点击显示 收起
Department of Medicine
University of California, San Francisco
Moffitt/MZ General Clinical Research Center
40 Crags Court
San Francisco, CA 94131
E-mail: anthony_sebastian{at}msn.com
Dear Sir:
In a recent Journal article, Devine et al (1) offered further evidence in humans of a bone anabolic effect related to an increase in dietary protein intake from amounts below to amounts above currently recommended intakes—to wit, an anabolic effect that occurs despite the potentially offsetting catabolic effect of greater endogenous acid production due to higher protein intake.
In a cross-sectional study of elderly women, the investigators reported significantly more positive bone status with protein intakes of >87 g/d than with intakes of <66 g/d, the highest and lowest tertile groups of protein intake, respectively. It is important, however, that they did not report the average values of protein intake in those 2 groups or the within-group distribution of values. We would like to have that information because of the possibility that the healthier bone status in subjects with higher protein consumption reflects primarily beneficial bone effects in those subjects only when they are compared with subjects who are consuming decidedly suboptimal protein intakes. In the low-protein group, a major unmet need of bone for a protein anabolic effect may render the bone particularly responsive to the catabolic effect of the dietary positive net acid load, even though the net acid load may be reduced by the lower protein intake. With little countervailing protein anabolic effect, the catabolic effect of any degree of positive dietary net acid load may prevail unchecked, taking its toll over the years.
In the overall study group examined by Devine et al (n = 1077), in which protein intakes ranged from 25 to 136 g/d and averaged (± SD) 80.5 ± 56 g/d, the subjects in the lowest tertile group, who consumed < 66 g protein/d, had an average protein intake between 25 and 65 g/d, and thus this group presumably included many subjects with decidedly suboptimal habitual protein intakes. Moreover, the investigators found significantly different positive bone effects of dietary protein between the lowest and highest tertiles of protein intakes but not between the middle and highest tertiles, which may suggest that there is a prerequisite for a decidedly suboptimal protein intake in the comparison group against which higher protein intakes are associated with improved bone health.
Extrapolating interventionally, I would suggest that a person's bone, when avid for protein's anabolic effect because of the person's habitually suboptimal protein intake, will experience a net anabolic effect when the person consumes a greater amount of protein, despite an accompanying greater acid load from the higher protein intake. (Indeed, the acid load may increase little, because the protein-deficient body retains rather than catabolizes the additional protein.) Increasing protein intake from, say, 40 to 110 g/d may have a net anabolic effect on bone because the habitual low protein consumption has made the bone particularly receptive to protein's anabolic effect, thus providing a greater counterbalance to the catabolic effect of the potentially higher net acid load. But increasing protein intake by equal increments—from, say, 80 to 150 g/d—may have a net catabolic effect on bone because the habitual higher protein consumption has made the bone less receptive to protein's anabolic effect, thus providing a lesser counterbalance to the catabolic effect of the potentially higher net acid load.
Devine et al may have estimated the dietary net acid load for each subject, as they did for dietary protein content. With that information and on dividing their subjects into 2 groups, they may have found, on multivariate analysis, an independent negative determinant in dietary net acid load that had a lesser effect (ie, a lower standardized regression coefficient) than did protein in the group with protein intakes between the lowest and the median and a greater effect than did protein in the group with protein intakes between the median and the highest.
The investigators possibly could have made that analysis from their data. They may have estimated the dietary net acid load for each subject by using data from the same food-frequency questionnaire that they used for estimating dietary protein and calcium content. Presumably, the food-frequency questionnaire could also have provided estimates of dietary potassium intake for each subject, in which case the investigators could have used the Frassetto algorithm (dietary net acid load = 0.91 x protein intake in g/d – 0.57 x potassium intake in mEq/d + 21) (2) to obtain an estimated value for dietary net acid load for each subject.
In performing such an analysis, Devine et al may contribute quantitative estimates of the degrees of opposing anabolic and catabolic bone effects of dietary protein in elderly subjects over the range of protein intakes observed, and they may confirm in adults the findings of Alexy et al (3) in children and adolescents, also recently reported in the Journal. In an editorial accompanying that article, we (4) further discussed the subject of the opposing anabolic and catabolic bone effects of dietary protein and suggested a way to maximize protein's anabolic effect by supplying diets that are both protein-rich and net base–producing.
ACKNOWLEDGMENTS
The author had no conflict of interest.
REFERENCES