Lactose, fructose, sorbitol: which intolerance you actually have

There is a moment that many people with abdominal complaints know. You have a list. On the list are milk, apples, onions, perhaps wheat, perhaps dried fruit. The list has grown over years, line by line, one new line after every bad evening. And still the belly has not become quiet.

Then comes a test. The test says: fructose positive. Or lactose positive. Or both. And for a moment that feels good, because at last something has a name.

Except that a breath test does not measure whether you are ill. It measures whether a sugar was completely taken up in the small intestine. Those are two different questions, and the mixing up of these two questions is the reason why so many people sit there with a very short menu and a very long road of suffering.

This text sorts the three sugars apart: lactose, fructose, sorbitol. It explains what the hydrogen breath test can do and where it goes wide of the mark. It says why self tests and IgG panels contribute nothing in this field. And it shows why avoidance is a tool on a timer and not a life plan.

Intolerance, allergy, malabsorption: three words, three different things

Imagine the small intestine as a very long, very well staffed border. Everything you eat has to get through. For every substance there is its own gate. For milk sugar a pair of scissors that cuts it into two parts. For fruit sugar a revolving door with limited capacity. For sugar alcohols almost nothing at all.

What does not get through does not disappear. It moves on into the large bowel. There, bacteria are waiting that are delighted by every undigested carbohydrate. They ferment it, gas is produced, and water is drawn into the bowel tube. That is the whole mechanism. No defect, no inflammation, no immune system.

Why the terms do not mean the same thing

The review by Misselwitz and colleagues in the journal Gut cleanly separates three levels for lactose, and this separation can be transferred to the other sugars. Lactase deficiency is the absence of the enzyme in the small intestine. Lactose malabsorption is any cause that prevents uptake, so the secondary ones as well. Lactose intolerance is only present when a person with malabsorption actually gets abdominal pain, bloating or diarrhoea after lactose.

The German language review by Christiane Schäfer in the Bundesgesundheitsblatt explicitly classifies the carbohydrate malassimilations as dose dependent but not immunological reactions. It distinguishes even more finely: hypolactasia versus lactose maldigestion, fructose malabsorption versus fructose overload, combined forms, and the isolated reaction to sorbitol. Fructose overload simply means: too much at once, with a completely normal uptake capacity.

Histamine also belongs in this row of terms that get muddled, but it works to a completely different logic: there it is about a biogenic amine and about its breakdown, not about a sugar in the bowel tube. How that fits together is set out in detail in the article on histamine intolerance and DAO deficiency.

Lactose: why the majority of humanity is the norm and we are the exception

Perhaps you have asked yourself why of all things milk is such a topic. Why there is a lactose free product on every second shelf, while the same milk is not even drunk in large parts of the world.

The answer is one of the nicer stories in biology, and it turns the usual account on its head.

Lactase is childhood equipment

Lactose is a double sugar made of glucose and galactose. For it to get through the intestinal wall it has to be split. For that, the enzyme lactase sits on the surface of the small intestinal villi. Every mammal has it as long as it is being suckled. And in almost every mammal the activity drops off after weaning, because it is no longer needed after that.

In humans it is exactly the same. The fall in lactase activity over the course of childhood is the normal state. What calls for explanation is the opposite: that a part of humanity keeps this enzyme for a lifetime. This lactase persistence is the special case, not the deficiency.

In people of European descent it essentially goes back to a single change of letter, an exchange of C for T at position minus 13910 in front of the lactase gene on chromosome 2. This variant switches the gene switch on permanently.

Exactly here lies a practical consequence that comes up almost daily in a Berlin practice. An analysis of worldwide data on the frequency of lactase persistence came to the conclusion that the known genotype data do not explain the observed frequencies in large parts of West and South Africa, South East Europe, the Middle East and parts of Central and South Asia. Variants are therefore still missing there. A negative genetic test in people with origins outside Northern Europe therefore says less than it appears to say.

The review by Ségurel and Bon in the Annual Reviews counts five known variants and records that the trait is monogenic in Eurasia but predominantly polygenic in Africa. The authors expressly call the story, despite its textbook status, far from being clear. They ask, for example, why the frequencies are low in Central Asian herding peoples and high in some African hunter gatherer groups.

How strongly origin, dietary history and metabolism are connected at all is a topic of its own. The article on genetics, metabolism and origin follows that up. Here it stays with the one sentence that counts for practice: the genetic disposition describes a starting position, not what you experience today.

The figure that changes something in practice

The most useful part of the lactose story is not a percentage, it is a gram figure. The evidence review by the Agency for Healthcare Research and Quality evaluated original work from 1967 to 2009 and came to a very clear finding on the dose.

These figures come from the AHRQ report and from its short version in the Annals of Internal Medicine. Both stress the same limitation: most of the included studies examined people with lactose malabsorption, not with a confirmed intolerance. And the authors explicitly found no sufficient evidence that a lactose free solution with zero to two grams lowers the complaints compared with more than twelve grams.

Two years later the same group repeated the experiment with double the amount, that is twice 240 millilitres daily over seven days each. There too the complaints did not rise meaningfully on normal milk. Another finding was interesting: the group that described themselves as intolerant reported more gas complaints in both periods, so on the lactose free milk as well. The complaints are real. The trigger they were attributed to was simply not correct.

How dose dependent the whole thing is is shown by a Chinese study with 60 people with diarrhoea predominant irritable bowel syndrome and 60 healthy controls. They received 10, 20 and 40 grams of lactose in random order, blinded to the dose. At 40 grams, 93 percent of the controls malabsorbed, but only 68 percent had complaints. In the irritable bowel group the complaint frequencies were higher at every dose, at 10 grams 18 versus 3 percent, at 20 grams 47 versus 22 percent, at 40 grams 85 versus 68 percent.

The decisive sentence of this work stands at the end: self reported lactose intolerance did not correlate with the result of the breath test. It did correlate very much with dairy products being avoided. The label therefore steers behaviour without matching the measurement.

Fructose: one transporter, one partner and a rare disease that is called something else

With fructose the confusion begins with the name. There are two completely different conditions that sound almost the same, and mixing them up is the one genuine danger spot in this whole field. So first the mechanism, then the clear distinction.

The revolving door is called GLUT5

Free fructose is not split. It has to get through the intestinal wall as a whole, and for that there is a transporter of its own with the name GLUT5. It works without expenditure of energy and with limited capacity. When more fructose arrives than the revolving door lets through per unit of time, the rest stays in the bowel tube.

At this point many texts are too quick. The paediatric review by Ebert and Witt expressly records that the involvement of defective transporters such as GLUT5 or GLUT2 in the development of fructose malabsorption is a matter of debate. The sentences you read everywhere, that GLUT5 is defective, are therefore not an established fact. Just as plausible is a simple physiological capacity limit that exists in all people and only lies at different heights.

The classic review by Rumessen brings the point down to a formula that explains everyday life better than any list: fructose as sucrose or in an equimolar combination with glucose is taken up well, and only the fructose in excess of the glucose is malabsorbed. The same work also describes that sorbitol can increase the malabsorption of free fructose.

The Monash group around Peter Gibson placed this in a larger picture in 2007. Up to half the population cannot completely take up a load of 25 grams of free fructose. The average daily intake worldwide lies between 11 and 54 grams. And their key sentence is that the confusion about the clinical meaning of fructose malabsorption would become smaller if it were regarded not as an abnormality but as a physiological process that opens up an option for dietary adjustment.

How differently the numbers can turn out is shown by a comparison with an Austrian outpatient analysis: there, at 25 grams of fructose, only 11.1 percent of 306 referred adults counted as fructose malabsorbers. The two values do not contradict each other, they come from different populations and different definitions. Anyone who forms a third, smooth figure out of them loses exactly this information.

Fructose also has a second building site that has nothing to do with the gut, namely the liver. Anyone interested in that will find it in the article on fructose, liver and metabolism. For this text it stays with the statement that gut tolerance and liver metabolism are two separate questions.

The hydrogen breath test: what it can do, how you prepare, when it goes wide of the mark

The breath test has something compelling about it. You blow into a tube, a number appears, and the number feels objective. Many people come into the consultation with exactly this feeling: at last something tangible.

The test is a good procedure too. It is not invasive, it is inexpensive, and it answers a real question. It just answers a narrower question than most reports suggest.

The principle in three sentences

Human cells do not produce hydrogen. Bacteria do. What is not taken up in the small intestine reaches the large bowel, is fermented there, and part of the gas produced passes into the blood and is breathed out through the lungs. There it can be measured. The Rome consensus conference of 2009 formally recorded this principle and at the same time named the considerable methodological differences between centres as the core problem.

Doses and thresholds

The preparation, and why it decides the result

A badly prepared breath test is worse than no breath test, because it produces a number that then justifies a diet for years. The European guideline is unusually concrete here.

Why the test can be negative even though you react

Here lies the first blind spot. Not every person produces hydrogen. The European guideline records in statement 2.14 that at least 10 percent of adults do not form enough hydrogen for a usable result. In children up to 10 to 15 percent are reported.

The second reason is the methane producers. In some bowels there are archaea that immediately process the hydrogen produced and turn it into methane. Anyone who measures only hydrogen then sees nothing, although plenty is being fermented.

The European guideline urges balance at this point: the additional methane measurement can be useful when hydrogen excretion is low, but the sensitivity gained by it is bought with a lower specificity. So you see more, and part of that is noise.

Anyone who produces a lot of methane often has a symptom pattern all of their own, by the way, with constipation instead of diarrhoea. That is a chapter of its own and is set out in the article on methane producers and intestinal methanogen overgrowth.

Why the test can be positive without it meaning anything

There is a second reason for false positive findings, and clinically it is the more important one. A review by Montalto and colleagues points out that with a small intestinal bacterial overgrowth the large amount of bacteria can ferment any sugar unspecifically. Then hydrogen rises with every challenge, and out comes the misleading diagnosis of a sugar malabsorption where there is actually a colonisation problem. Anyone who no longer tolerates anything should therefore have a small intestinal bacterial overgrowth considered instead of lengthening the list of forbidden sugars further. The American guideline on this topic also defines the breath test criteria as a separate matter accordingly.

Genetic test or breath test

This question comes up often, and it has a clear structure. The LCT genotype describes a disposition. It says something about whether your lactase activity has probably fallen in adulthood. It says nothing about how you are today, and it does not capture a secondary lactase deficiency at all.

The European guideline therefore carries the polymorphism only in the background text and does not recommend the genetic test as a routine clinical procedure. Added to that are the two findings from above: the known variants do not fully explain the frequencies outside Northern Europe, and in the UK Biobank the genotype was only weakly connected with milk consumption.

Sorbitol and the other sugar alcohols: the quiet source

There is a constellation that turns up regularly in the consultation and is almost never in the food diary. Someone eats consciously, avoids sugar, looks after their teeth, chews several sugar free chewing gums across the day, takes sugar free throat sweets and uses a mouth rinse. And is puzzled by loose stools in the afternoon.

Sugar alcohols are the blind spot of this field, because they do not look like sugar and are not called sugar.

The group at a glance

The polyols include sorbitol with the number E420, mannitol as E421, isomalt as E953, maltitol as E965, xylitol as E967 and erythritol as E968. They taste sweet, deliver less energy than table sugar, and they are barely fermented to acid by the caries bacteria in the mouth. Replacing sugar with sugar substitutes can therefore contribute to the maintenance of tooth mineralisation. That is why they are so widespread in chewing gum and dental care.

Their shared feature from the gut point of view: most of them are taken up only slowly and incompletely. One exception is erythritol. It is taken up for the most part already in the small intestine and excreted unchanged in the urine, so it barely reaches the large bowel. In a controlled study of twelve healthy men, about 78 percent of the amount ingested was recovered in the urine, and signs of gastrointestinal intolerance did not occur. From the gut point of view erythritol is therefore usually tolerated better than sorbitol or maltitol.

Better tolerated by the gut does not mean harmless, though. A group around Marco Witkowski and Stanley Hazen found in 2023, in three cohorts of 1,157, 2,149 and 833 people attending for cardiac assessment, an association between erythritol levels in the blood and major adverse cardiovascular events. The adjusted hazard ratio comparing the highest with the lowest quartile was 1.80 and 2.21. An enhanced reactivity of the platelets was seen in cell culture and in the animal model, in humans this mechanism is not established. These are observational data, not a proven cause and effect, and the authors themselves call for studies on long term safety. For this text it means: with the sugar alcohols too, the question remains one of quantity rather than a free pass.

As a natural occurrence you find sorbitol above all in stone fruit and pome fruit, especially in plum, pear, apple and apricot. As an addition it sits in sugar free sweets and chewing gums, in cough sweets, in many products with the note no added sugar, and in toothpaste and mouth rinse.

There is a practical marker for this that you can read on every packet. In the EU a food to which more than ten percent polyols have been added has to carry the note that it can have a laxative effect when consumed in excess. This sentence is not a platitude, it is a statement of quantity in words. For naturally contained sorbitol, for example in prunes, this note does not appear. The amount can still be considerable.

The closing sentence of this old work is remarkable. Back in 1985 the authors pointed out that diet products with sorbitol are very popular and that the complaints they cause can lead to extensive diagnostics and in the end to a lifelong diagnosis of irritable bowel syndrome.

The review by Montalto places sorbitol soberly: uptake is dependent on dose and concentration and depends on the available absorption surface. A malabsorption finding is expressly not an expression of a specific intestinal damage. So it is not proof of disease, it is a question of quantity.

Why pear and prune are a double hit

Now the circle closes back to the fructose section. Sorbitol can further worsen the uptake of free fructose. And of all things, the fruits with the highest sorbitol content often also have an unfavourable ratio of fructose to glucose. Pear, apple and prune therefore bring both with them.

That explains one of the most common everyday questions on this topic, namely why someone tolerates apple badly and banana well. It is not down to the fruit as such, but to the combination of fructose excess and sorbitol.

Anyone who wants to take the systematic route at this point ends up at the FODMAP concept, in which polyols are the P. How the three phase approach works there is set out in the article on the FODMAP diet and its correct use. And when diarrhoea is in the foreground, it is worth first taking a look at the overlooked causes of chronic diarrhoea, because sorbitol is only one of many.

Why self tests and IgG panels so often lead in the wrong direction

There is a feeling I can understand very well. You have had complaints for years, nobody finds anything, and then someone offers you a test that checks 90 foods at once and gives you back a colour coded list. Red, yellow, green. Order at last.

The problem is not the longing for order. The problem is that this particular test cannot answer the question you are asking.

What an IgG test actually measures

A task force of the European Academy of Allergy and Clinical Immunology systematically assessed the data on food specific IgG4. Their findings are unusually clear for an expert panel.

The Canadian professional society for allergology added two arguments in a position paper in 2012 that often move more in everyday life than the immunology does. First the costs, which frequently lie in the range of 400 to 700 dollars. Second a safety argument that matters more to me than the money: a person with a genuine IgE mediated food allergy and a risk of anaphylaxis may well have an unremarkable specific IgG against their allergen. On the basis of such a result they could be wrongly encouraged to eat this food again. In children the risk of exclusion diets with consequences for growth and nutrient supply is added.

The German professional society for allergology and clinical immunology holds the same line and describes the testing of food specific IgG in its guideline on food allergy as unsuitable. With that, the European, the Canadian, the American and the German professional societies say the same thing on this question. That happens rarely.

One distinction matters to me at this point. The criticism is directed against a test procedure, not against people and not against providers. Anyone who has had such a test done has done nothing wrong. And anyone who felt better after the diet that followed is not imagining it. It is just that it is then very probably due to something other than the measured IgG.

Why leaving things out often changes something anyway

Exactly here it becomes interesting, and here lies the part that makes the matter fair. There is a very elegant piece of evidence that a dietary change can ease complaints without the suspected component having been the reason.

The German position paper on non coeliac gluten and wheat sensitivity names exactly this methodological core problem: high nocebo and placebo effects in food challenges. It also points out that on a gluten reduced diet the whole food selection often changes, for example with more vegetables and more soluble fibre, so that digestion and transit time shift independently of the gluten avoidance. Anyone who wants to go deeper will find that in detail in the article on gluten and gliadin without coeliac disease.

One more word on stool tests with large bacterial panels, which are often offered alongside in this context. They answer a different question from the one about sugar tolerance, and how much they can say depends very much on what exactly is measured. The article on stool testing, PCR and dysbiosis diagnostics follows that up in detail.

What can change something in practice, and why avoidance is a beginning and not a goal

If you have read this far, you know the three mechanisms and the limits of the tests. Now comes the part that arrives in everyday life. It consists of four building blocks, and none of them is a ban.

First: amount instead of ban

The most robust figure of this whole field is already above. 12 to 15 grams of lactose are tolerated by most people with lactose malabsorption. Above 12 grams complaints become clearer, from 24 grams noticeable, and 50 grams lead to something in almost everyone. If you spread 24 grams across the day, that is generally tolerated.

The NIH consensus conference of 2010 recorded two things from the same data base: that the actual frequency of lactose intolerance in the population cannot be reliably determined, and that a complete avoidance of dairy products can have disadvantages for the nutrient supply. On this, the AHRQ report found weak indications from 55 observational studies with 223,336 people that a low milk consumption can go along with a higher fracture risk. Weak means weak, that is not an argument for milk. It is an argument against a hasty complete cutting out.

How much dairy products contribute overall and where the debate about that stands is a topic of its own. It is set out in the article on the question of whether dairy products are healthy or not.

Second: spreading and company

Two simple adjusting screws shift tolerance without anything being cut out. The first is the spreading across the day, because the uptake capacity is a rate and not a daily total. The second is the company of other foods, because a meal slows the passage and gives the small intestine more time.

With fructose the glucose company is added. People with fructose malabsorption show raised breath hydrogen and complaints after pure fructose, but report markedly less when fructose is given together with glucose. This favourable effect is still not explained today, but it is well described.

With lactose there is another everyday point that is often overlooked. A systematic review on pre and probiotics records that in many people with lactose malabsorption the clinical signs are absent after usual amounts of milk, and especially after products such as yoghurt and cheese, in which part of the lactose has already been broken down by living bacteria or by ripening.

Third: the microbiome adapts

The same idea was later checked in a targeted way. In a randomised, double blind study of 85 people with lactose intolerance, a special galactooligosaccharide was given over 35 days, after which dairy products were reintroduced and followed up for 30 days. 50 percent of those who had abdominal pain at the start reported none at the end. After the reintroduction, those treated stated six times more often that they tolerated lactose. The accompanying microbiome analysis showed a rise in bifidobacteria in 27 of 30 of those treated.

And now the placing that belongs with it, because otherwise nobody supplies it: some of the authors of this study were employed by the manufacturer of the preparation. That does not devalue the work, but it belongs to be named. The AHRQ review rated the evidence for colonic adaptation and for probiotics overall as insufficient. A systematic review on pre and probiotics found a strongly strain and dose dependent efficacy with considerable heterogeneity and only a single prebiotic study.

Fourth: enzyme preparations, viewed soberly

Lactase preparations are meant to support the digestion of milk sugar. Whether that also reliably improves the complaints is less clear than the impression on the shop shelf suggests. The systematic review by the AHRQ group describes the included studies as so differently designed that the results could not be pooled, and it found no demonstrated advantage of a nearly lactose free solution over more than twelve grams. In infant colic, a meta-analysis of two randomised trials came to no difference in crying time or fussing time. A dose deliberately does not stand here, because that belongs in a conversation and not in a blog text.

Digestive enzymes in general are a chapter of their own and are set out in the article on when digestive enzymes make sense. One order matters to me in this, which I keep to almost always in practice: before I think about enzymes, I ask whether there is enough stomach acid up top at all. A digestion that starts too quietly in the stomach is hard to make up for further down with enzymes. Why that is so is set out in the article on low stomach acid and betaine HCl.

Fifth, and most important: avoidance has an expiry date

The German language review in the Bundesgesundheitsblatt puts this so clearly that it can stand here almost word for word: extensive dietary restrictions with regard to the sugar in question should, with the exception of lactose, not be maintained over a longer period. What is recommended instead is a three or four step approach.

A review by the same group sums up the data on FODMAP reduction like this: supported in the short term by several meta-analyses, in the long term at least 50 percent experience relief after uncontrolled follow ups once they have gone through all three phases. And there are indications that less elaborate approaches such as a usual irritable bowel diet or a low lactose diet can contribute a similar amount. The three phase approach is set out in detail in the article on the FODMAP diet, and what the microbiome needs in the reintroduction phase is set out with the prebiotics and resistant starch. On the question of which bacterial preparations make sense at all, there is the article on probiotics in spore form or capsule.

Three groups need accompaniment rather than self experiment: pregnant women, breastfeeding women and children. In them a longer restriction can lead to gaps more quickly than it feels. And anyone who leaves out dairy products permanently should cover calcium and vitamin D deliberately from other sources and have that worked through medically once.

When an intolerance appears newly: the question that is rarely asked

There is a constellation that always makes me attentive. Someone has drunk milk for forty years without ever thinking about it. And for the past half year it no longer works.

This story has an important property: the genome is hardly the reason, because the gene switch for lactase does not change any more in adulthood. What can change is the small intestinal mucosa on which the enzyme sits, and the amount that still gets through.

Secondary lactase deficiency

The lactase sits right at the outside on the tips of the small intestinal villi. That makes it vulnerable. Everything that damages the mucosa or alters the passage can hit it. The review in Gut expressly names as causes of a lactose malabsorption, alongside the primary genetic one, also the secondary lactase deficiency through infections and other conditions that affect the mucosal integrity of the small intestine.

The adjusting screw that has nothing to do with the sugar

There is one more factor that rarely comes up in advice conversations and is nevertheless one of the best explanations for field 3 of the four field table: visceral perception. The same amount of gas in the bowel does not lead to the same experience in all people.

The review in Gut records that people with visceral hypersensitivity in connection with anxiety or irritable bowel syndrome have a raised risk of complaints independently of lactose digestion. The fructose review says the same for fructose: in people with functional bowel disorders the probability is higher that complaints come out of the same amount.

That is expressly not a devaluation and not an indication that something is imagined. On the contrary: in the blinded study that showed that only 47.5 percent could correctly assign lactose as the trigger, not a single psychological feature was found that distinguished people with a nocebo reaction from the others. So it is not down to personality. It is down to the assignment being objectively difficult.

Visceral perception is moreover something that can be worked on. The route leads through the nervous system and not through the menu, and it is set out in the article on the gut brain axis and vagus stimulation.

And if all three tests are unremarkable

Then the matter is not settled. A paediatric review on carbohydrate intolerances lists, alongside lactose, fructose and sorbitol, further conditions, among them sucrase isomaltase deficiency, glucose galactose malabsorption and trehalose intolerance. Some of them are rare, but they exist.

And more common than all of that is the simple sum: it does not have to be a single sugar. The review in Gut records that the clinical result of the treatments in lactose remains modest, because lactose is only one of several poorly absorbed carbohydrates that can cause complaints through the same mechanism. That is exactly the reason why this article treats the three together and not separately. Anyone who suffers above all from air will find the mechanisms behind it in the article on bloating and its causes.

And at the end of this section once more the sentence from right at the top, because it is the most important of the text: blood in the stool, black tarry stool, unintended weight loss, fever, night time complaints, vomiting, difficulty swallowing, a new persistent change in bowel habit from around 45 to 50 years of age, anaemia or a family burden of bowel cancer or inflammatory bowel disease belong in medical assessment. Not interpreted, not left aside, not answered with a diet.

Frequently asked questions

Where this topic connects to the rest of the body

A carbohydrate intolerance rarely stands on its own. It hangs on the dietary history, on the uptake of micronutrients, on sleep quality and on how a person deals with a long list of bans.