LC Mitchell, LCK Low, J Karlberg

Abstract

The aim was to examine how well short-term growth in lower leg length (LLL) and total body length are correlated with weight gain in early life in healthy infants. This is an observational study with weekly follow up for 2 to 13 weeks for ages between birth and 21 weeks. The study was done at the Mother's Choice Adoption Center, Hong Kong. Weekly measurement of lower leg length (LLL), total body length and weight were made in 22 healthy full term Hong Kong Chinese infants (12 boys and 10 girls). The sample represents all new infants coming to the adoption center during the reqruitment period, January to May 1994. The technical error of LLL was 0.53 mm, which corresponds to 20% of the weekly growth rate in early life. In comparison, the technical error for length was 2.8 mm, or around 40% of the mean weekly growth rate in length. The main finding was a significant correlation between the gain in LLL and weight for short-term growth in early life in normal healthy babies: over 2, 3 and 4 week intervals the correlation were 0.32, 0.40 and 0.51, respectively. On the other hand the correlation coefficients between the gain in length and in weight were much lower for intervals below 4 weeks. We conclude that LLL and not length can be recommended for short-term growth studies in early life for intervals shorter than 4 weeks; the shortest interval seems to be 2 weeks in normal children, but possibly shorter under catch up growth. For intervals of 4 weeks or longer no recommendation can be made at present. LLL measurements may be useful in nutritional, hormonal and/or metabolic studies in early life by adding reliable information about longitudinal bone growth.

Introduction

Lower leg length (LLL) gives a more accurate estimate of short-term linear growth than total body length due to less measuring error influences.^1,2 In early life LLL can be measured by means of a recently developed hand-held knemometer.³ It has been used, for instance to study short-term growth influences of medical treatment, such as dexamethasone therapy in prematurely born babies with bronchopulmonary dysplasia.⁴ Other areas of use are human nutritional, endocrine and metabolic growth studies in early life. In extreme situations of disturbed nutrition in conjunction with a high incidence of diarrhoeal diseases, the gain in both weight and length has been shown to be highly affected.⁵ With the LLL device, it might be possible to observe nutritional, endocrine and/or metabolic influences on early linear growth in less extreme situations in which changes in length are unreliable. Furthermore, the aim of this study was to examine how well short-term growth in LLL versus total body length are correlated with weight gain in early life in healthy infants.

Methods

Between January and May 1994, 22 healthy full term Hong Kong Chinese infants (12 boys and 10 girls) residing at Mother's Choice Adoption Center, Hong Kong were enrolled into the study; mean birth weight was 3118 grams (range 2675-3900 grams), mean length of gestation 39.2 weeks (range 37-42 weeks) and a mean maternal age 16.5 years (range 14 - 20 years). The deliveries were uneventful in all babies.

Postnatally the babies were followed weekly and the period of follow up varied, since the duration of stay at the adoption centre varied individually; the duration of follow up ranged between 2 and 13 weeks for ages between birth and 21 weeks. Apart from a complete physical examination, five measurements of lower leg length (LLL) and total body length (length) were performed (by LCM) in a total of 201 measuring sessions. Weight was measured once at each session. All measurements were taken in a standardized manner using calibrated equipment.

The hand-held knemometer (manufactured by the FORCE Institute, Denmark) measures LLL by the means of a pair of electronic calipers which automatically record the measurement when the pressure applied on the heel reaches a preset value. The measuring system is based on a magnetic encoder with a resolution of 0.01 mm.³ Five sequential measurements were performed and the result of the measurement sequence was expressed as the average of the last four measures. The technical error was computed in a sequential manner: (a) the difference between the individual values and the mean value were computed for each measuring session giving the residual values; (b) the SD of the residual values was determined for each session; (c) the mean of these SD values over all sessions gave the technical error.³ The technical error for length was derived in a similar manner.

The body measurements were converted into standard deviation scores (SDS); Danish reference values were used for LLL (both sexes pooled, unpublished, personal communication with KF Michaelson). The NCHS growth reference served as reference for length and weight.⁶ The growth rate was expressed as the increase in SDS, or delta SDS over different interval lengths. The statistical analysis comprised of one sample t test and Pearson's correlation coefficient. Only two tailed tests were used.

Results

The technical error of LLL as determined over all measuring sessions was 0.53 mm in our hands. In the first two months of life, when the mean LLL growth velocity is 2.0-2.5 mm/week, the technical error corresponds to about 20% of the weekly growth rate. In comparison, our technical error for length was found to be 2.8 mm corresponding to around 40% of the mean weekly growth rate in length (6.9-9.1 mm/week).

The individual LLL values representing the mean of the last four measures for each measuring session are shown in Fig. 1 together with the reference values of Danish children. The mean SDS values for both LLL, length and weight over the different postnatal ages are shown in Fig. 2. The mean LLL SDS values were not significantly different from the reference mean (p<<0.05), but both mean weight SDS and mean length SDS were significantly below zero (p<<0.0001).

Fig. 1 Lower Leg Length in Chinese infants

Reference values for lower leg length in early life for Danish healthy infants (published with permission, KF Michaelson). The star represent all mean measuring session LLL values (n=210) of the longitudinally followed Chinese children (n=22).

Fig. 2 Mean SDS of Lower Leg Length, Length and Weight in Chinese infants

Mean SDS values lower leg length, length and weight, respectively; Danish and NCHS growth reference. Only weekly mean values with more than 9 observations were included. Both sexes were pooled.

Bivariate correlations for the change in SDS between the three body measurements are given in Table for different interval lengths, i.e. for 1, 2, 3 and 4 weeks. The correlation coefficient between LLL gain and weight gain reached a value above 0.30 for an interval of 2, 3 and 4 weeks, while it was below this figure for all four intervals between length gain and weight gain, and between length gain and LLL gain.

Discussion

The main finding of this study of normal healthy Chinese babies is that we found a significant correlation between the gain in lower leg length (LLL) and the gain in weight for short-term growth (2 week intervals or longer) in early life in normal healthy babies. However, not until the interval length was 4 weeks did the correlation between length gain and weight gain became of any clinical relevance (r > 0.20). In normal infants we should expect a positive correlation between skeletal growth and body weight gain. However, this correlation coefficient cannot be close to 1.0, since body weight reflects not only skeletal, but also muscle, fat and body liquid weight.

The reason that LLL is a more reliable measure than length for short-term growth is that the technical error is small as it has been reported in previous studies.^1-4 Our results show that the technical error of LLL is about 20% of the weekly growth rate in early life, while the figure was 40% of the weekly growth rate for length. However, other factors than the technical error may also contribute to measuring error influences, such as soft tissue changes between measuring sessions and seasonal variations in the growth rate.⁷ The correlation between short-term growth and long term growth is virtually zero.⁷ This partly seems to be related to the seasonal influences on growth and perhaps also to the proposed pulsatile short-term growth spurts.⁸ For this reason the interpretation of the results gained from short-term studies are completely different from the results of long-term studies. However, it is only in the short-term study that the growth rate can be related to season, disease, serum growth factors and their serum binding proteins, receptor activity, nutritional factors, body composition, treatment or other indicators. The results of such studies can provide us with knowledge about the biology and the dynamics of the growth process, but nothing reliable about the long-term aspects of growth.

The mean LLL length of our study population was similar to the Danish reference mean while both the mean total body length and the mean weight were significantly reduced. Normal Hong Kong Chinese babies have birth length and weight similar to that of Western countries.⁵ The babies in our study were born to primiparous young unmarried mothers. It is therefore not surprising that the birth length and weight of our study group were below the mean NCHS standard. This discrepancy in the growth of LLL and total body length has not been reported previously in Chinese infants, or other ethnic groups. It may be a result of sampling variation or the use of two different growth reference values, but can also stand for ethnic growth differences. Another plausible explanation is that LLL growth is not affected very much by prenatal maternal-placental related intrauterine factors, while both total body length and birth weight are.

Our conclusion is that LLL and not length can be recommended for short-term linear growth studies in early life for intervals shorter than 4 weeks. This conclusion is not only based on the relatively low measuring error of LLL, but also to the fact that length gain was shown to be poorly correlated with not only LLL gain, but also weight gain. LLL has a role to play in studies of nutritional, hormonal and/or metabolic influences on growth in early life by adding reliable information about short-term longitudinal bone growth.

Acknowledgement

The Faculty of Medicine, University of Hong Kong (CRC grants 335/045/0017 and 337/045/0009).

References

1. Valk I, Langhout Chabloz A, van Gust W. Intradaily variation of the human lower leg length and short term growth - A longitudinal study in fourteen children. Growth 1983;47:397-402.

2. Hermanussen M, Geiger-Benoit K, Burmeister J, Sippell W. Knemometry in childhood: accuracy and standardisation of a new technique of lower leg measurement. An Hum Biol 1988;15:1-16.

3. Michaelson KF, Skov L, Badsberg JH, Jorgenson M. Short-term measurement of linear growth in preterm infants: validation of a hand-held knemometer. Ped Res 1991 ;30:464-8.

4. Gibsson AT, Pearse EG, Wales KH. Growth retardation after dexamethasone administration: assessment by knemometry. Arch Dis Child 1993;69:505-9.

5. Karlberg J, Jalil F, Lam B, Low L, Yeung CY. Linear growth retardation in relation to the three phases of growth. Eur J Nutr 1994;48:525-44.

6. World Health Organization. Measuring change in nutritional status: guidelines for assessing the nutritional impact of supplementary feeding programmes for vulnerable groups. Geneva: WHO, 1983.

7. Karlberg J, Gelander L, Albertsson-Wikland K. Distinctions between short- and long-term growth studies. Acta Paediatr 1993;82:631-4.

8. Lampl M, Veldhuis JD, Johnson NIL. Saltation and stasis: a model of human growth. Science 1992;258;801-3.