RURAL/URBAN RESIDENC~ AND FERTILITY
II. THE SURVEY METHODOLOGY AND ANALYTICAL BACKGROUND
As part of the general survey on fertility, the interrelationships among infant and childhood mortality, socio-economic factors and fertility, a study involving the measurement of weights, heights and arm circumference of children aged 1-5 years was undertaken. Measureoents were recorded on a specially designed foro for approximately 700 children representing a 10 percent subsanple of households.
Unlike the study of other variables which was carried out on a longitu-dinal basis, the anthroponetric study was purely cro99@sectional. A1th~u8h
it is appreciated that a combination of the two perspectives should provide a richer description of the growth processes of children, the use of the cross-sectional method was thought to enhance the study of variations and permit the accurate placement of each child at a point in time in relation to other children. This does not of course over-look the lioitations of the technique, principal among which is that while the nean value of weight or height or arm circumference provides the best estinate for the group, it does not really describe the ac tual gr,1wth r)f the indivlildual chilla. This is
because of sampling differences fron one age group to another. Furthernore, children attain their growth spurt at different times. In particular during infancy and up to about 4 years, growth of the long bones of the extremities is quite erratic; it only stabilises durin?, mid-childhood (4-10 in girls and 4-12 in boys) because sonatic growth and physiolo~ical processes become stable.
11
Thus, cross-sectional data especially during the growing period by age are less valuable than lonv.itudinal data since it is the velocity of r.rowth which lIIay be affected by nutrition. However, the advanta~e of usin~the method for comparinn, different p,roups with common back~rnund stands.
The data on weight was obtained by weighing each child using a simple portable bathroom scale on which children were made to stand upright. The observed weight was then recorded by the enumerator on the schedule designed for the investigation. All enumerators were uniformly trained on how to
~easure and record, and use and adjust~ the scale to the zero point before weighing in order to minimise observer and recording errors and biases.
Since the sample of children was small, the measurements were tallied and processed manue.lly to obtain the tables, graphs and charts for the present analysis.
1/ M.M. Maresh, "Linear growth of long bones of extremities from infancy through adolescence"., American Journal of Disabled Child, 89: 725-742, June 1955.
M.M. Maresh, "Variations "n patterns of Linear growth and skeletal maturation", Journal of American Physio-Therapy Association, 44 (l!'l) , October 1964.
Hdght w!'!s measur"d with the child standing upright with his/her b:1Ck against the wall. A ruler placed flat on the child's head w~S used to nark-off the corresponding height on the wall and then measured with a tape measure and recorded against the n~me of the child on the schedule. Rere again marked varinbility in observer and recording errors and biases were kept to a minimum by giving the same level nne type of proup training to ' enunerators*
Tailors' tape measures made of cloth were also used to measure arm
circ~ference. Enunerat~rs were instructed to apply the tare p,ently and not so tirht but firmly around the left arm half-w,>y between the shoulder and the tir of the elbow. The me!lsured left arm was kept strt'i<:ht :md l ' left to hang freely down by the side of the child. The measurements were recorded on the schedule against the name of the child. A manual tally of all the records was mad~ and used to construct arm-circumferenc~-for-a~e
charts for each of the survey locations.
The schedule for the investip,ation also collected information on the general health condition of each child in the subsamplc stuey in which 705 children were actually observed. Of these 92 were from the urban low density stratum, 202 from the urban hi~h density stratum, 307 fro~ the urban squatter areas, and 104 from the rural stratun. The schedule 3180
had space f0r recording the name. age and survey serial number of the
mcth~rs 0f the children in the sub-sample.
III. ANTHROPOMETRIC FINDINGS: BODY ;,mIGl-1T AND HEIGHT
Table 1 presents separately by age observed and standardized measur~s
of height and weight for each of th" four sample strata in the survey. The ages of the children were recorded in the survey as thGir respective ages last birth'!!'!y· Because of the sample size, anc.lysis is confined to both Sexes to"ether.
It can be seen from Tabl2 1 and Fig. 1 that there is an approximately linear increase in mean height and mean weight as age of children increases.
ijoth observed and standardized aver~ges for the four areas support th1s
"bs ervat ion , althoueh it has tn be interpreted cautiously in view of the re.ther small sample size. Even when the tabulations were prepared sc?ara-tely for "lale and feMale children. the same tendency for height ane! weip,ht to increase as age increases was observed. It is appp-rent that the
linearity of the relationship could have been nore direct but for the smallness of the sample size.
Iable 2 shows mean heieht and weif,ht for males and fenales in the four survey locations. Children in urban areas were both taller and heavier than those in the rural are~.s and those in high density areaS were taller and heavier than their counterparts in low density and squatter areas. Female children in low density areas were however heavier ~nrl taller thrtn those in high density and squ"tter areas. Children in the rural area were short"r and lighter and werO 81so more heterogenously spread around the mean weight
"nd height thn urbrtn children. As ex"nplified in Table 2, the standard deviations an,} the coefficients of varialoility fvr both sexes and the total population confirm this.
---p
Mean observed CO). and Standardized (S) Height and Weight of children by age
Height (in MM) Weight (in KG)
AGE U R.8 AN RURAL I1RllAN
(in months)
Low High
Squatter Total Low High
Squatter Total
Dens it:/: Dens1t:/: Densit;r Densit;r
~ S 0 S 0 S 0 S 0
sll
0s
0s
0s
0s
10 - 11, 773 726 722 731
71'
724 739 723 591 635 9.0 8.4 3.3 8.4 S.? 9.0 8.7 8.715 - 19 754 741 744 736 736 738 743 738 70S 737 11.3 11.~ 9.2 9.1 9.3 9.4 9.9 9.4 20 - 24- S2l 799 805 805 797 805 808 803 700 737 11.0 10.7 10.6 1').6 10.3 11).4 10.6 10.5 25 - 29 92J 321 &24 817 812 319 ~52 818 844 851 13.0 11.6 11.5 11.4 11.5 11.6 12.0 ll.5 30 - 34 8~7 356 865 ·351 833 854 962 853 679 749 11.4 11.0 12.~ 12.6 11.7 12.0 12.0 11.9 35 - 39 938 896 891 39R 879 893 903 895 884 927 13.312.7 13.? 13.3 12.9 11.1 13.1 13.1 40 - 44 948 923 953 932 H2 932 938 930 890 1031 14.1 13.8 13.5 13.3 13.8 14.1 U.S 13.8 45· - 49 1005 972 985 985 957 977 982 974 917 977 15.4 14.9 15.0 15.0 14.3 11, .• 6 14.9 14.7 50 - 54 D26 993 1010 992 969 994 1002 991 903 1046 15.5 15.0 16.5 16.2 15.7 16.1 15.9 15.8 55 - 59 1003 lDOR 1031 1010 994 1~08 1011 1009 999 971 111.1 14.1 14.7 14.4 14.3 14.5 11 •• 4 14.4
}) Rural data was standardized differently on the basis at the combined sub-samples for rural and urban areas for each age group.
&URAL
0 S
e.o
ry.58.9 9.2 10.0 11.4 11.4 11.1 f-'
""
9.4 13.2 \II 12.4 12.5 ll.7 12.2 13.7 14.4 12,6 13.3 IS.a 15.1
niGHTS AND HEIGHTS BY AGE
FOR CHllDREM UNDER 5 YEARS OF AGE 1M ZAMBIA ... ,
URBAN HIGH DENSITY URIAN L4 .. Q.fNSITV URBAN SOUATT'ER RURAL ... u TOTA·l
'loCI·" •• \ ... 1 . . . 1
(a) OBSERVEO HEIGHTS 1 _ _ . _ t .... ..,.
"' .00 0·''', f _ .... l
,: ~: ~ ::'. ~: :~ ~
;7_ ~
no ".~ tl. no~~
:
I " 1.0 100 10! .O~ '. ' . . . . ZOO 50 4. 50 10 " " 0 10 50 '0 SO t. .n,!) 10 so 1 . . . 0 I U . to I . III so t o ' .... ,".~-. . i;;-.L.,.~.-'-.~.~.,...-.."
(II) STANDARIZ£D HEIGHTS
i~·:t~ : ~ ::: ~ '::~ ~
!;~,
I I ; ; /;;~::: ~)~ g:
'0 it 10 am!lO 40 o. 10- to 10 .0 4. 10 (O)08SERvitD ';;£'G~OTS 4.0 10 10 iO ell ' 0 I ' 10 . . •• .0 loCI I . •• Ie
'U/"
,u'U ...
,u;".
•:::
" ,~ ::: /
" . .::~.
" I:::t
10"11.~
i 'DO 100 100 ";'0
;.4 •. ... . , ' •.
' 0
. ..
"- .. -
' 0 ,... - .. - -
(b) STAHDAR'ZEO WEIGHTS , .. -
"... -
70... - . - ..
1'- • ~..
~- ....
·:"i' /
~., .:::
' 1 . 0 . /~ :::: /
,.1:::~ ::f~
.aD RI~ j;
,0 ' 1 1 rc l ! " so 4. , 'II " 10 ,~~
'0 / 10 SO 40 50. 4.0·~~
II- RO so 4.0 !IO .. o.~~
. 10 to so 40 lie ~ : : . •• " . . "-AG£· IN MON TH8
Besides, the greater homogeneity and compactness of the distribution of chilrlren in the urban area shows a limited de~ree of variability when the standard deviations or coefficients of vari{ltion for low density~ hiz,h density and squatter areaS 3re considered. Similnrly, no marked cerree of variation was found between mole and fenale childr~n in terMS of their respective distribution around the mean weight and height. Irrespective of the sex of the children, it is also clear from Table 2 that the
because of low susceptibility to diseases.
In the present study, a relative scale has been used to identify children who were underwei~ht for their age in their respective survey environment. In the first instance, children were classified in such n way"ls to be "ble to infer the extent of what is commonly referred to as
"first degree malnutrition!!. This 't-lC1S donG. by means of a men.sure of grnwth
retard3ti~n which included all children within each sal'lple stratul'l "hose weights were less than 9C per cent of the mean weight for the age group
taken as a percentar,e of the number of children in each age group.
Table 3 presents the results of the analysis of growth retardation.
Over one-fifth (21 per cent) of children in the low density areas of Lusaka city were affected by "first degree. malnutrition". The corres-ponding fip;ures for the high density areas, squatter areas and the rur?l areas were 25 per cent, 28 per cent and 30 per cent respectively. This suggests the existence of an inverse relationship between socio-economic status and the incidence of growth retardation; the poorer areas r~corded
higher levels of "first degree malnutrition". Another interesting relation-' Ghip in Table 3 is brought out by the st~ndardised percentage of children whose weight were less than 90 per cent of the mean weight for the age grouP. The st:mdardisec}) values shnw that the percentare with ret"rded growth declines with increasin~ age up to the age group 50-54 months in the four survey locations. Figure 2 also illustrates the inverse relationship between growth retardation and~::e of children in all the four loc"tions.