ABSTRACT
Three indicators were used to assess the onset of growth faltering: (a) average weight - and length - for - age as percentage of the NCHS references or (b) the Cambridge references based on breastfed infants and (c) incremental growth over 4 weeks in which "faltering" was defined as a weight increment less than minus 2 standard deviations, calculated by Waterlow from USA and British data or no gain in length. The population covered were singleton infants born in two villages in Madura in September 1982 through December 1984 (N = 391). Weight and length were measured at 4 week intervals. The first two indicators did not properly identify the age period at which growth became unacceptable according to weight or length increments. This was mainly due to the large variation and the skewed distribution of weight and length of Madurese infants. Arbitrarily a magnitude of 20% weight falterers was used as a cut-off point for the onset of growth faltering. Among Madurese infants it would be at or before 3-4 months. It is recommended to analyse weight increments from data, generated by growth monitoring at POSYANDUS for mapping of the age period at risk throughout Indonesia.
