Growth Velocity of Infants from Birth to 5 Years Born in Maku, Iran

Background: Growth velocity standards are essential for proper evaluation of child growth. The goal of this study was to construct weight, height and head circumference growth velocity charts for infants. Methods: This study includes 256 infants (124 boys and 132 girls) born in Maku, Northwest of Iran, and monitored from birth until they were 5 years. The weights and heights of the subjects were recorded at birth, one, two, four, six months and 1, 1.5, 2, 3, 4 and 5 years of age, while the head circumferences were measured until they were 1.5 years old. In this study, the LMS method using LMS chart maker software, was utilized to obtain growth velocity centiles. Results: Growth velocity charts for weight, height and head circumference (5th, 50th, 95th percentiles) were obtained. The velocity growth charts decreased rapidly from birth to 2 years and then remained relatively constant up to 5 years for both sexes. The growth velocity of boys was higher than girls through the first year of age but became equal at 12 months of age and no significant difference was seen up to 5 years. Conclusion: Growth velocity studies are really sparse in Iran. In this study, longitudinal data were used to obtain growth velocity centiles. Furthermore, the weight and height velocities of infants from Shiraz, southern Iran, and U.K were compared.


Introduction
Growth is a dynamic process whereby a living organism increases in size over a period of time. The accurate monitoring of physical growth in infants is of value for the assessment of health status, identifying deviations from normality and determining the effectiveness of interventions (de Onis et al., 2012).
Anthropometric measurements are significant indices to predict infant health and future outcome (S. M. Ayatollahi & Shahsavari, 2002). Weight, height and head circumference of new born infants are the most common parameters for measuring the physical growth of infants (Fok et al., 2003).
Growth velocity is a desirable approach that depicts the age-dependent changes in velocity that characterize human postnatal growth (De Onis & World Health Organization, 2009). Standards for velocity are required to determine whether the rate of infant growth over the past month, had been within normal limits for his or her age, sex, etc. It has advantages as an epidemiological tool when compared with distance charts. The latter is the result of a cumulative growth experience, whereas growth velocity represents what is happening at present (Hosseini, Maracy, Sarrafzade, & Kelishadi, 2014).
Several studies have assessed infant growth using distance charts and constructed reference growth curves during the past two decades in Iran (S. M. Ayatollahi & Shahsavari, 2002;S. M. T. Ayatollahi, 2001S. M. T. Ayatollahi, , 2012S. Ayatollahi & Ahmadi, 2001;S. M. T. Ayatollahi & Shayan, 2008;Emdadi, Safarian, & Doosti, 2011;Heydari, 2009;Mirfazeli, Besharat, Rashedi, & Rabiee, 2009). Based on the longitudinal data however, the only reference growth velocity standard in Iran was provided by the first author (S. M. T. Ayatollahi, 2005) in Shiraz, one of the five main cities of Iran. This paper presents growth velocity standards for height, weight and head circumference of a birth cohort, followed longitudinally from birth to 5 years of age in Maku, and compared the relevant anthropometric velocity measures with that of the study conducted in Shiraz (S. M. T. Ayatollahi, 2005).

Methods
Maku, one of the West Azerbaijan province cities with a rural population, is located in the Northwest of Iran, sharing borders with Turkey. It has a semi-arid cold climate and a human population of 202939, including 8.13% of the total population of the province (http://fa.wikipedia.org).

Subjects
A birth cohort of 256 apparently healthy neonates (124 boys and 132 girls) born in Maku in 2004 were recruited. The subjects were followed for five years until 2009 and visited at the health centers. A questionnaire on the demographic and health status of the neonates and their parents including anthropometric measurements was completed. The weighs and heights of the subjects were measured by trained auxologists at birth, one, two, four, six months and 1, 1.5, 2, 3, 4 and 5 years of age and head circumferences were recorded just up to 1.5 years of age. Heights were measured in supine position until one year of age and then in standing position in centimeter. Weights were measured to the nearest 10 g until the second year of age using a baby scale and onwards to 0.1 kg.

Statistical Analysis
The birth weight of 10 subjects (3.9%) were under 2500 g (range 1600-2400g). No subject was observed that failed to thrive. Only one subject left the study by age of 4 years and another 18 subjects (7%) by age of 5 years, giving a total attrition rate of 7.4%.
The growth velocity formula is given asV ∆ , where M and M are measurements at adjacent occasions, and t  is the time interval between them. Due to the fact that each measurement has its own measurement error, the variance V is given by var V σ 2ε ∆t ⁄ , where  is the measurement error and  is the population standard deviation of the true measurement velocity (Bairagi, 1986). This indicates that the variability of velocity (and hence the spacing of centile curves) depends on the time interval t  between measurements. The velocity standards must therefore be tied to a particular time interval, to ensure that the two components of error are weighted appropriately. For length and weight during infancy, intervals between 2 weeks and 3 months have been proposed (Bairagi, 1986;Emery, Waite, Carpenter, Limerick, & Blake, 1985;Guo et al., 1991;Piwoz, Peerson, & Brown, 1992).
Coleʾs LMS method (Cole & Green, 1992) was performed to construct smoothed weight, height and head circumference velocity centiles. The LMS method summarizes the changing distribution using the median(M) curve, the coefficient of variation (CV) of the distribution (called the S curve) and the skewness of distribution, expressed as a Box-Cox power (the L curve) (Tanner, Whitehouse, & Takaishi, 1966a, 1996b. Height is usually normally distributed, while most other anthropometric measurements tend to be skewed. The benefit of the LMS method is that the normalized growth centile standards are obtained using SD scores which deal with skewness that may be presented in the distribution of measurements, such as height or weight. The SD score (standard deviation score) or the Z score was calculated using the formula below: Where, measurement is weight, height or head circumference velocity for an infant of age t months (Bong, Shariff, Mohamed, & Merican, 2012). The LMS method assumes that skewed data can be normalized using a power transformation which stretches one tail of the distribution and shrinks the other tail simultaneously, thus removing the skewness. Using penalized likelihood the L, M and S curves can be fitted as cubic splines by non-linear regression, and the extent of smoothing required can be expressed in terms of smoothing parameters or equivalent degree of freedom. Finally, any required centile curve 100 ( ) t C  as specified by the corresponding Normal equivalent deviate Z , for tail area  , could be derived from (Bong et al., 2012).
Statistical analysis was performed using statistical analysis software SPSS version 11.5 (SPSS Inc., Chicago, USA). LMS chart maker (professional) and Excel softwares were used for construction of weight, height and head circumference velocity centiles. Furthermore, the goodness of fit of data was evaluated by the Z-score method and by the Q-Q plot using the LMS chart maker software (Pan & Cole, 1997). www.ccsen