The Soda Debate Fizzes On


To the Editor:

We were surprised by the stridency of the statement of Drs Amato, Araujo, and Paniagua that our conclusion about the displacement effect of soft drinks was “non-evidenced-based opinion.” An editorial does not provide opportunity for a comprehensive review of a complex literature, but we are grateful for the opportunity to stress here that the absence of deleterious effects of caffeine and phosphorus, the two principal components of colas that are generally implicated in bone status, has been well established in extended human metabolic balance studies.(1, 2) Furthermore, and in support of a displacement effect, Barrett-Connor et al.(3) showed that the deleterious effect of caffeine in the Rancho Bernardo cohort was confined to those individuals with the lowest calcium intakes and that with an adequate calcium intake, caffeine was without effect.

Many more such studies could be cited, but in this brief space, we believe it important to deal with the rat experiment by García-Contreras et al.,(4) which Dr Amato et al. cite as evidence of a direct harmful effect of soft drinks. While clever in conception, the pair-feeding experiment to which they allude was flawed in execution. The cola-consuming rats imbibed a quantity of cola equal to roughly one-half their body weight each day. This huge cola intake more than doubled their phosphorus intake relative to the pair-fed controls, and its sugar more than doubled the energy intake from food as well. Altogether apart from the massive water and sugar diuresis this intake would have produced, it is also out of all proportion to what would be obtained with human cola consumption.

Despite the high energy intakes, the animals were manifestly globally malnourished, as evidenced by a substantial drop in serum albumin.(4) In any event, and much more to the point, the more than doubling of phosphorus intake in this experiment would have resulted in substantial binding of the calcium in the rat chow, effectively reducing the calcium intake of the cola-consuming rats. This is actually displacement with a vengeance. Not only was the cola displacing solid food, but it was likely reducing its effective calcium content as well.

Finally, the extremely high cola intake in this small experiment (N = 7) cannot be extrapolated to a human situation, and specifically in this case, to the paper by McGartland et al.,(5) to which our editorial was a response. Whereas the distribution of cola intakes was positively skewed in the McGartland study, mean cola consumption in the girls was ∼250 g/day (i.e., ∼8 oz), providing a vastly smaller contribution to energy and phosphorus intakes than in the rat experiment cited by Amato et al.(4)

To our knowledge, there are no data at soft drink intakes more typical of human consumption that indicate that soft drink constituents are harmful to the calcium economy. On the contrary, there is much negative evidence indicating that those constituents are effectively harmless, as long as total nutrition (and specifically calcium intake) is maintained.

Finally, it should have been clear from our editorial that we were in no sense apologists for the carbonated beverage industry, and we certainly do not favor high intakes of such liquids. An argument against such consumption that we could have emphasized (except that it was not particularly pertinent to the McGartland paper) is the role that high carbonated beverage intakes play in the current worldwide epidemic of obesity.