This is the old United Nations University website. Visit the new site at http://unu.edu
Growthfaltering rates in California, Guatemala, and Tamil Nadu: Implications for growthmonitoring programmes
Reynaldo Martorell and Meera Shekar
The median weights for the Berkeley, Guatemala, and Tamil Nadu children are plotted relative to the WHO/NCHS reference curves (ninetyfifth, fiftieth, and fifth percentiles) in figure 1 (see Figure. 1. Median weights for boys in three populations compared with WHO/NCHS reference curves) and figure 2 (see Figure. 2. Median weights for girls in three populations compared with WHO/NCHS reference curves). The medians for Berkeley were consistently above the reference median, while those for both the developing country samples were below the fifth percentile. Also, the medians for Tamil Nadu were consistently lower than those for Guatemala.
The differences in median weight for boys at 24 and 12 months were 2.8 kg in Berkeley and 1.7 in both Guatemala and Tamil Nadu, compared with 2.4 kg in the WHO/NCHS reference population. In girls, the weight gains were 2.6, 1.9, and 1.5 kg respectively in Berkeley, Guatemala, and Tamil Nadu, compared with 2.4 kg in the reference sample. Compared with incremental growth charts that use Fels Research Institute data [18], the weight growth velocities in boys and girls exceeded the fiftieth percentile in the Berkeley sample but were around the tenth percentile for the other two samples.
Length values were not available for Tamil Nadu. The median lengths for Berkeley exceeded the reference median, particularly after 12 months of age, but the medians for Guatemala were much lower than the fifth percentile.
The distribution of threemonth weight increments is given in table 1. Here, the unit of analysis is childperiod, as each subject may contribute as many as four values to the total. Analyses restricted to only individuals with complete data for all periods give results similar to those shown in the table. As expected, the weight increments were larger in Berkeley than in either Tamil Nadu or Guatemala. The degree of faltering was similar in boys and girls in all three populations.
TABLE 1. Distribution of threemonth weight increments measured between 12 and 24 months of age
<=0 g 
1299 g 
300499 g 
>=500 g 
Total 

No. 
% 
No. 
% 
No. 
% 
No. 
% 

Berkeley  
all  38 
7.0 
37 
6.8 
65 
12.0 
403 
74.2 
543 
boys  24 
9.1 
21 
8.0 
27 
10.2 
192 
72.7 
264 
girls  14 
5.0 
16 
5.7 
38 
13.6 
211 
75.6 
279 
Guatemala  
all  453 
13.7 
601 
18.2 
622 
18.9 
1,623 
49.2 
3,299 
boys  244 
13.8 
307 
17.3 
356 
20.1 
864 
48.8 
1,771 
girls  209 
13.7 
294 
19.2 
266 
17.4 
759 
49.7 
1,528 
Tamil Nadu  
all  348 
14.8 
345 
14.7 
757 
32.3 
897 
38.2 
2,347 
boys  174 
14.2 
184 
15.0 
404 
33.0 
464 
37.9 
1,226 
girls  174 
15.5 
161 
14.4 
353 
31.5 
433 
38.6 
1,121 
Periods are 1215, 151A, 1821, and 2124 months.
Chisquare tests with three degrees of freedom were applied to the following: Berkeley boys vs. girls (5.66, NS), Guatemala boys vs. girls (4.93, NS), Tamil Nadu boys vs. girls (1.35, NS); Berkeley vs. Guatemala (120.76, p<.001), Berkeley vs. Tamil Nadu (231.81, p< .0001), and Guatemala vs. Tamil Nadu (149.13, p<.001)
The definition of faltering in the second year of life as a weight gain of less than 300 g over a threemonth period is widely used. The percentages of all threemonth increments that indicate faltering according to this definition are shown in figure 3 (see Figure. 3. Percentages of all threemonth weight increments of less than 300 g (based on data from table 1)). Some 17.1% of the increments in boys and 10.7% in girls in Berkeley were classified as faltering. The rates were two to three times as great in Guatemala and Tamil Nadu.
In table 2 the children are classified according to the number of episodes of growth faltering they experienced within the age ranges studied (1224 months in Berkeley and Guatemala, with a possible range of 04 episodes; 1221 months in Tamil Nadu, with a possible range of 03 episodes), using both definitions of faltering: a weight gain of 0 g or less (i.e., no change or loss of weight) over three months, and one of less than 300 g. The analysis is restricted to individuals with complete information for all possible childperiods. The sexes are combined because of the general similarity in the distribution, particularly in Guatemala and Tamil Nadu. Although faltering was considerably more frequent in Guatemala and Tamil Nadu, it did occur in many Berkeley children, a few of whom had repeated episodes of faltering.
TABLE 2. Children classified by number of episodes of growth faltering, 1224 months of age (Berkeley, Guatemala) or 1221 months (Tamil Nadu) according to two criteria—growth <=0 g and growth < 300 g over three months
Episodes 
Berkeley 
Guatemala 
Tamil Nadu 

No. 
% 
No. 
% 
No. 
% 

<=0 g 

0  112 
82.4 
348 
54.6 
139 
66.5 
1  13 
9.6 
232 
36.4 
63 
30.1 
2  8 
5.9 
53 
8.3 
6 
2.9 
3  3 
2.2 
4 
0.6 
1 
0.5 
4  0 
0.0 
0 
0.0 
— 

Total  136 
100.1 
637 
99.9 
209 
100.0 
< 300 g 

0  86 
63.2 
112 
17.6 
77 
36.8 
1  31 
22.8 
290 
45.5 
96 
45.9 
2  14 
10.3 
184 
28.9 
34 
16.3 
3  4 
2.9 
46 
7.2 
2 
1.0 
4  1 
07 
5 
0.8 
— 

Total  136 
99.9 
637 
100.0 
209 
100.0 
For Berkeley and Guatemala, based on weight measurements taken at ages 12,15,18, 21, and 24 months. For Tamil Nadu, the data were abstracted from the records in such a way that threemonth increments can be determined only for ages 1215, 1518, and 1821 months The sexes are combined in all the groups.
The percentage of children who faltered (<= 0 g and < 300 g) in at least one period from 12 to 24 months is shown in figure 4 (see Figure. 4. Percentages of children faltering in growth in at least one three month period between the ages of 12 and 24 months (based on data from table 2; for Tamil Nadu, the values based on three periods are projected to four).). The values for Tamil Nadu are adjusted for the fact that faltering was observed over only three periods; using the methods described earlier, the values shown project the rate of faltering to four periods of observation.
With a weight gain of 0 g or below as the criterion for faltering, 17.6% of the children in Berkeley faltered at least once between 12 and 24 months of Figure. 4. Percentages of children faltering in growth in at least one threemonth period between the ages of 12 and 24 months (based on data from table 2; for Tamil Nadu, the values based on three periods are projected to four) age, compared with 45.4% and 42.0% in Guatemala and Tamil Nadu respectively. In other words, faltering in at least one period was about 2.5 times as likely to occur in Guatemala and Tamil Nadu as in Berkeley. When a gain of less than 300 g is used as the criterion, the percentage of children in Berkeley who faltered at least once rises to 36.8%, compared with 82.4% in Guatemala and 73.6% in Tamil Nadu. Using the second definition, the rate of faltering in at least one period was twice as great in Guatemala and Tamil Nadu as in Berkeley.