| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
Department of Medical Statistics and Evaluation Imperial College of Science, Technology and Medicine Du Cane Road London W12 0NN United Kingdom E-mail: b.lausen{at}rpms.ac.uk
Dear Sir:
Itoh et al (1) state new evidence in their recent paper in this Journal that protein intake is a main factor of calcium excretion. They report detailed findings on the relation between protein, sodium, and calcium intakes and calcium excretion. Moreover, they provide estimates of the size of the effect. In what follows, I will explain some weaknesses in their data interpretation that may have resulted in overestimation of the effects. In addition, I have evidence of a different finding in healthy children and adolescents, and hence conclude that in these populations there is no evidence that dietary salt and dietary protein are main factors of calcium excretion.
Itoh et al state that errors in the measurement of daily dietary protein intake are relatively high. It is well known that measurement error in factor variables causes biased effect estimates in standard multiple regression and that the bias depends on the accuracy of the measurement—see review by Reeves et al (2). Therefore, the authors used urinary urea excretion and urinary sulfate excretion to assess the effect of the components of protein intake. Urinary measurements of urea, sulfate, calcium, and sodium in Itoh et al's study were likely to have correlated measurement errors because the measurements were made in the same urine sample. Consequently, a crucial assumption of standard multiple regression analysis was not valid. The inappropriate use of multiple regression analysis led to the unjustified conclusion that dietary salt and protein intakes have a major effect on calcium excretion. Two other studies also report evidence that dietary salt is a main factor regulating calcium excretion; however, both of these studies suffer from the same misuse of multiple regression and consequently their findings are not valid (3, 4).
In a preliminary interim analysis of data from the Dortmund Longitudinal Study (5) in healthy German children and adolescents aged 2.8–18.3 y (1985–1997: n = 187 boys and 850 observations and n = 180 girls and 807 observations), Lausen (6) showed that interindividual variation, reciprocal relative growth velocity [(
cm/y)/cm], and dietary protein intake (g/d) are the most important factors regulating urinary calcium excretion (mg/d). Other significant factors are magnesium intake (mg/d), sodium intake (mg/d), calcium intake (mg/d), and phosphorus intake (mg/d). The longitudinal data show that individual excretion explained 69% of the variance in the data (R2 = 0.69).
In summary, evidence for dietary protein and dietary salt as main factors of calcium excretion when intakes of protein, salt, and calcium are within usual ranges is not convincing.
REFERENCES
Department of Home Economics Tokyo Kasei Gakuin University 2600 Aiharamachi, Machida-City Tokyo 194–0292 Japan E-mail: ysuyama{at}tka.att.ne.jp
Dear Sir:
We appreciate greatly the sound criticism of Lausen on the weakness of the data interpretation in our study. Recently, we also realized the bias in the analysis of the relation of the measured variables in the same 24-h urine specimens. As Lausen correctly pointed out, the positive correlations observed among the measurements of daily calcium excretion were possibly due to errors commonly associated with urinary specimens, especially errors that occur in 24-h urine collections, although we carefully excluded subjects who appeared to have had problems with urine collection. Thus, we reanalyzed the data from our previous study using daily excretions of various urinary constituents corrected for daily urinary creatinine excretion (1). The results of multiple regression analyses still showed significant and positive correlations between calcium-creatinine, urea-creatinine, calcium-creatinine, and sodium-creatinine ratios after adjustment for sex, age, and calcium intake in both age groups: 20–49- and 50–79-y-olds.
Of course, we recognize that multiple regression analysis does not completely eradicate the bias. However, multiple regression analyses showed that the daily dietary intake of protein estimated from dietary records was significantly and positively correlated with daily calcium excretion in the same population (1).
The reciprocal association between urinary calcium excretion and dietary protein intake observed by Lausen in healthy children and adolescents is an interesting and important finding. The association is reasonable because protein is an important constituent of bone, and young individuals, contrary to adults, require large amounts of protein for bone growth and are in strong positive nitrogen balance. It is highly possible that low protein intakes prevent effective utilization of dietary calcium for bone formation, resulting in an increase in urinary calcium excretion.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  T. Remer, T. Dimitriou, and C. Maser-Gluth Renal Net Acid Excretion and Plasma Leptin Are Associated with Potentially Bioactive Free Glucocorticoids in Healthy Lean Women J. Nutr., February 1, 2008; 138(2): 426S - 430S. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
