Aim: To test the correlation of per capita A1 beta-casein (A1/capita) and milk protein with: 1) ischaemic heart disease (IHD) mortality; 2) Type 1 (insulin-dependent) diabetes mellitus (DM-1) incidence.
Methods: A1/capita was estimated as the product of per capita cow milk and cream supply and its A1 beta-casein content (A1/beta) (calculated from herd tests and breed distribution, or from tests of commercial milk), then tested for correlation with: 1) IHD five years later in 1980, 1985, 1990 and 1995, in 20 countries which spent at least US $1000 (purchasing power parities) per capita in 1995 on healthcare; 2) DM-1 at age 0-14 years in 1990-4 (51 were surveyed by WHO DiaMond Project; 19 had A1 data). For comparison, we also correlated 77 food, and 110 nutritive supply FAO (Food and Agriculture Organization)-based measures, against IHD and DM-1.
Results: For IHD, cow milk proteins (A1/capita, r = 0.76, p <0.001; A1/capita including cheese, r = 0.66; milk protein r = 0.60, p = 0.005) had stronger positive correlations with IHD five years later, than fat supply variables, such as the atherogenic index (r = 0.50), and myristic, the 14-carbon saturated fat (r = 0.48, p <0.05). The Hegsted scores for estimating serum cholesterol (r = 0.42); saturated fat (r = 0.37); and total dairy fat (r = 0.31) were not significant for IHD in 1995. Across the 20 countries, a 1% change in A1/capita in 1990 was associated with a 0.57% change in IHD in 1995. A1/capita correlations were stronger for male than female mortality. On multiple regression of A1/capita and other food supply variables in 1990, only A1/capita was significantly correlated with IHD in 1995. DM-1 was correlated with supply of: A1/capita in milk and cream (r = 0.92, p <0.00001); milk and cream protein excluding cheese (r = 0.68, p <0.0001); and with A1/beta in milk and cream (r = 0.47, p <0.05). Correlations were not significant for A2, B or C variants of milk beta-casein. DM-1 incidence at 0-4, 5-9 and 10-14 years was equally correlated (r = 0.80, 0.81, 0.81 respectively) with milk protein supply. A 1% change in A1/capita was associated with a 1.3% change in DM-1 in the same direction.
Conclusions: Cow A1 beta-casein per capita supply in milk and cream (A1/capita) was significantly and positively correlated with IHD in 20 affluent countries five years later over a 20-year period--providing an alternative hypothesis to explain the high IHD mortality rates in northern compared to southern Europe. For DM-1, this study confirms Elliott's 1999 correlation on 10 countries for A1/capita,1 but not for B beta-casein/capita. Surveys of A1 beta-casein consumption in two-year-old Nordic children, and some casein animal feeding experiments, confirm the A1/capita and milk protein/capita correlations. They raise the possibility that intensive dairy cattle breeding may have emphasised a genetic variant in milk with adverse effects in humans. Further animal research and clinical trials would be needed to compare disease risks of A1-free versus 'ordinary' milk.