IJE Advance Access published online on August 24, 2008
International Journal of Epidemiology, doi:10.1093/ije/dyn169
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Hypothesis: Is Lactose a Dietary Risk Factor for Ischaemic Heart Disease?i
308 Cricklewood Lane London NW2 UK
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionAcknowledgementsReferences |
|---|
Summary The prevalence of primary adult lactose malabsorption (LM) in 23 ethnic groups was matched with national data on milk consumption and mortality rates from ischaemic heart disease (HD). In 6 other ethnic groups prevalence of LM was related to unquantified assessments of milk consumption and frequency of IHD.
On the available data, populations with a prevalence of LM over 30%, and whose consumption of milk is low or is largely in low lactose form, have a lower risk of IHD mortality than populations with a prevalence of LM under 30% and a high milk consumption. There is evidence against attributing these findings to genetic linkage between susceptibility to IHD mortality and persistent lactose absorption, or to differences in socio-economic development, cigarette consumption or intake of animal fats. The findings are compatible with an hypothesis that, if the correlation reported previously between milk consumption and IHD mortality is causal, lactose could be the responsible dietary factor.
Accepted 23 July 2008
In a previous study on data from 43 countries, a strong correlation was found between national per caput consumption of total milk (milk and milk products, excluding butter, expressed in terms of fresh milk) and mortality rates from ischaemic heart disease (IHD).1 The amount of milk consumed tends to be related to the prevalence of the persistence into adult life of lactose absorption (although this is generally considered to be an inherited characteristic, and not due to induction of intestinal lactase activity by ingestion of milk).2,3 It seemed likely, therefore, that IHD mortality rates would correlate positively with prevalences of persistent lactose absorption, and negatively with prevalences of primary adult lactose malabsorption (LM), which is the usual way in which the lactose absorption status of a population is expressed. Hence the available data were examined for relationships between milk consumption, prevalence of LM and IHD mortality.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionAcknowledgementsReferences |
|---|
The published data on ethnic prevalences of LM3 were matched where possible with national per caput consumption of total milk4 and IHD mortality rates.5 Matching proved possible for 19 of the 43 countries of the previous study,1 and for 4 additional countries6 (which were omitted from the previous study because their data on IHD mortality rates were classified by WHO as ‘incomplete or of unknown reliability’) to give 23 matched ethnic groups and countries.
The prevalences of LM were obtained from the comprehensive table of published results compiled by Simoons.3 Weighted averages were calculated for the ethnic groups on whom there were more than one reported study. Reports on or including children were excluded.
The matched data were divided into 2 groups, designated A and B, with prevalences of LM respectively of over and under 30%. This was based on Simoon's classification of prevalence of LM as >60% (high), 30-60% (intermediate) and <30% (low), but the high and intermediate groups were combined to make Group A because of their small sizes.
Prevalence of LM in 6 other ethnic groups was related to reported but unquantified assessments of milk consumption and frequency of IHD.
The term ‘hypolactasia’ was avoided because many of the investigations to determine the prevalence of LM relied on lactose tolerance testing, without disaccharidase assay of small intestinal biopsy.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionAcknowledgementsReferences |
|---|
The matched data divided into Groups A and B are shown in Tables 1 and 2. A statistical analysis of the results is given in Table 3.
|
|
|
Groups A and B were roughly comparable with regard to size (10 and 13 respectively) and average number of subjects investigated for ethnic prevalence of LM (172 and 203). The groups differed geographically, racially and socially. The ethnic groups in Group B, except for the Spaniards, consisted of northern Europeans and their overseas descendants, who were unrepresented in Group A. Group B contained only developed countries, whereas Group A included some developing countries.
Group A had the lower mean per caput milk consumption and IHD mortality rate. The mean annual mortality rate for Groups A and B combined was 152 per 100 000, and in Group A only Israel had a mortality rate above this level.
In Groups A and B combined there was a strong positive correlation between milk consumption and IHD mortality rates, and there were strong negative correlations between prevalence of LM and both milk consumption and IHD mortality rates.
Table 4 shows the available data on the Bantu of South Africa, Chinese, Eskimos of Greenland, Nilotes and Nilo-Hamites, Punjabis and USA Blacks.
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Milk consumption tends to be inversely related to prevalence of LM2,3 The strong negative correlation found in the present study between national per caput milk consumption and ethnic prevalence of LM therefore indicates that the countries were appropriately matched with the ethnic groups.
The strong positive correlation between milk consumption and IHD mortality rates in the 23 countries is a finding that was expected from the previous study on data from 43 countries.1
The new finding is that, on the data available, high and intermediate prevalence of LM (>30%) is generally associated with a lower IHD mortality rate (Table 1) and probably prevalence (Table 4) than low prevalence of LM (<30%)(Table 2). The explanation for this observation could be that (1) persistence of lactose absorption in adults is genetically linked with a predisposition to IHD, (2) low prevalence of LM is associated coincidentally with environmental factors that favour IHD, (3) overconsumption of milk is an aetiological factor in IHD, or (4) lactose is atherogenic or thrombogenic.
- (1) Is persistence of lactose absorption genetically linked with predisposition to IHD? Persistence of lactose absorption into adult life is inherited, probably as an autosomal dominant character, and populations with a low prevalence of LM have probably evolved mainly by natural selection in dairying areas.2,3 It is conceivable, therefore, that the association of low prevalence of LM with high IHD mortality may be due to genetic linkage between persistent lactose absorption and a predisposition to coronary disease. However, the low IHD mortality rate in Spain (the lowest in Group B) despite a prevalence of LM in Spaniards of only 15% (Table 2) is evidence against such linkage, and suggests that the liability of north-western Europeans18 and their overseas descendants to IHD is either inherited separately from their characteristic of persistent lactose absorption, or is environmental or dietary.
- (2) Is low prevalence of LM associated coincidentally with environmental factors that favour IHD? This possibility is suggested by the countries in Group B, all of which are fully developed, unlike those in Group A. It is supported by the occurrence of high IHD mortality rates in 2 populations of predominantly lactose malabsorbers living in developed countries – USA Blacks (Table 4) and Israeli Jews (Table 1). However Japan, ethnic prevalence of LM of 78%) has a low IHD mortality rate (Table 1) despite advanced socio-economic development and high cigarette consumption;19 and, conversely, in the Punjab (ethnic prevalence of LM of 17%) IHD is not uncommon (Table 4) and affects non-smoking women, living sheltered lives, in the same proportion as females in Western countries.17
- (3) Is overconsumption of milk an aetiological factor in IHD? In the previous study per caput consumption of milk correlated more strongly with IHD mortality rates than per caput consumption of meat, fats and oils, and sugar (in 43 countries), butter (in 21 countries) and coffee and cigarettes (in 12 countries).1 The present findings indicate that national IHD mortality rates could be primarily correlated either positively with milk consumption or negatively with prevalence of LM. Similarly, the low risk of IHD in China and among Greenland Eskimos could be primarily related to low milk consumption or high prevalence of LM (Table 4). Data suggesting that of these 2 possibilities, milk consumption is more likely to be the primary association are the low IHD mortality rate in Spain, where milk consumption (the lowest in Group B) is low for predominantly lactose absorbers (Table 2), and the high IHD mortality rates in USA Blacks (Table 4) and Israelis (Table 1), who have high milk consumption for predominantly lactose malabsorbers.
- (2) Is low prevalence of LM associated coincidentally with environmental factors that favour IHD? This possibility is suggested by the countries in Group B, all of which are fully developed, unlike those in Group A. It is supported by the occurrence of high IHD mortality rates in 2 populations of predominantly lactose malabsorbers living in developed countries – USA Blacks (Table 4) and Israeli Jews (Table 1). However Japan, ethnic prevalence of LM of 78%) has a low IHD mortality rate (Table 1) despite advanced socio-economic development and high cigarette consumption;19 and, conversely, in the Punjab (ethnic prevalence of LM of 17%) IHD is not uncommon (Table 4) and affects non-smoking women, living sheltered lives, in the same proportion as females in Western countries.17
A suspicion that overconsumption of milk may have an aetiological role in IHD has been expressed in a number of reports, whose authors have considered possible causes to be dairy fats,20 the large size of fat particles in unboiled milk, 21allergenicity of milk proteins, 22–25 the content of the enzyme xanthine oxidase,26 the contribution to calorie overnutrition and lactose.1 A factor other than animal fats is suggested by the diet of the Greenland Eskimos, among whom IHD is rare.11 Dairy products are very scarce in their diet, but the content of total fats and of saturated fats and monounsaturated fats together is about the same as that of the Danes.12
- (4) Is lactose responsible? If sucrose is an IHD dietary risk factor,27 lactose could be as well. Both sugars yield glucose on hydrolysis, and galactose like glucose and fructose stimulates insulin secretion,28 so that lactose might also eventually impair glucose tolerance.27 If lactose is atherogenic or thrombogenic, populations with a high consumption of milk would be protected from this dietary risk factor by a combination of high or intermediate prevalence of LM and the intake of milk largely in fermented or other low lactose form. This might explain, at least partly, the low risk of IHD in the Masai14 and the Samburu,15 urban Blacks of South Africa,10 Greeks (Table 1) and some northern Indians.21
The Masai14 and the Samburu15 have a high milk consumption. Their prevalences of LM do not appear to have been investigated, but the findings in other Nilo-Hamites and in Nilotes (Table 4) suggest that their prevalences are likely to be intermediate or high. Moerover, milk is consumed by the Masai largely in fermented form,26,29,30 which might be a factor in their low levels of blood cholesterol,30 (and presumably prevents symptoms of lactose intolerance in the lactose malabsorbers).
The Bantu of South Africa have a high prevalence of LM (Table 4). The urban Blacks consume milk, but sour milk is more popular than fresh milk for drinking.31
In Greece milk consumption is high and the IHD mortality rate relatively low (Table 1). The prevalence of LM is intermediate and, except in cities and large towns, adults do not drink much milk, most of which is processed into cheese or other dairy products.3
Northern Indians differ from southern Indians in being at less risk of IHD,21 having generally a higher prevalence of LM,3 and consuming milk as fermented products or boiled.21
Thus it appears that, in comparison with populations who have a low prevalence of LM and a high milk consumption, populations with high and intermediate prevalence of LM have a low risk of IHD mortality if their consumption of milk is low or is largely in low lactose form. The findings do not seem to depend on genetic linkage between susceptibility to IHD and persistence of lactose absorption, since the IHD mortality rate may be high in a population with a high prevalence of LM if milk consumption is high for predominantly lactose malabsorbers (USA Blacks, Israeli Jews), and low in a population with a low prevalence of LM if milk consumption is low for predominantly lactose absorbers (Spaniards). Moreover, there is evidence against attributing the findings to differences in socio-economic development and cigarette consumption (Japanese, Punjabis) or in intake of animal fats (Greenland Eskimos). The findings are therefore compatible with an hypothesis that, if the correlation between milk consumption and IHD mortality is causal, lactose could be the responsible dietary factor. This hypothesis could be tested by a direct comparison of the IHD risk in lactose absorbers and malabsorbers, and possibly by the experimental investigation of the atherogenic and thrombogenic potentials of lactose.
|
Acknowledgements |
|---|
|
TopAbstractMethodsResultsDiscussionAcknowledgementsReferences |
|---|
The author thanks Dr. Keith Ball, for some helpful comments.
|
Notes |
|---|
iFirst published. Int J Epidemiol., September 1980; 9: 271 - 276
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionAcknowledgementsReferences |
|---|
1 Segall JJ. Is milk a coronary health hazard? Br J Prev Soc Med (1977) 31:81–85.[Web of Science][Medline]
2 Bayless TM, Paige DM, Ferry GD. Editorial: Lactose intolerance and milk drinking habits. Gastroenterology (1971) 60:605–08.[Web of Science][Medline]
3 Simoons FJ. The geographic hypothesis and lactose malabsorption. Amer J Dig Dis (1978) 23:963–80.[CrossRef][Medline]
4 United Nations. Statistical Yearbook. 1972 (1973) New York: UN.
5 United Nations. Demographic Yearbook. 1973 (1974) New York: UN.
6 United Nations. Demographic Yearbook. 1975 (1976) New York: UN.
7 McMichael J. Letter to the Editor: French wine and death certificates. Lancet (1979) 1:1086–87.[Medline]
8 Worcester N. Diets for people. In: China Now (1978) Nov/dec: 12–14.
9 Salaff JW. Mortality decline in the People's Republic of China and the United States. Population Studies (1973) 27:551–76.[CrossRef]
10 Sertel HC. The rarity of coronary heart disease in South African Blacks. S Afr Med J (1978) 54:99–105.[Web of Science][Medline]
11 Bang HO, Dyerberg J, Nielsen AB. Plasma lipid and lipoprotein pattern in Greenlandic west-coast Eskimos. Lancet (1971) I:1143–54.
12 Bang HO, Dyerberg J, Hjorne N. The composition of food consumed by Greenland Eskimos. Acta Med Scand (1976) 200:69–73.[Web of Science][Medline]
13 Atherosclerosis and Epidemiology Study Groups. Primary prevention of the atherosclerotic diseases. Circulation (1970) 42:A55–95.[Medline]
14 Mann GV, Shaffer RD, Anderson RS, Standstead HH. Cardiovascular disease in the Masai. J Atheroscler Res (1964) 4:289–312.[Web of Science][Medline]
15 Shaper AG. Cardiovascular studies in the Samburu tribe of Northern Kenya. Am Heart J (1962) 63:437–42.[CrossRef][Web of Science][Medline]
16 Rab SM, Baseer A. High intestinal lactase concentration in adult Pakistanis. Br Med J (1976) I:436.
17 Malhotra RP, Pathania NS. Some aetiological aspects of coronary heart disease. Br Med J (1958) I:528–31.
18 World Health Organisation Regional Office for Europe. Myocardial Infarction Community Registers (1976) Copenhagen: WHO. 59–64.
19 Beese DH. Tobacco Consumption in Various Countries (1972) London: Tobacco Research Council.
20 Briggs RD. Rubenberg ML, O’Neal RM, Thomas WA, Hartroft WS. Myocardial infarction in patients treated with Sippy and other high-milk diets. Circulation (1960) 21:538–42.
21 Malhotra SL. Protective role of dietary factors in coronary heart disease. Practitioner (1976) 217:929–34.[Web of Science][Medline]
22 Annand JC. Hypothesis: Heated milk protein and thrombosis. J Atheroscler Res. (1967) 7:797–801.[Web of Science]
23 Davies DF, Davies JR, Richards MA. Antibodies to reconstituted dried cow's milk protein in coronary heart disease. J Atheroscler Res (1969) 9:103–07.[Web of Science][Medline]
24 Annand JC. Further evidence in the case against heated milk protein. Atherosclerosis (1972) 15:129–33.[Medline]
25 Davies DF, Johnson AP, Rees BW, Elwood PC, Abernethy M. Food antibodies and myocardial infarction. Lancet (1974) I:1012–14.
26 Oster KA. Plasmalogen diseases: A new concept of the etiology of the atherosclerotic process. Am J Clin Res (1971) 2:30–35.
27 Yudkin J, Roddy J. Levels of dietary sucrose in patients with occlusive atherosclerotic disease. Lancet (1964) ii:6–8.
28 Keele CA, Neil E. Samson Wright's Applied Physiology (1971) London: Oxford University Press. 487.
29 Biss K, Ho K, Mikkelson B, Lewis L, Taylor CB. Some unique characteristics of the Masai of East Africa. N Engl J Med (1971) 284:694–99.[Web of Science][Medline]
30 Mann GV, Spoerry A. Studies of a surfactant and cholesterolaemia in the Masai. Am J Clin Nut (1974) 27:464–69.[Abstract]
31 Manning EB, Mann JI, Sophangisa E, Truswell AS. Dietary patterns in urbanised Blacks. S Afr Med J (1974) 48:485–98.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||