sounds like an absolute no-brainer, but that 36 % was a relative risk reduction. When you looked at the actual clinical trial, the ASCOT-LLAA trial for those interested, the absolute risk of having a heart attack for the group taking the placebo was 3%. For the group taking the statin, was around 1.9%. So the drug didn't reduce the risk by 36 out of 100. It reduced the risk by about 1.1 percentage points from 3 % to 1.9%. They simply took that tiny 1.1 % absolute difference, divided it by the 1.1 % baseline, and voila, a mathematically true but practically deceptive 36 % relative reduction was born. So what are we left with going back to the Shanghai Breast Cancer Study? It's important research and it does suggest an association between higher soy consumption and better outcomes for breast cancer survivors in a population where soy is a dietary staple. However, that 30 % reduction, while not entirely meaningless, needs to be understood in its full context. It's a relative risk reduction, not an absolute one, and the absolute impact depends heavily on the rarely stated baseline risk. It's based on a comparison between higher and lower soy consumers within a population where soy is already common, not against a soy-free group, which would be very telling, I think. And you've got the healthy user bias. And that definitely could play a role, making it challenging to attribute the benefits only to soy. So ultimately, while the study hints at potential benefits, it's a far cry from a definitive, isolated cause and effect of pronouncement. It's a call for more joined-up thinking, I think, when we see these sort of headlines. I don't want to leave you feeling angry or defeated though. fact, waking up to how these stats are manipulated should make you feel empowered. That's the whole point of it. I mean, this is exactly why I'm trying to get this information out to you in a calm, factual way. It's why I take the time to break down the nuance behind the headlines and why I always provide the links in the description so you can check what I'm saying for yourself. hope with genuine grounded understanding and a healthy healthy dose of common sense as well. And on a practical note, regarding your own health tracking, always remember that interpreting blood, for instance, is important in the context of the way you are eating. And that's absolutely crucial. Context is everything. Whether we're reading a sensationalized news headline or looking at lipid panel, always put it into context. And if you feel the same as me about this, please share this video with the people you care about. It might just save them a lot of unnecessary anxiety or some false hope. Or you could drop a comment below if this has touched a nerve or if you've been frustrated by these sort of misleading headlines yourself. And finally, please hit the subscribe button. It is completely free to subscribe and quite honestly, it gives me the encouragement I need to keep digging into the research and making videos like this one. Sorry, I got a little bit upset about the personal bit in the middle. Don't let the headline writers dictate your health. Look after yourselves, ask the hard questions, and I'll see you in the next one.