Acta Padiatr 81: 264-5. 1992
SHORT COMMUNICATION
Prediction equations for total body water during early infancy Nancy F Butte, William W Wong and Cutberto Garzal USDAIARS Children's Nutrition Research Center, Baylor College of Medicine and Texas Children's Hospital, Houston, Texas, USA; Division of Nutritional Sciences', Cornell University, Ithaca, New York, USA
In the classic publication of Friis-Hansen (I), prediction equations for total body water (TBW) were provided for infants and children from birth to age 16 years. Equations using weight and height were derived from 73 measurements of TBW by deuterium dilution. In the 1 I-day to 0.5-year age-group, there were 19 infants. We have found that the Friis-Hansen equation, specified for this age-group, overestimates our measurements of TBW in infants by 12%. We therefore present new equations derived from measurements of 2H and I8O dilution spaces for the prediction of TBW during early infancy. Forty term, appropriate for gestational age, breastfed and formula-fed infants were studied at one month (n=20, mean age (SD) 32 (4) days, 15 males/five females) or four months (n= 20, age 1 I4 ( I 0) days, I 1 males/nine females) (2). At the time of study all infants were healthy and thriving; mean weight was 4.75 (0.5) and 6.60 (0.61) kg and mean length was 54.9 (1.9) and 62.0 (1.8) cm at one and four months, resbectively. TBW was determined by deuterium and oxygen- 18 dilution. A baseline urine sample was collected from each infant before oral administration of a dose of 2H2'80,equivalent to 200 mg 'H20 (99.8 atom percent 2H) and 300 mg I80/kg body weight. Post-dose urine samples were collected daily thereafter for 14 days in sterile, pediatric specimen bags (U-Bag, Hollister, Inc., Kurtsville, MO). Urine samples were analyzed for 'H using a Finnigan Delta-E gas-isotope-ratio mass spectrometer (Finnigan MAT, San Jose, CA) and for I8Ousing a VG-SIRA 12/ ISOPREP-I 8 mass spectrometer system (VG Isotec, Cheshire, UK). Initial isotope dilution spaces (TBWI) were calculated from the changes in isotopic enrichment of 2H or '*O in urine ( a d ) as TBW I (kg) =
d-A.6, 103- a ' 6 d
where d is the dose of 2H20 or H2IXO(g); A is the laboratory water (g) used in the gravimetric dilution of the isotopic water (a, g); 6, is the change in isotopic abundance of the laboratory water following the addition of the isotopic water; and lo3 is for the conversion of g to kg. Mean TBW, corrected for changes in 'H or IXOdilution spaces during the study, was calculated as TBW(kg) = (TBWI-TBW2)/ln(TBWI/TBW2) where TBWl is the initial and TBW2 the final isotope
dilution space. TBW2 was estimated from weight gain based on the assumption that the dilution spaces changed in proportion to weight during the course of study. The mean TBW of the 40 infants was 3.49 (0.49) kg or 62.3 (5.6)% body weight based on 2H dilution, and 3.41 (0.47) kg or 61.0 (5.9)% based on I80dilution. Application of the Friis-Hansen equation for infants aged 11 days to 0.5 years
TBW=0.177.WT0.790.LN0.425 where TBW and weight (WT) are in kg and length (LN) is in cm, resulted in a calculated TBW value of 3.93 (0.71) kg or 69.5 (l.2)% body weight. This value was 12.2 (9.2)% higher than our estimations based on 2H measurements. In the above equation, Friis-Hansen used the average regression coefficients for weight and height across ages 1 1 days to 16 years, because no statistically significant difference was found between coefficients. Application of equations using regression coefficients specific to the age-group 1 1 days to 0.5 years TBW =0.264-WT0.79'.LN0.325 s=0.068 r2=95.5% TBW=0.851 -WT0.881 s=0.066 r2=95.7% yielded values of TBW equal to 3.91 (0.68) kg and 3.92 (0.67) kg, respectively. Compared with our estimations, TBW was overestimated by 11.6 (8.6)% and 12.0 (8.4)% using these age-specific equations. We propose new equations for the prediction of TBW based on the 2H and I80dilution spaces of our 40 infants. Because regression coefficients for weight and height on TBW were not statistically different for specific equations derived for one- and four-month-old infants, the data were combined for all 40 infants. A linear regression of the natural logarithm of TBW on the natural logarithms of weight and length resulted in the following equations: s
TBW(2H)=0.389.WT0.549.LN0.3W 0.060
T B W ( ~ H= ) I . I3 2 . w ~ O . ~ ~
0.059
TBW('80)=0.375.WT0.516.LN0.324 0.065 TBW("O)= 1.160.WT0.623
0.064
? 82.3 82.6 78.2 78.5
95% CI 3.39-3.52 3.39-3.52 3.32-3.46 3.32-3.46
95% PI
3.06-3.90 3.06-3.90 2.97-3.86 2.94-3.86
The 95% confidence interval (CI) for the regression line and the 95% prediction interval (PI) for individual values were calculated at the mean weight and length of the 40 infants. TBW is used in the study of growth and development
ACTA PEDIATR R I (1992)
( I , 3), assessment of nutritional management (4)and estimation of milk intake from deuterium dilution (5). Isotope dilution space of I8O is a closer approximation of TBW than that of 2H, because I8O exchanges with non-aqueous body constituents to a lesser extent than 2H (6,7). In some applications (3,however, use of 2H dilution space is required. The 19 infants in the 1 1-day to 0.5-year age-group studied by Friis-Hansen were described as normal; mean weight was 4.21 (1.43) kg and mean length was 54.5 (5.5) cm. TBW average 3.01 (0.93) kg or 72.4 (0.06)% of body weight. A body hydration of 72.4% for these 1.7(1.4)-month-oId infants is higher than most recent reports in the literature (3). In accord with Friis-Hansen (l), we found the prediction equations for TBW to be highly dependent upon weight and independent of age. Because the equations that we propose in this study are based on a larger sample size and on analytical improvements in mass spectrometry since 1950, they are more appropriate for the prediction of TBW during early infancy.
References I . Friis-Hansen B. Changes in body water compartments during growth. Acta Paediatr 1957;46(suppl I 10):1-68
Prediction equations for TB W
265
2. Butte NF, Wong WW, Ferlic L, Smith EO, Klein PD, Garza C. Energy expenditure and deposition of breast-fed and formula-fed infants during early infancy. Pediatr Res 1990;28:631-40 3. Fomon SJ, Haschke F, Ziegler EE, Nelson SE. Body composition of reference children from birth to age 10 years. Am J Clin Nutr 1982;35:1169-75 4. Fjeld CR, Schoeller DA, Brown KH. A new model for predicting energy requirements ofchildren during catch-up growth developed using doubly labeled water. Pediatr Res 1989;5:503-8 5. Coward WA, Cole TJ, Sawyer MB, Prentice AM. Breast-milk intake measurement in mixed-fed infants by administration of deuterium oxide to their mothers. Hum Nutr Clin Nutr 1982;36C:141-8 6. Schoeller DA, van Santen E, Petersen DW, Dietz W, Jaspan J, Klein PD. Total body water measurement in humans with "0and 'H labeled water. Am J Clin Nutr 1980;33:2686-93 7. Trowbridge FL. Graham GG, Wong WW, Mellits ED, Rabold JD, Lee LS, Cabrera MP, Klein PD. Body water measurements in premature and older infants using H2'*0 isotopic determinations. Pediatr Res 1984;18:524-7
Received December 27, 1990. Accepted May 15. 1991
SHORT COMMUNICATION
Prediction equations for total body water during early infancy Nancy F Butte, William W Wong and Cutberto Garzal USDAIARS Children's Nutrition Research Center, Baylor College of Medicine and Texas Children's Hospital, Houston, Texas, USA; Division of Nutritional Sciences', Cornell University, Ithaca, New York, USA
In the classic publication of Friis-Hansen (I), prediction equations for total body water (TBW) were provided for infants and children from birth to age 16 years. Equations using weight and height were derived from 73 measurements of TBW by deuterium dilution. In the 1 I-day to 0.5-year age-group, there were 19 infants. We have found that the Friis-Hansen equation, specified for this age-group, overestimates our measurements of TBW in infants by 12%. We therefore present new equations derived from measurements of 2H and I8O dilution spaces for the prediction of TBW during early infancy. Forty term, appropriate for gestational age, breastfed and formula-fed infants were studied at one month (n=20, mean age (SD) 32 (4) days, 15 males/five females) or four months (n= 20, age 1 I4 ( I 0) days, I 1 males/nine females) (2). At the time of study all infants were healthy and thriving; mean weight was 4.75 (0.5) and 6.60 (0.61) kg and mean length was 54.9 (1.9) and 62.0 (1.8) cm at one and four months, resbectively. TBW was determined by deuterium and oxygen- 18 dilution. A baseline urine sample was collected from each infant before oral administration of a dose of 2H2'80,equivalent to 200 mg 'H20 (99.8 atom percent 2H) and 300 mg I80/kg body weight. Post-dose urine samples were collected daily thereafter for 14 days in sterile, pediatric specimen bags (U-Bag, Hollister, Inc., Kurtsville, MO). Urine samples were analyzed for 'H using a Finnigan Delta-E gas-isotope-ratio mass spectrometer (Finnigan MAT, San Jose, CA) and for I8Ousing a VG-SIRA 12/ ISOPREP-I 8 mass spectrometer system (VG Isotec, Cheshire, UK). Initial isotope dilution spaces (TBWI) were calculated from the changes in isotopic enrichment of 2H or '*O in urine ( a d ) as TBW I (kg) =
d-A.6, 103- a ' 6 d
where d is the dose of 2H20 or H2IXO(g); A is the laboratory water (g) used in the gravimetric dilution of the isotopic water (a, g); 6, is the change in isotopic abundance of the laboratory water following the addition of the isotopic water; and lo3 is for the conversion of g to kg. Mean TBW, corrected for changes in 'H or IXOdilution spaces during the study, was calculated as TBW(kg) = (TBWI-TBW2)/ln(TBWI/TBW2) where TBWl is the initial and TBW2 the final isotope
dilution space. TBW2 was estimated from weight gain based on the assumption that the dilution spaces changed in proportion to weight during the course of study. The mean TBW of the 40 infants was 3.49 (0.49) kg or 62.3 (5.6)% body weight based on 2H dilution, and 3.41 (0.47) kg or 61.0 (5.9)% based on I80dilution. Application of the Friis-Hansen equation for infants aged 11 days to 0.5 years
TBW=0.177.WT0.790.LN0.425 where TBW and weight (WT) are in kg and length (LN) is in cm, resulted in a calculated TBW value of 3.93 (0.71) kg or 69.5 (l.2)% body weight. This value was 12.2 (9.2)% higher than our estimations based on 2H measurements. In the above equation, Friis-Hansen used the average regression coefficients for weight and height across ages 1 1 days to 16 years, because no statistically significant difference was found between coefficients. Application of equations using regression coefficients specific to the age-group 1 1 days to 0.5 years TBW =0.264-WT0.79'.LN0.325 s=0.068 r2=95.5% TBW=0.851 -WT0.881 s=0.066 r2=95.7% yielded values of TBW equal to 3.91 (0.68) kg and 3.92 (0.67) kg, respectively. Compared with our estimations, TBW was overestimated by 11.6 (8.6)% and 12.0 (8.4)% using these age-specific equations. We propose new equations for the prediction of TBW based on the 2H and I80dilution spaces of our 40 infants. Because regression coefficients for weight and height on TBW were not statistically different for specific equations derived for one- and four-month-old infants, the data were combined for all 40 infants. A linear regression of the natural logarithm of TBW on the natural logarithms of weight and length resulted in the following equations: s
TBW(2H)=0.389.WT0.549.LN0.3W 0.060
T B W ( ~ H= ) I . I3 2 . w ~ O . ~ ~
0.059
TBW('80)=0.375.WT0.516.LN0.324 0.065 TBW("O)= 1.160.WT0.623
0.064
? 82.3 82.6 78.2 78.5
95% CI 3.39-3.52 3.39-3.52 3.32-3.46 3.32-3.46
95% PI
3.06-3.90 3.06-3.90 2.97-3.86 2.94-3.86
The 95% confidence interval (CI) for the regression line and the 95% prediction interval (PI) for individual values were calculated at the mean weight and length of the 40 infants. TBW is used in the study of growth and development
ACTA PEDIATR R I (1992)
( I , 3), assessment of nutritional management (4)and estimation of milk intake from deuterium dilution (5). Isotope dilution space of I8O is a closer approximation of TBW than that of 2H, because I8O exchanges with non-aqueous body constituents to a lesser extent than 2H (6,7). In some applications (3,however, use of 2H dilution space is required. The 19 infants in the 1 1-day to 0.5-year age-group studied by Friis-Hansen were described as normal; mean weight was 4.21 (1.43) kg and mean length was 54.5 (5.5) cm. TBW average 3.01 (0.93) kg or 72.4 (0.06)% of body weight. A body hydration of 72.4% for these 1.7(1.4)-month-oId infants is higher than most recent reports in the literature (3). In accord with Friis-Hansen (l), we found the prediction equations for TBW to be highly dependent upon weight and independent of age. Because the equations that we propose in this study are based on a larger sample size and on analytical improvements in mass spectrometry since 1950, they are more appropriate for the prediction of TBW during early infancy.
References I . Friis-Hansen B. Changes in body water compartments during growth. Acta Paediatr 1957;46(suppl I 10):1-68
Prediction equations for TB W
265
2. Butte NF, Wong WW, Ferlic L, Smith EO, Klein PD, Garza C. Energy expenditure and deposition of breast-fed and formula-fed infants during early infancy. Pediatr Res 1990;28:631-40 3. Fomon SJ, Haschke F, Ziegler EE, Nelson SE. Body composition of reference children from birth to age 10 years. Am J Clin Nutr 1982;35:1169-75 4. Fjeld CR, Schoeller DA, Brown KH. A new model for predicting energy requirements ofchildren during catch-up growth developed using doubly labeled water. Pediatr Res 1989;5:503-8 5. Coward WA, Cole TJ, Sawyer MB, Prentice AM. Breast-milk intake measurement in mixed-fed infants by administration of deuterium oxide to their mothers. Hum Nutr Clin Nutr 1982;36C:141-8 6. Schoeller DA, van Santen E, Petersen DW, Dietz W, Jaspan J, Klein PD. Total body water measurement in humans with "0and 'H labeled water. Am J Clin Nutr 1980;33:2686-93 7. Trowbridge FL. Graham GG, Wong WW, Mellits ED, Rabold JD, Lee LS, Cabrera MP, Klein PD. Body water measurements in premature and older infants using H2'*0 isotopic determinations. Pediatr Res 1984;18:524-7
Received December 27, 1990. Accepted May 15. 1991
