If eating saturated fat is so bad, why do the French, who every day eat much more of it than the Anglo-Saxons, suffer from less than a third the rate of heart disease of Brits and on average live four years longer than Americans? Nearly a third of French saturated-fat intakes comes from dairy products. Since the late 1980s, when epidemiologists noticed fourfold differences in mortality between the UK and France, the so-called French paradox has been a subject of much debate and speculation.1
For many years the British–French rivalry has extended from rugby matches, politics and trading insults to trading mortality statistics. Since records in France started to be accurately collected they have reported considerably fewer deaths from heart disease and a longer lifespan than the British. The French are proud of this, but many UK colleagues tell me that much of the difference is due to a reluctance to record deaths properly, with the same ‘Anglo-Saxon rigour’. Others disagree, asserting that misclassification could only explain at most 20 per cent of the difference, and point to a consistent north–south difference across Europe. Even within France itself there is a wide north–south difference, which suggests that most of the variation between UK and France is due to the healthier habits of the southerners.
What do the French do differently to give them this amazing advantage? The list is long. The regular wine drinking, cheese or yoghurt with every meal, the long dinner conversations about politics, culture and food, the relaxed attitude to marital affairs, the thirty-five-hour week, spending the whole of August on the beach, their love of regular strikes and street demonstrations, or their 70 per cent tax on the super-rich? Or maybe it’s simply that they appreciate their food more and enjoy the pleasure of savouring with family and friends the small tasty portions over several courses. Their choice of food is quite different too. They often eat raw meat, such as in steak tartare and steaks cooked rare and dripping with blood, earthy-smelling sausages made from intestines, unpasteurised cheeses, raw oysters and seafood, snails and frogs’ legs. They also cook practically everything in garlic and butter or olive oil.
The French enjoy a regular diet full of living things. Cheese, wine and yoghurt all teem with living microbes that during fermentation help make the food they eat so tasty and stop it going mouldy. Red wine drinking as the explanation of the difference between the two countries in heart disease rates has become the most popular theory, and has done much to boost red wine sales in the UK and America, as we will see later.
Could high-fat cheese be healthy?
Meat and cheese are perhaps the two most popular high-saturated-fat foods. Let us look at cheese first. Everyone with high cholesterol is familiar with the doctor’s advice to reduce or cut out cheese and to take a statin drug. Most cheese is made up of 30 to 40 per cent fat, and most of this is saturated fat which is traditionally considered the fat to avoid. The rest of the fat in cheese is of the poly- and mono-unsaturated varieties. Only about 1 per cent is actually cholesterol.
The French eat a hell of a lot of cheese – 24 kg (53 lb) a year per person – nearly double that of the average American and Brit (13 kg). Most cheese in France is eaten as real cheese purchased from a shop rather than as in the US, and to a lesser extent in the UK, in processed food products. These differences were even greater in the 1970s when US and UK total consumption was only a third of its current level. Charles de Gaulle famously asked in 1962, ‘How can you govern a country with two hundred and forty-six different types of cheese?’
De Gaulle was uncharacteristically modest in underestimating his country’s riches: France has probably as many as a thousand different varieties now, with many protected by law in the way they are traditionally made and with an Appellation d’Origine Contrôlée certificate of origin like the classification for wine. Of the top ten bestsellers at least four cheeses are unpasteurised, which the French believe gives them more taste and special properties. There are twenty-seven words for describing the different tastes and the enormous complexity of cheese. It contains a wide variety of microbes including bacteria, yeasts and fungi, and hundreds of species plus thousands of known and unknown strains.
The more artisanal the cheese-making process, the less sterile the conditions and the more diverse the microbes that grow in and on the cheese. The hundreds of natural microbe species plus the yeasts and moulds, particularly on the rind, provide more taste and better textures than the more industrial preparations. Despite the fears of other countries, outbreaks of cheese-related food poisoning are very rare. The French have a large cheese-science industry backing up their global market and are starting to seriously research the role of microbes. Unsurprisingly, the French cheese-research centres report mainly positive news about French cheese.
Some human clinical trials have shown that cheese supplements could be used to maintain the microbiome in people taking antibiotics, which normally knock out a large proportion of our healthy species. Unpasteurised hard cheese, when given with antibiotics, has been found to speed up recovery times and reduce bacterial resistance compared to sterile industrial cheeses. It was postulated that the cheese microbes may be helping maintain the greater diversity of microbes in our guts.2
When I visited friends in the Savoie region of France recently, the process of making traditional Alpage (high altitude) Comté cheese, which uses a recipe that hasn’t changed for centuries, was explained to me. The explanation (over much wine and cheese) took an hour, but the basic process involves mixing cold and warm cow’s milk outside in the spring mountain air (other cheeses use an added enzyme), which chemically cuts off the tail of the milk protein, which in turn allows it to curdle into lumps that stick to the fat. This lumpy mixture is passed through a fine linen net to drain away some of the liquid, then stored on old wooden shelves in a damp cellar. There the cheese is regularly rubbed with a milky rag from a vat on the cellar floor full of milk whey and brine, which gives the cheese a good crust teeming with microbes, including bacteria and fungi which change the acidity and the taste. For French cheeses the key to the many tastes are the other substances that the milky rag may be dipped in – such as, in the past, horse urine, which gave acidity as well as distinctive flavours.
Most real cheeses are left to age and mature (including hard cheeses like cheddar) and have a crust or rind that contains other, larger, microbes called cheese mites, visible under a powerful magnifying glass. These greedy creatures eat the microbes and cheese on the crust and make tiny holes to improve the flavour, but they are usually brushed off before the cheese goes to the shops. One cheese, Mimolette, used to arrive with so many mites crawling on it that the US health authorities banned it. After its banning, the bright orange cheese, a French seventeenth-century copy of aged Gouda, became a big black-market seller. Cheese fans liked the earthy taste of the rind. The mites, transparent and understandably chubby, can be seen wriggling happily in a hardcore YouTube video as they munch away on the cheese. The video comes with a warning that you may never want to eat French cheese again.3
These mites emphasise that cheese is very much alive – a living entity full of microbes – from the specialist milk bacteria, lactobacilli, to yeasts and fungi that are responsible for the tasty blue veins in cheeses such as Roquefort and Stilton. The US FDA (Food and Drug Administration) in their wisdom have also decided that having bacteria in cheese is a bit risky (unlike firearms), and have banned a number of other artisanal cheeses made from raw unpasteurised milk such as Comté, Reblochon and Beaufort. They have even recently announced a likely crackdown on cheeses that are allowed to mature on ‘old-fashioned’ wood surfaces that are difficult to sterilise. The perceived risk to the American public of eating traditional foods versus the ‘healthy’ industrial alternatives tells us lots about the balance of risk assessment in current health and dietary policy.
So while the safety-conscious FDA and the commercially minded US Department of Agriculture push ever more industrial cheese-type products of the sterile processed kind, containing little if any live bacteria, the French prefer their traditional cheeses. Even the ones sold in supermarkets are packed with trillions of microbes. If you leave some of these cheeses out of the fridge you can sometimes see them changing shape as the bacteria and yeast interact and battle with each other, breaking down the milk to produce energy. The high acid levels produced by the bacteria keep rival microbes away and stop the cheese going rancid.
The only signs of real cheese slowly going off are usually a mould that forms on the surface, like the famous penicillin strain or, in some cases, the strong smell of ammonia in the cheeses made with more moisture, like Taleggio, Limburger or Epoisse, which need eating more quickly. On their own the toxins produced by the fungi in cheeses can be harmful, but are broken down and safe inside the cheese. Whatever the smell, if the cheese tastes good, apparently the rule is that you can eat it safely.
I was particularly interested to see if the French paradox could be explained by the ingestion of huge amounts of friendly microbes contained in the cheese the French consume daily. I therefore conducted an intensive fromage experiment on myself (and four other willing volunteers from my lab). I wanted to test the best variety of French cheeses to provide a wide variety of microbes, so I asked an expert at my local cheese shop to advise me.
After a few days of discussion (and tastings) he came up with three unpasteurised cheeses for me: Brie de Meaux, blue-veined strong-tasting Roquefort, and a runny, smelly Epoisse that can be eaten with a spoon when ripe. I was to eat a large quantity – 180 grams a day (a normal generous serving is 30 grams). To help wash it down and to stick with French tradition I allowed myself two glasses of good full-bodied red wine, and for any hunger pangs three yoghurts a day. I usually have a bit of cheese once or twice a week, but I abstained the week before while I collected some stool samples for testing my normal levels before the three-day diet period.
For a cheese lover like me, this seemed a doddle. Day 1’s breakfast was easy – a nice big slab of Brie de Meaux on some brown bread; lunchtime was my Roquefort on some crackers, with an apple to dilute the strong taste; and the evening meal was a salad and my delicious Epoisse with bread and wine – perfect. The next day’s food was the same, and breakfast was easy; but lunch with the Roquefort was getting tough to digest – maybe because it is a whopping 31 per cent fat. In the evening the cheese I ate was still tasty but I started to feel quite full.
When Day 3 came I felt relieved that this was nearly the end of the experiment. From the morning onwards I had an odd bloated feeling, and as a by-product of ingesting much less fibre I ended up constipated for a few days. I felt full although my calorie count was not excessive. Each day from the cheese alone I was consuming about 800 calories and about 45 grams of saturated fat, way more than the ‘recommended’ allowance. And that was without counting my other foods and the yoghurt. I continued collecting stool samples for two weeks to see how long the effects of the cheese microbes would last.
Up till ten years ago, the only way to detect microbes was to grow them into visible colonies. You had to tease them to grow in culture plates for several weeks – and in those days we thought there was just a limited number of interesting bacteria. But it turned out that only about 1 per cent of our gut microbes are easy to grow in culture and these are the ones generally harmful to us – pathogens, in other words. New gene-sequencing methods have totally changed the process and uncovered the other 99 per cent of species we live with, most of which are never harmful.
I was keen to see the sequencing results when they came back from my collaborator Rob Knight and his lab in Colorado (the lab is now in San Diego). From my samples they extracted the combined DNA from all the microbes, then using gene-sequencers measured just one gene that all bacteria have in common, called the 16S gene. Each different species of bacteria has a distinctive version of the 16S gene which gives off a unique individual signature. When the analysis was finished around 1,000 species were arranged in groups and sub-families, which could then be compared across different people.
My baseline results were slightly surprising: the microbes from the stool samples coming from my gut looked slightly more like Venezuelans than most Americans. The two commonest groups (phyla) of gut microbes are Bacteroidetes and Firmicutes. I had a higher starting level of Firmicutes than I expected. One big question was whether any of the cheese microbes had survived the journey through my stomach and small intestine. It used to be thought that the stomach acid was so strong that it killed all microbes. Luckily for cheese microbes, this isn’t true. After just a day of the cheese diet my gut microbes had started to change, with big increases particularly in a number of the lactic acid bacilli (lactobacilli) and in the yeast penicillium.
The effects of the lactic acid microbes lasted a few days after the cheese stopped, then things started to return to normal, suggesting that the microbes couldn’t survive without more supplies. These results were reassuringly similar to a much more detailed experiment by a team at Harvard run by Peter Turnbaugh, who followed six volunteers on a meat and dairy diet (to be discussed later).4 After two weeks the good news was that I had increased the diversity of my microbes by a small but significant amount. However, it turned out that the results of the other four volunteers’ cheese diets could not have been predicted, and some didn’t change at all.
Whatever our diet, our microbes always carry our personal signature. This experiment showed that the particular composition of microbes that we are hosting varies a lot and may be the reason why many of us react in different ways to the same food. After my supersize cheese experiment – it took about two weeks for my bowels to return completely to normal – I felt like eating cheese again, showing that, like a kid in a sweetshop, you can sometimes have too much of a good thing.
The saturated-fat scare has hit the rails
Heart scares about eating too much dairy were widely reported in the 1980s and 90s, and have persisted. Some of these were caused by animal experiments in which mice or rats were fed large amounts of saturated fat, which increased their blood lipid levels and gave them signs of heart disease. But mice and men differ in many ways especially related to diet and health. Other scares came from epidemiology and we now know that many of those early studies were flawed, particularly the observational ones.
There were many other possible explanations for the large national differences between heart disease rates, discussed earlier. Critics that were brave enough accused the powerful anti-fat guru Ancel Keys of being very selective in the countries and the data he used. Others came to the opposite conclusions with the same data.5 Over the following years further studies were inconsistent and inconclusive. Nevertheless, the prevailing dietary-fat-leads-to-heart-disease hypothesis took root.
For many years the medical and scientific community who raised objections to this hypothesis were shouted down as crazy heretics. The invention and widespread use of the statin drugs reinforced opinions. These drugs, went the theory, in contrast to diet, rapidly reduce blood cholesterol levels and cut heart disease and mortality. Guidelines in the UK and the US suggest one in four adults should now be taking them. It was assumed this was due to statins’ cholesterol-lowering effects, but it turns out that was a red herring. The drugs’ main benefit comes from anti-inflammatory actions on the blood vessels, and they have both good and bad effects on many other diseases.6 With fresh eyes we can now look back more objectively at the accumulated diet data. A 2015 study re-examined the six early trials of the 1970s and 80s and found that although diets could reduce levels of cholesterol – contrary to the conclusions at the time – they had no effect on reducing heart disease.7 A meta-analysis summarised twenty-one large observational studies exploring saturated-fat consumption across the world comprising a total of 347,000 people. Of the 11,000 who later developed heart disease over the next twenty years, no association was found between amounts of saturated fat in the diet and subsequent heart disease or stroke.8
Now, the tide of evidence has started to shift in the opposite direction.
The ideal way to solve the problem of saturated fat would be to perform a gold-standard randomised clinical trial of a high-dairy versus low-dairy diet and watch the levels of heart disease. But this was thought to be neither ethical – a full-fat milk and cheese diet was ‘too dangerous’ – nor practical – it would need to last years and be very expensive. One compromise was to carry out six-week diet studies to explore changes in heart risk factors. One such study gave forty-nine volunteers an initial low-fat diet for six weeks, then added an extra 13 per cent of calories either as cheese or butter for another six weeks. The cheese group didn’t increase blood lipid levels or cholesterol at all, whereas the butter group did, showing that not all saturated fat is the same.9
The results now seem clear, particularly if you separate cheese from butter consumption. Far from being a risk for heart disease, full-fat cheese (but not butter), despite the saturated-fat content, now shows not only no harmful effects but a consistent protective effect on heart disease and mortality.10 So although we can’t depend on the reliability of these observational epidemiological studies, which have misled us in the past, we now have a reasonable hypothesis that the regular eating of traditional cheeses could actually prevent some heart and other health problems, owing to the extra microbes. Highly processed cheese or boiled or grilled cheese contains few viable microbes and doesn’t have the same benefits. Other products like milk and fermented products containing microbes may also offer some advantages, as we shall see.
As for explaining the French (or Mediterranean) paradox, cheese definitely plays a role, although no one will now be able to solve the mystery for sure. This is because death rates in both the UK and France, along with most of the Western world, have plummeted as a consequence of effective treatments that didn’t exist thirty years ago. Although the numbers of people with heart disease are still high, we can now keep them alive much longer after a heart attack. This is mainly thanks to minor surgery to unblock arteries, to drugs keeping our blood thin and to our blood pressure remaining well controlled.
The cheese pizza diet
Dan Janssen is thirty-nine years old and comes from a small town in Maryland, famous as the place where Babe Ruth got married. He likes cheese as well as pizza. In fact he likes it so much he has eaten it every day for the last twenty-five years – for every meal.
He tolerates the tomato sauce but won’t touch any other vegetable toppings. He washes it down with a sugary cola and usually eats a whole 14-inch pizza himself, which contains 45 grams of saturated fat and 1,300 calories. He clearly has an obsessive eating disorder but seems strangely normal in most other ways. He is slim, and apart from needing insulin for his diabetes, which he has had since he was a kid, he looks relatively healthy. His doctors have suggested he change to a healthier diet – but are amazed (and not a little annoyed) that his blood cholesterol and blood pressure are fine; and he controls his insulin with injections. He spent several years working in the local Domino’s pizza restaurant before eventually deciding to start a woodworking business.
When people tease him or say ‘You’re going to die!’ he retorts ‘We’re all going to die. But I’m going to die with pizza in my belly.’ His fiancée Madeleine, who like him is a vegetarian, tries to get him to eat vegetables (technically, tomatoes are a fruit). He has tried to please her, but gags and can’t finish the slice with even just a few vegetables on top. ‘Why ruin a good pizza with toppings!’ She has encouraged him to see a therapist, who believes his problems started as a kid.
‘Like when I was four or five, we lived in the backwoods of North Carolina, where I went to day care in a lady’s home. She would try to feed all of us Brunswick stew every day, which is not something you would ever feed a five-year-old. It’s either chicken, pork, or rabbit with beef and okra, lima beans, corn, potatoes, and tomatoes. I would protest and try to run away, but she would grab me. I can’t remember whether she would beat me or spank me, but I know that she would throw me in a closet as my punishment. I would sit in there crying and screaming for a couple hours until my mom came to pick me up.’
When asked if he had changed his habits after seeing his therapist regularly he said, ‘No. In fact, one of the reasons I like to go see her is because she’s in the city and I can go to Joe Squared [a pizza place nearby] after to get pizza.’
Exceptional stories like this are hard to explain via our conventional understanding of nutrition. Of course, we don’t know if Dan will drop dead next year or live to be a hundred, but if the latter, we’d be surprised. His intakes of saturated fat are very high – way over most countries’ official ‘recommended levels’ of 20 to 30 grams a day, and he eats very little fibre. But what if some people have actually adapted to this type of high-fat dairy diet and yet remain healthy? In the influential Seven Countries study after the war Ancel Keys highlighted the island of Crete as the place where the lowest cholesterol levels (half of US levels at the time) and the lowest rates of heart disease were to be found.
Cretan cholesterol and centenarians
The Cretan villages were in mountainous areas, isolated and poor. Most of the people were shepherds or fishermen, and despite very hard lives and no real medical facilities there were large numbers of centenarians among them. What Keys and his colleagues didn’t highlight at the time were the huge amounts of animal and vegetable fat and dairy produce these villagers were eating. A geneticist colleague of mine, Ele Zeggini, has been looking more closely at a sample of these villages fifty years later. It turns out they are all quite different, isolated from each other, and with strong local dialects and customs.
One small village that Keys didn’t look at comprises just over four thousand people called Anogia (which means ‘upper earth’ in Greek). It is nearly 3,000 feet up the north face of Mount Idi, where they rarely eat fish but do eat a lot of goat’s cheese and yoghurt every day. The only real difference in their diet over the last few hundred years is that they can now afford to eat meat (usually goat) more regularly, whereas before it was only for special occasions. They are also getting quite lazy and would rather drive 400 yards than walk it.
These villagers are part of a national nutritional study and have regular check-ups and blood tests. What they show is that their total cholesterol levels in their blood are higher (just over 5 mmol/l) and so in theory less healthy than in some places in Greece and similar to north Europeans’; but importantly, while they do get cancer, unlike the rest of the country they show no signs of heart disease.
What Ele and her team found was that a high proportion of the population carried a mutation in a gene called APOC3, which explained the extra-elevated levels of the good lipid transporter found in blood, HDL, plus low levels of the harmful lipid triglyceride, providing protection for the heart despite the high-fat diets. Now, this isolated and partly inbred village with plenty of cousins had something in common with another unexpected population the other side of the world who also consume vast amounts of cheese and dairy – the Amish in the US. Amazingly, they also share the same rare genetic mutation providing the heart-protection gene that usually occurs in less than 1 in 50,000 people.11
This story shows how populations might adapt to odd diets and environments over relatively short times – a bit like the Masai in East Africa with their high-fat intakes of meat and milk and their blood drinking, or nomadic Mongolians who exist on fermented milk, meat and little else.
As well as the genes changing, it could also be that the microbes adapt. Microbes can add a generation every thirty minutes and indeed, they adapt a lot faster than we can. I haven’t yet been able to test Dan the Pizza Man, but he may well have some cheese-loving gene mutations, as well as definitely hosting cheese-loving microbes in his gut. Microbes do not yet appear on food labels, so it’s hard to guess how many friendly microbes are still alive in the cheese toppings of a Domino’s pizza. Apparently, they are made up of a blend of frozen cheeses and starches. However, unless you are really sure of your genes and microbes, I don’t think I would recommend the cheese-pizza diet.
Industrial cheese is a by-product of the excess-milk lakes that built up as milk went out of fashion in the US and Europe. This was led by giant processed food companies like Kraft Foods, who in the 1950s developed methods for transporting cheese with a shelf life of several months around the US. They had hit products like Cheez Whiz, which coming as a bright-orange sauce was as far from artisanal European cheese as you can get. The process involved boiling and spinning the cheese or treating it with chemicals so that the fat and the milk blended together. This produced a sterile product (no live microbes) that could be, and still is, added to more or less anything to enhance taste, addictive qualities and consistency.
The combination it has worked best with is pizza, which has become arguably the world’s most popular meal. It has steadily and frighteningly become the principal source (14 per cent) of the US saturated-fat intake and a third of total calories; and one-third of young Americans eat it daily. This is astonishing for a food that was only created in its modern form in Naples in 1889 (for Queen Margherita) and brought to the US in 1905. Most of the pizza eaten today is not, of course, the artisanal fresh product common in Italy, but cheap, processed and frozen in a business worth over $40 billion in the US alone. Some advertised pizzas have so much cheese stuffed into the topping and crust that just one slice equals 14 grams of fat and 340 calories. As long as it is cheap and durable, there seems to be no end to what cheese can be added to.
US cheese consumption has increased fourfold since 1970, ironically mirroring a decline in milk drinking because of fears of its fat content. Although it contradicts their own diet guidelines, the US Department of Agriculture (USDA) and the farmers are happy.12 Exports of ‘real’ American cheese pizza have also boomed, especially in fast-expanding Mexico.
Fromage nature and mosquitoes
Another non-traditional form of cheese is made when you mix certain bacteria with milk. You can personalise these cheeses to uniquely suit you as an individual. All you have to do is take swabs of your armpits, belly button and between your toes and mix what you have garnered with milk, then add some lactic acid bacilli, and hey presto your own personal and very individual cheese! Christina Agapakis from Los Angeles (UCLA) created such works of art with Norwegian sensory artists for a recent exhibition in Dublin called Selfmade. The cheeses look like any regular cow’s or sheep’s cheese and each one is named after the donor of the bacteria. The bacteria used for making normal cheeses are closely related to those in the darker, less washed parts of our bodies.
The famously pungent Limburger cheese is made from the same bacteria that many people have between their toes (Brevibacteria linens), the ones that cause smelly feet. The composition of bacteria inhabiting these areas may make you more, or less, attractive to other animals. Mosquitoes in particular seem very sensitive, and different species will avoid some bacterial odours and make a beeline for others, explaining why some people appear immune to insect bites. Our lucky UK twins were tested recently in a lab study putting their hands into a plastic bubble full of mosquitoes, then counting the attempted bites. There were big differences, and the attractiveness of some people to mosquitoes was clearly shown to be genetic.
Smell is a very subjective sensation, as the UCLA team found out when they prompted subjects before a controlled smelling experiment. Those told beforehand that the bacteria smelled like cheese subsequently reported it as a nice smell, but others who were told it was taken from a body said it smelled disgusting. Christina has apparently tasted ‘her belly-button cheese’ and it was ‘just like a normal mild cheese’. Human cheeses haven’t yet become a regular part of our diet – but who knows? As the ultimate selfie, it could catch on.
Bulgarian health food
Yoghurts are another common source of our saturated fats, although the amounts vary widely across styles, particularly when you include the recently introduced low-fat varieties. Yoghurts are made from goat’s, cow’s or sheep’s milk and come in many different forms, thicknesses and consistencies. Greek yoghurt, which is more solid as the liquid is strained off, is now very popular. Traditional Greek yoghurt also contains the most saturated fat, often 14 grams per pot, and it also has plenty of vitamin B12, folic acid and calcium. Basically, the more traditional and natural the mix, the more the amount of saturated fat. The popular low- or zero-fat yoghurts obviously contain the least, but these use either artificial sweeteners or several spoonfuls of sugar or its equivalent in concentrated sweet fruit to compensate for the lack of flavour, and usually contain fewer vitamins and nutrients.
In the early 1900s, the pioneering Russian immunologist Dr Elie Metchnikoff was working as the first serious yoghurt scientist. He was an eminent researcher who in 1908 shared (with the scientist Paul Ehrlich) the Nobel Prize for showing that white cells were not villains but actually good guys when it came to fighting infection. He suggested for the first time that bacteria, like white cells, had been misrepresented as always evil, and that a symbiotic relationship exists between them and us. He said that ‘intestinal flora are the principal cause of the too short duration of our life, which flickers out before having reached its goal … It is hoped that the opening century will witness the solution to this great problem.’
He came up with his theory after observing that Bulgarian peasants ate a lot of local yoghurt and despite hardship managed to live relatively long lives. He proposed a novel idea for those days – one that seems quite obvious now – that a link exists between better health and longer life. His theory was that ageing is caused by rotting toxic bacteria in the gut and that consuming lactic acid (producing bacteria in milk and yoghurt) could counteract this, thus prolonging life. He took his own medicine and drank sour milk every day from then on. He outlived both his wives, who didn’t share his tastes, and after working at the Institut Pasteur in Paris lived to the age of seventy-one.
One of his admirers was Isaac Carasso, a wealthy Jewish Catalan who was working in the Balkans just before the First World War when he heard of Metchnikoff’s work and thought it had great potential. His firm became the global company Danone, now worth around 35 billion euros. Another disciple was the Japanese doctor Minoru Shirota, who in the 1920s in Kyoto was looking for remedies for preventing infection. He cultured special strains of so-called friendly lactobacteria, which he modestly named after himself (Lactobacillus casei Shirota). His commercial skills led to the worldwide marketing of the yoghurt brand Yakult in 1935. It’s not clear how much yoghurt he drank himself, but he lived to be eighty-three.
Today, there could be an ecological downside to this surge in Greek yoghurt production around the world. The unused whey protein left over from the filtered yoghurt is so acidic that it is illegal to get rid of it in the normal way as it could devastate plants and wildlife. In the north-east US 150 million gallons of toxic whey form giant lakes as it waits to be disposed of. Eco-pioneers are experimenting with mixing it with animal manure, using the bacterial fermentation to produce methane gas, which although rather smelly could potentially allow yoghurt to power generators.
Despite the ‘unhealthy’ saturated fat and concentration of calories, dairy products could, paradoxically, help us lose weight. In several trials comparing dairy and non-dairy diets, the dairy group in each study lost slightly more weight. This was only significant if the groups were also calorie-restricting – that is, trying to lose weight at the same time – but the dairy consumers lost a significant amount of body fat while gaining lean muscle. This suggests that some constituents of dairy products could reduce visceral belly fat, which if true would be a major bonus. Although it is still unclear whether the actual saturated-fat levels are crucial or not, the passenger microbes may be.13
One of the recurring themes of this book is the surprising lack of good evidence for any health benefits or harm potential of any particular food product. Specifically for yoghurt eating and weight loss there have been only two randomised clinical trials, both small, short-term and inconclusive. There are, however, six large population (albeit observational) cohort studies following the yoghurt habits of over 150,000 people over time. Four of the six had a positive finding, such as a recent Spanish study of 8,000 men and women followed over six and a half years which found some slight weight loss when full-fat yoghurt was being consumed. This translated into a 40 per cent reduction in the risk of becoming obese if they ate at least one serving daily.14 So again we see that increasing your proportion of calories from dairy certainly doesn’t make you gain weight, as we previously believed,15 and could even help a bit if you also want to lose a few pounds.
Short-term yoghurt experiments show improvements in the B vitamin thiamine, produced uniquely by human gut microbes. Other studies in mice with some special strains of lactobacillus (including the Bulgarian strain) have shown improvements in immunity.16 But direct and consistent evidence of immune benefits in humans remains absent, apart from one small study which showed a reduction in common colds in the elderly. Our own twins studies are, however, showing some clear and important links in humans between microbes, diet and the immune system, which I will discuss later.
Super-microbes and probiotics
All yoghurt contains in large amounts the same group of bacteria that start the fermenting process in milk; the lactic acid bacilli, or lactobacilli, already referred to. These microbes help the digestion of lactose. Yoghurts vary in the amounts and the precise strains they contain, and in whichever other bacteria may be present naturally or artificially added. Most of the bacteria in natural yoghurt are strains that don’t normally live in our intestines. When so-called gut-friendly bacteria are added to food in sufficient amounts and they are claimed to have potential health benefits, they are called ‘probiotics’.
Probiotics are now big business. There has been much controversy about the health claims of such additions to yoghurts and other milk products. They are often sold separately in health food shops to combat antibiotic use or stomach upsets; others are said to increase and boost the immune systems of both healthy and diseased individuals, and they are marketed with an increasingly bizarre list of health claims. Most of the species sold commercially are related to lactobacilli or bifidobacteria.
The best scientific evidence that probiotics actually work comes not from yoghurt commercials but from research into the prevention of a nasty and potentially fatal disease known as antibiotic-induced disease, which often occurs in highly susceptible pre-term babies or elderly patients. Antibiotics (which we discuss again later) are widely used to treat infections when certain rare pathogenic bacteria species grow out of control. These often work, but they usually cause collateral damage, killing many friendly species and altering the natural community, the microbiome. This can lead to some of these pathogenic bacteria having no natural enemies and taking over; they become resistant to even the most powerful antibiotics.
A combination of lactobacilli and bifido microbes is added to many types of yoghurt, which are recommended to prevent the severe gut infection C. diff (full name, Clostridium difficile). This is a pathogenic bacteria that can take over the antibiotic-treated gut in a significant proportion of hospitalised people, particularly females and the elderly. A recent meta-analysis of twenty-one clinical trials found a 60 per cent benefit in taking these probiotics for three weeks; although it doesn’t always work, on average one C. diff case was prevented for every eight prescriptions of probiotics, making it highly cost-effective.17
However, on the high street and the internet the market is not well controlled, and as well as their efficacy often being absurdly exaggerated many probiotic products contain contaminated or even dead bacteria, or sub-optimal numbers. This combination of factors has led authorities in Europe and the US to clamp down on the claims that the yoghurt companies can make without first conducting better trials. This has left the manufacturers in a catch-22 situation, as the authorities are now treating probiotics like new drugs, and if being tested for health benefits they have to show rigorous evidence of both efficacy and safety, which would cost millions. The companies argue that probiotics are just food and that the FDA don’t demand a trial when a new type of porridge is released. For the moment neither side is giving ground, and large-scale responsible probiotic trials in humans seem a way off.
A meta-analysis of several probiotic trials concluded that there was little evidence of consistent benefit, either because they don’t work or because most are small and short-term studies.18 One notable exception is the trial of a particular strain of a probiotic called Lactobacillus reuteri in people with hypercholesterolaemia (genetically very high blood cholesterol, mentioned earlier).
Could microbes be the new fat eaters?
There is an emerging link between microbes and our blood lipids. Germ-free mice suffer from high lipid and cholesterol levels because they lack a key product that accumulates in mice’s gall bladders called bile salts. It turned out that microbes were crucial in enabling these bile salts to do their normal mopping-up job. A group of Montreal researchers tested patients with very high cholesterol levels first by feeding them the probiotic in yoghurt over two weeks – with favourable results.19 They then put the same microbes into capsules and gave them to a similar group of patients over nine weeks. They found a 10 per cent reduction in all the harmful blood lipids and an increase in the protective ones which matched the beneficial change in the mice’s bile salts.20
What is strange is that despite the increasing general evidence of the health benefits of yoghurt, there is still little evidence that the probiotic friendly bacteria actually survive and replicate in the human gut. Of the basic starter lactobacilli probiotic, studies show that only 1 per cent passes from the acid stomach to the duodenum and then the trail goes cold. Most studies can find no later trace of viable microbes in the stools, or evidence that they survived in the colon.21 Increasing severalfold the concentration of microbes in the yoghurt or using a slightly different strain can sometimes pay off, but overall it looks like no one single microbial type suits us all. The common probiotic strains in commercial yoghurt may just work well in one person but not in another.
This may be because a particular gut has the ‘wrong environment’ owing to some individual chemical signals or conditions; or like a new kid at school, the microbe in question is vastly outnumbered, gets bullied by the other microbes and never settles. One small but careful yoghurt trial was performed, and with some interesting results. Seven young female identical twin pairs volunteered to take yoghurt containing five well-known species of bacteria commonly found in many commercial brands. They all ate the yoghurt twice daily for seven weeks. The US team, led by the microbiome pioneer Jeff Gordon, reassuringly found that fairly high levels of the five microbes reached the colons of the twins, particularly one of the bifidos which lasted for a week after the yoghurt stopped.22
However, against expectations these microbes were apparently not having much impact. The team found no difference to the composition of the other resident gut microbes, who seemed unfazed by the arrival of the newcomers. When they then put the same five microbes into a highly controlled group of mice, they got the same result. They could detect the yoghurt microbes but they were not perturbing the other residents. Some scientists would have stopped there, but the team did a series of complex tests to show that the yoghurt microbes had had major effects behind the scenes. They had somehow massively ramped up the activity levels of genes that control the breakdown of complex carbohydrates and sugars found in fruit and vegetables.
So the consumption of yoghurt microbes alters the way we break down other foods and initiate anti-inflammatory pathways. Our microbial communities work together in large interacting networks to metabolise our food in many different and complex ways. So introducing a single microbe may not be sufficient to alter the balance of the multitude of others, but it could still alter the overall metabolic balance of the community and affect our health.
There is a caveat. I worry that many commercial yoghurt brands are overselling their benefits, especially when they are adding only low doses of one or two of their special patented microbial strains. Many low-fat yoghurts are full of sugar (or puréed fruit), which could negate the benefits as sugar stops bacteria growing. Since yoghurt bacteria usually can’t survive in most humans unless replenished, you may need to eat them every day to be effective. It could also be that the exact strain of bacteria or community of bacteria is crucial, and that we need to be matched up more precisely with our microbial soulmates.
Personalised microbes and designer yoghurts
Studies of probiotics have produced generally good, consistent results in lab mice that are genetically similar and have identical environments, but human studies are disappointing, showing little replication between studies. This could be because human guts possess an amazing diversity of microbial species between people. This poses the question, do our own human genes determine which microbes are attracted to us? This is important, as it might explain why some probiotics or yoghurts work for some and not for others.
Scientists like me with a genetics background often believe that genes are important for everything in our bodies that is biologically useful. Not everyone agrees. Non-geneticists say that the remarkable diversity between humans suggests that it is the random effects of our surroundings and of the food we eat that are the major determinant. Two earlier twins studies in the US produced no conclusive evidence of a genetic influence. But when I heard microbiome expert and future collaborator Ruth Ley present the results at a meeting I thought the studies were too small, and that we could answer the question properly with my cohort of 11,000 twins.
I had spent twenty-odd years performing twins studies on hundreds of different traits, from religious beliefs and sexual preferences to vitamin D and body-fat levels, in order to determine whether characteristics or diseases were mainly influenced by genes or by environment. The design is simple: you compare the similarity of identical twins with the similarity of non-identical or fraternal twins. If the similarity is greater in the identical pairs, then genes are definitely involved, as identical twins share all their genes and non-identical twins share only 50 per cent of them, like normal siblings. Using a bit of simple maths you can work out what proportion of the differences between people are due to their genes: this is called the per cent heritability.
Our faithful twins were persuaded to part with a tiny stool sample, which was kept frozen and sent to Ruth Ley at Cornell University. She and her team extracted the DNA and then sequenced one highly variable gene, 16S (mentioned earlier), that uniquely separates the species. By identifying the proportions of the thousand or so major microbe groups for each person we could start to compare them. The first thing we noticed was the lack of similarity between any two people – the variety was amazing. Overall, even our identical twins shared only slightly over 50 per cent of the major patterns of microbes, and the general population share only around 40.
It was only when we divided the microbes into the major classification groups, or phyla, that a much clearer pattern emerged. Although diet and other environmental effects predominated for many major groups, like the Bacteroidetes, many of the smaller subgroups of bacterial families, like lactobacillus and bifidos, found to have major effects on diet, obesity and disease, were also seen to be under genetic influence. This control is, however, partial and over 60 per cent of the effects on these microbes still come from the environment.23
Our results were a surprise to most scientists in the field. What this means is that the number and type of microbes (like lactobacilli) that grow best in our guts are in part controlled by our genes – it’s a bit like certain flowers or shrubs preferring different types of soil. This could explain why the current limited range of probiotics work in some people and not others. Until a wider range of probiotics are developed and ways of delivering them to our colons are improved, real foods containing both a wide range of beneficial bacteria and microbe fertilisers would have a better chance of helping most of us than gambling on just a few added species of bacteria.
Working with a thousand of our twins and with Mario Roederer at the US National Institute of Health, we have just performed a novel experiment on the immune cells found both in our intestines and in our blood that communicate most with the microbes, called Tregs (short for T-regulatory cells). We found that they vary between people and are strongly controlled by genes.24 So our genes determine to some extent which microbes live and prosper in our guts – using our soil analogy – and then, importantly, how our immune system reacts and responds to them.
So genes are not inflexible, as previously believed, but can be turned up or down like dimmer switches. This ‘epigenetic process’ is how we humans adapt to new surroundings and new diets. It is also how we communicate with our microbes and how our microbes can switch our genes on and off and manipulate us. In this way, over time our diets (or microbes) could slowly alter the way our genes fertilise (the soil of) the colon, allowing a greater variety of microbes to eventually settle and grow.
Listen to the brain in your gut
Microbes in our gut were essential in allowing us as newborn babies to develop a normal brain and nervous system. There is good evidence that microbes in general and lactobacilli and bifidos in yoghurt in particular can affect key centres in the brain via the brain–gut axis. The gut is the second-largest network of nerves outside of our heads and has been called our second brain. It has been estimated that the kilometres of neurones and nerve connections in our intestines are the same size as the brain of a cat – an animal we admire for its cunning, self-centredness and nine lives. We should listen more to our guts.
A complex system of signals relays information to and from the brain and the gut. These connections control many functions, particularly eating and digesting food, but we are finding they do much more: for instance, they can influence our mood. Sufferers and their doctors have long known that stomach upsets send signals to the brain to initiate nausea, prevent eating, reduce activity and dampen mood, which can lead to short-term depression.
We looked recently at rare twin pairs containing one depressed twin and one happy one. In the affected twin’s blood we found altered levels of the key brain chemical serotonin. This chemical comes from our food except when we are fasting, when our gut microbes manufacture it for us. So changes in our microbes will alter this key brain chemical and, potentially, our mood. This could explain the odd sensations of a euphoric high reported by some people after fasting.
There are known to be at least sixteen specific messenger chemicals, called gut hormones, released from the intestines into the blood that send signals to the brain to eat more or less food. They are carefully regulated by our genes and by what we eat. In times of stress, the brain can also alter the function of the bowel via our emotions, which in turn can alter other hormones, leading in a vicious circle of symptoms to the disruption of microbiomes and even to depression. But these messenger hormones don’t work alone.
We are learning that the immune system is another player in the crucial gut–brain connection. Cells like the Tregs, mentioned earlier, are in constant contact with both our microbes and our brains and act as messengers between the two.25 In humans, one of the commonest complaints affecting our intestines is irritable bowel syndrome (IBS), which classically affects women more than men and has its peak incidence between the ages of thirty and sixty. The cause is unknown, although stress is a factor. What has always confused doctors and made them sometimes think the symptoms – and the disease – were invented consciously or subconsciously by the patients is the fact that 50 per cent of cases are associated with psychiatric symptoms such as anxiety or depression or chronic non-specific pain.
Stress, swimming pools and IBS
According to Sally, the worst thing about having IBS over the last twenty-seven years has been ‘never being quite certain what to expect and how long it will last. It occurs before interviews, after shocks, like when my daughter was assaulted, or just when I feel very anxious. The bouts generally last from a few weeks to several months. I did have an investigation to see if it was cancerous. It’s funny, though, because when I was growing up I had no idea that people “went to the loo” every day. My bowels used to go up to three weeks before I emptied them. And yes, it did used to hurt. Since having IBS I can go about five or six times a day when it’s not great, when it equalises, and I now “go to the loo” most days.
‘I did have salmonella at one point in 1989, but that didn’t really affect them. My “trigger” seems just to be stress. My sister was raped and her boyfriend and my dad were beaten up and left for dead when I was seventeen, and my symptoms started a few weeks after that. I have tried various treatments, but they haven’t helped much. I admit my diet is not very good – I eat a lot of white bread and go to McDonald’s more than I should. In fact a burger often triggers pain or bowel problems. I’ve not tried regular probiotics or healthy yoghurt because if anything they frighten me!’
Sally is overweight at 13½ stone and hasn’t been able to lose weight. She puts her poor health and diet down to chronic stress and ten years of unemployment. We measured the composition of Sally’s gut microbes compared to a thousand control women. Sally had less of the common Bacteroidetes species and, interestingly, seven times more Actinobacter than other women, which are more common on the skin. She also had much less microbial diversity overall.
IBS is a disease that was hardly recognised fifty years ago and now affects over 10 per cent of people in most surveys, but is especially hard to define as there are no specific tests. It is characterised by altered bowel habit, bloating and abdominal pain, switching between bouts of constipation and diarrhoea, often ending in a dash to the toilet after meals. There have been over twenty gut microbe studies looking at groups of IBS patients. The results show clear abnormalities in the IBS sufferers but no consistent pattern indicating exactly which microbes are altered; but all, like Sally, have reduced levels of microbial diversity.26
A few studies have tried antibiotics as a treatment, with limited success. Some companies are making antibiotics more efficient by means of a special capsule that protects them until they reach the small intestine and colon. Probiotics have had some partial success in treating IBS symptoms in most of the forty-plus studies, although many of them have been small, short-term and unreliable.27 Up to half of IBS patients show some evidence of having experienced leaky intestines, when small amounts of chemicals or even microbes can cross from the gut into the blood. But whether this is directly related to the change in microbial diversity or happens after the disease has started is still unclear.
We all recognise that mood and eating habits often go together. Many human and rodent studies have shown that stress can lead to either losing weight or overeating and increasing blood lipids.
Only one group studied are immune to extreme stress and act cool under pressure. These ‘Die Hard’ Bruce Willis types are in fact germ-free mice that revert to normal wimps when bacteria are reintroduced into their guts. So, clearly, our microbes are crucial in transmitting anxiety. I remember crying when as a child I lost my mother at our local swimming pool; it is still a painful memory. Researchers can detect stress in rodents and have done the same to young mice. When, after separating them from their mothers, they forced them to swim laps, their microbiomes were disrupted and became less diverse. The anxiety and stress were consistently reversed by giving them lactobacilli probiotics, which suggests that yoghurt, not crisps, should be our treat after going swimming.28
Only a couple of good probiotic trials of mood in humans have been published. One showed women photos of angry and friendly faces and monitored their brain activity after four weeks of probiotics. They found no major changes in mood but did see a toned-down emotional response. Another, similar, trial involving both men and women showed a reduced blood cortisol level (which is increased by stress), also after a month of probiotics. One study, sponsored by a yoghurt company, showed that the microbes in yoghurt rather than the milk itself can light up key centres of the brain, again dampening down negative thoughts.29 Ice-cream was earlier reported to have the same effect – and proved to be a marketing sensation for Häagen-Dazs. So let’s not get too excited.30
Overall the story for the health benefits of probiotics is cautiously favourable. They clearly work well when the body is at its weakest and the microbes are disrupted or not yet formed, as in infants, after infections, or in the elderly. For the majority of normal healthy people (unlike lab mice), there is as yet no incontrovertible evidence from randomised clinical trials that for most of us taking them regularly provides any obvious benefit. But these are early days: we are testing only a handful of the protective microbes and we don’t yet know what are the ideal conditions to provide them with.
A small number of classic microbes have been used by yoghurt companies for over seventy years, and with sales rising they don’t want to change a winning formula. As discussed earlier, many commercial yoghurts while promoting their low-fat status contain lots of concentrated fruit and sugar or sweeteners which could inhibit the growth or function of many microbes. So avoid these and keep your yoghurt natural and with plenty of microbes. Only a tiny number of them survive, so to boost your chances look for products that are likely to be genuine and with plenty of spare microbes, meaning over 1 billion colony-forming units (CFUs), hopefully listed in the small print on the label.
The overall conclusion here is that for most of us saturated fat is not the villain to be avoided at all costs. The saturated fat many people eat in products like cheese and yoghurt is not, as we have so often been told, unhealthy, but likely to be beneficial. This is provided the food is ‘real’ and contains living microbes, and is not over-processed or full of other unwanted chemicals and sweeteners.