The Latest Lactose Intolerance Research
Updated December 21, 1998
Check here regularly for summaries of recent articles on lactose intolerance from the top
medical journals, along with my commentary.
"Allergy to supplemental lactase enzyme," by Karen E. Binkley. Journal of Allergy and Clinical Immunology, Vol. 97, No. 8 (June 1996), pp. 1414-16.
Allergy testing showed that the man showed a trace reaction to Aspergillus species, the fungus that is used commercially to manufacture the lactase enzyme.
Notice, however, that the man wasn't actually allergic to lactase; he was allergic to the fungus that manufactures it, a very different thing. Other patients known to be allergic to Aspergillus were also tested. Only one of them showed any signs of any reaction and even that was labeled "equivocal."
Basically, then, if you already know you are allergic to fungus, molds, or the like, you may want to be cautious about taking lactase tablets. For the vast majority of people, however, lactase pills remain a safe and useful product.
"Does Lactose Maldigestion Really Play a Role in the Irritable Bowel?" by Beth A. Tolliver, et. al. Journal of Clinical Gastroenterology, Vol. 23, No. 1, pp. 15-17.
In interviews conducted an average of 41 months later the subjects "showed no significant differences in abdominal pain, altered bowel habits, bloating or distention, mucus, and relief with defecation." However, "all lactose-maldigesting IBS subjects reported that identifying lactose maldigestion influenced their awareness of lactose-symptom relationships and brought some improvement."
The authors concluded that, "Nevertheless, lactose maldigestion was not the cause of IBS symptoms."
After the experiment, five out of six were avoiding lactose-containing foods, and, not too surprisingly, the same percentage experienced some improvement in symptoms.
None of these findings was very strong, so those of you with IBS may or may not find relief by going off milk. If you haven't already tried doing so, however, it may be worth the time and trouble.
This experiment also shows the necessity for test groups that mirror the overall prevalence of LI in the population. The non-LI IBS group was 89.5% white. Only 61.7% of the IBSLI group was.
"A Comparison of Symptoms After the Consumption of Milk or Lactose-Hydrolyzed Milk by People with Self-Reported Severe Lactose Intolerance," by Fabrizis L. Suarez, M. D., Dennis A. Savaiano, Ph.D., and Michael D. Levitt, M.D. The New England Journal of Medicine, Vol. 333, No. 1 (July 6, 1995), pp. 1-4.
This is the biggie, the one you heard about in newspapers, on CNN, on NPR, in a thousand health-oriented newsletters.
And what you heard was that LI wasn't actually a problem; in fact, that it was mostly all in our heads.
Here's their bottom line: "People who identify themselves as severely lactose-intolerant may mistakenly attribute a variety of abdominal
symptoms to lactose intolerance. When lactose intake is limited to the equivalent of 240 ml [8
ounces of milk, containing 12 grams of lactose] a day, symptoms are likely to be negligible and the use of
lactose-digestive aids unnecessary."
Each drank an 8 oz. glass of milk with breakfast every day for a week. Some of the milk had had lactase added to remove all of the lactose (no measurable residual).
Symptoms reported ranged from trivial on down, and no statistically significant differences occurred between the regular and the lactose-reduced milks.
The good is that this was a carefully devised experiment, designed to eliminate any of the factors in previous experiments which could have confused the results.
In particular it was a double-blind experiment, meaning that neither the doctors nor the subjects knew whether any particular glass of milk contained lactose. We know from other experiments that the way people respond to lactose has a large psychological component. And even the most careful, well-meaning experimenters can skew results inadvertently by conveying their expectations to the subjects.
Double-blind experiments aren't always possible in medical testing, but it has to be argued that no non-double blind experiment can really prove anything.
It's worthwhile to quote from their conclusions: "Although many patients are certain that they can link the ingestion of various foods to subsequent abdominal symptoms, it is extremely difficult to pinpoint accurately which, if any, of the constituents of the diet cause abdominal distress.... Our finding that 240 ml of milk was not associated with a significant increase in the severity of bloating, abdominal pain, or flatus suggests that people frequently misattribute a variety of abdominal symptoms to lactose intolerance. A larger study might have uncovered some differences. [italics added] Nevertheless, the symptoms that might be caused by lactose are unlikely to be substantial..."
The bad about this experiment is the same one that almost every LI experiment has. There are simply too few subjects involved to make a useful sample of the LI population.
Look, sampling works: polls do predict results; the Nielsen's are an accurate reflection of tv ratings. And a sample of 1,000, even of 300, is large enough, if that sample is a good random sampling of the population as a whole.
But 30 people? No way. A sample that small could be skewed in any number of ways. And of these particular 30 people, nine turned out not to even be LI when tested. So only 21 actual LI individuals were in the experiment (one black, 7 Asian, 8 white, 5 Hispanic).
We also are never told what kind of symptoms they had after taking the 15 g. of lactose for the test. If they didn't have symptoms when taking it straight, they surely were not likely to have symptoms when taking lactose with a meal, a known method of reducing symptoms (see the discussion in the Hertzler article below).
The power of the New England Journal of Medicine (NEJM) is such that this article generated a million additional stories. None of the ones I saw or heard noted that an editorial in the very same issue was much more cautious in its conclusions: "Some physicians may recommend that patients... discontinue the use of digestive aids. Others may recommend that symptoms still be the guide.... Nevertheless, the well-designed evaluation by Suarez, et. al. represents a substantial contribution. The interplay of mind and body is critical in the development of abdominal symptoms. More objective research onto the cause of abdominal symptoms should be enlightening."
That's the real meaning of the experiment. It is a strong suggestion, not a final answer.
P.S. For completists, letters in response to the article appeared in the Nov. 16, 1995 issue of the NEJM (Vol. 333, No. 20, pp. 1358-9). I incorporated a couple of their points in my commentary.
"Tolerance to small amounts of lactose in lactose maldigesters," by Tuula H. Vesa, Riita A. Korpela, & Timo Sahi. American Journal of Clinical Nutrition, Vol. 64, pp. 197-201.
Maldigesters reported more bloating, but there was no difference in the mean severity of symptoms and one-third of the maldigesters did not report any symptoms at all. In fact, the same number of maldigesters reported symptoms after drinking the lactose-free milk as the highest lactose milk.
Bottom line: "Our results seem to indicate that an entirely lactose-free diet is only rarely necessary."
On the flip side, using 39 maldigesters is not much more statistically significant than using 21. And Finns are probably even less representative of the American population than the subjects in the Savaiano and Levitt study.
So why even discuss it? Because of the editorial which accompanied it (next).
"Abdominal symptoms and lactose: the discrepancy between patients' claims and the results of blinded trials," an editorial by Fabrizis Suarez and Michael D. Levitt. American Journal of Clinical Nutrition, Vol. 64, pp. 251-2.
"Well-controlled trials have shown that the vast majority of lactose malabsorbers report no significant difference in bloating, abdominal discomfort, and diarrhea with ingestion of one cup of conventional milk daily....
"Thus, it appears that subjects who believe themselves to be intolerant to small doses of lactose may in reality have an underlying symptomatic state, IBS, for lack of a better term, which they misattribute to lactose intolerance....
"However, the concept that modest doses of lactose can be tolerated with negligible symptoms is anathema to most individuals who perceive themselves to be lactose intolerant....
"The conventional response to these individuals is that controlled trials do not lie and that the lactose-induced symptoms in these individuals are largely 'psychologic' rather than 'physiologic' in origin....
"Is it possible that lactose malabsorption actually induces symptoms in these subjects, but the methodology required for a blinded study somehow alters a subject's ability to accurately perceive symptoms? For example, does the need to focus on and grade symptoms perturb the normal, delicate mind-body interaction that results in the perception of mild abdominal symptoms, with a resulting underestimation of symptoms during the active treatment period?...
"[I]t would be of interest to determine whether the blinded techniques required to evaluate lactose intolerance symptoms possibly represent a biological uncertainly principle."
Well, if it's not mind-body interaction, then what could it be? Well, I've said it before, and it still seems to apply: Sirs, your test subjects are not a good random sampling of the LI population of this country. And they are almost certainly not representative of those who are most intolerant, the ones who show the most symptoms to the smallest amounts of lactose.
You have no way of knowing just how many people are in that group, because no one has ever figured out a way of measuring it. You have no idea what that group's make-up is, because no one has ever studied it for the population as a whole. They are the ones most likely to report problems, to refrain from drinking milk, to use lactase products, to be vocal about the subject. Are they being captured in studies? You don't know, because no one does.
Perhaps controlled trials do not lie, but their results can certainly be misleading. Perhaps most patients are confident that they can identify the foods that aggravate their problem because they really can. Nobody now knows what the lactase curve, how many people have how much remaining lactase, looks like. Nobody now knows what the milk-drinking curve, how much milk can be taken without symptoms, looks like. Nobody even knows whether the two curves match, or even correlate well for the vast LI population, a population whose ethnic background, age distribution, and even gender cannot be said except in the broadest of generalities.
Until a true broad-based study of the LI population is done, all such small-scale studies such as yours are groping in the dark, rendered moot by the fact that they are not generalizable to the larger population.
Maybe when and if that day occurs, it will turn out that you were absolutely right in your evaluation. For now, you are in the position of being the blind men feeling the elephant. (You might call that a mind-body problem as well.) Your looking at the elephant's tail and trunk, tiny slivers of the whole, leaves the rest of us comprising the elephant unaccounted for.
More work needs to be done before anybody can say anything definitive about lactose intolerance, let alone that it is purely or even mainly psychologic in origin.
"Lactose malabsorption as influenced by chocolate milk, skim milk, sucrose, whole milk, and lactic cultures," by N. Dehkordi, M.S., R.D., D. R. Rao, Ph.D., A. P. Warren, M.S., R.D., and C. B. Chawan, Ph.D. Journal of the American Dietetic Association, Vol. 95, No. 4 (April 1995), pp. 484-6.
I have a whole chapter in my book (Milk Is Not for Every Body) devoted to the various ways you can better tolerate lactose, based on a bunch of different journal articles. These researchers read all those articles and decided to put them once again to the test and see what happened.
Note, however, that this was not a double-blind experiment; in fact, not even single blind. The subjects knew exactly what they were having, and that could easily affect the results.
And the sample sizes here are even smaller, much too small to be meaningful, even if the results do correlate with previous studies. But if anything can be said, it's this: having milk with a meal reduces symptoms.
"How much lactose is low lactose?" by Steven R. Hertzler, R.D., Bao-Chau L. Huynh; and Dennis A. Savaiano, Ph.D. Journal of the American Dietetic Association, Vol. 96, No. 3 (March 1996), pp. 243-6.
While the higher amounts of lactose (6 grams and above) significantly increased breath hydrogen production (the standard test for LI), symptoms of abdominal pain did not start to increase until after the 12 gram dose, and flatulence increased only after the 20 gram dose. No differences in severity of diarrhea were noted after any of the five doses.
Their conclusions: "lactose maldigesters may be able to tolerate foods containing 6 g lactose or less per serving, such as hard cheeses and small servings... of milk.... [L]arger doses of lactose may still be tolerated, particularly if they are consumed with other foods.
The small number of subjects, as always, is a real problem. Ethnic backgrounds are not identified.
The point of this experiment is a bit hidden. They're really arguing that the older, non-double blind, experiments weren't really very good. If your subjects never know whether they are drinking lactose or not, they never get serious symptoms, even after drinking 20 grams of lactose. In this way, this is to be taken as a confirmation of the Suarez experiment.
However, they don't go quite as far in suggesting that our problems are psychological, only that taking people entirely off milk products when they are LI may not be necessary. Have small amounts, with food, and you should be okay. Who can argue with that for most people?