9 1 .. .. .9 . . ..OIH.. .9 ..f.
. . . 1..... . . . 9.1. . . . 11 9 .. . .1.91.....1191.. 91511.. .9199901

. O...

9..-9!].l 9993190,-
91:9991149991199rd991.9119 9.9..
. . .. . . .9999. ..191...I9119. 91.999
1 . -1.. .1 . l . 1999.9 0 1 9 1f: 1 9 9090'!.' 99'1Drl 9‘19 . . 1II
. . 1 . 9.1.... . 91' 9... .9 9 .99 1.91 191.1

. r 9.0.). I I.l99.’. “9.9.9.0009wt01ffiff” .
. . . . . 1 99 1 . .. 9 19 9 9 999.... I .9 I9.91 ‘1‘9’0.’ . 9..

. 1. . . 9 .. C

, is. .
PATRICIA -.

9...9910911....90.9.7.9.1l119.l1'.19.1. 1.
9¢.t9 .J1.1§

9 1 ., 9 . 0.9. . . .. 13...! . . 1 9 . 9.9.... . 1 , :9. .1.. . ._.. .. 9 ..9 . L. 1. . 1... 9 . 1 1 . . .I 1 13.019'0 _
99991.9( 99‘191.9.. 9. 9.1.90 ..9 .91 99 9.9090909199990.19H909.99.1 o .1 ...-9 .9. 1 1.1. 1. . I 9 1 .. 9O . .. ,1 , 1 . I. . 1.. .11 .9 .. .19 1 1 . 1 1 95.19.91. . 1.9.1.199. .9 9999 .9. 1. . 91.9. .9.9.9 9 1 919. (99999”.99vohw’991.91¢1 olJJ—0a41990 L’9>sa.3<r. 1*. dart”, (9" 5.99 0'. .JOJ’OdNJ. 99
V .11 . .9 .. . 911... .. . .. . . 1 . 1 .. .1. . .1...11 .1 .11 .1 . 11.1.94... ...9....9..I91..... .../.....993L1J9’J“ ..99111 .9.: 9.11.0 1.99.91.119199911 979,9» ’
.... 1199.9 .1.... 9 .1 1. 1 I . . .1 9 .1 . . . . 1.. . . .111 111919.19. 1 1. ....9199 .91.. .1 9..91 .1. 9a.. 9J9.) r:091~.919 91'99 99 .75..
‘91.. .. .... .1. ... ... ..............11...1.. ...... .1 . . 1 . r . 1 . .. .1. 1 . . . ... .........1.......... .1. 1. .1 . 11.911u.11....9119....1.1’11 ...11. 1191.99 9.919199 .9..919 1.11919" 11.1919 ourxf/aufid.i..9rfodoa 9..... .
19 .9 1 .1 .1 9 . . I . . 9 1. . . .. . . . 1 I . . 9. . 9 . 9 . . 1 9 9 9 9. 9 9A 9 9 0 0 . . 9 . .91
1 1 1 1 . . 1 1 . . 1 1 .. . . 1 99. 9 9. . 9 . 1 . . . 1 . ... .. 1 . . 1. . . . 1. 1 1 . 9 . . .1 1. . 1 . .1.. . . 1.... .. 11.11 a. 1 .... 9. .9.99.a 9 1. 0 399999 . ..9 14.00.9r119a04..9!9.0 9*119999./.9911199 99 0 9;.0a9r9 (9 9.00.0aomo”90mu010099 90919,
...-1 . 9.11... 9... 1 . . .. . 1 1. . o . .. 9. ... 19.91. . 1.19 1 9 .1... .. .999.....91.1t.9..0999.. 2.9.9.1951}... .9..999..ldx90..90 9 919 9 119.19911J9’
W 139... 9 r.1. 9. . 99 1 19....... 1.1 1 . . 1 9 . .1 . . L . 1 ... 1 9 . .. [.119 . 1 .1 11. . .1.-1.....9 .11.. 9.9999990u.r-9..1.:11.199994.9 91‘0“O1.9999 49.0194/O/‘9ikw .1?J0”0.05199£’99J0k 00.9"0
1 9.... .1 1 1 . 9 11.9. 1. .1 . . . .. . . . . ....1 . . 9... . .1.. 9 99.49.99. 991.. 9 9.4.9991...’ I ’9... .1" 9
.19.... . 1 . . . . 1 . 1 . . .. 9 . .1. .1 1. .1. ....1 .. .1 .. .97 9\1(.9.99¢. 99 9 99.. 9 9. 1 9 {.9’
so . 1 1 1 . .1 1 1 19 . . 1. . . 9 . ... .1. . 1 1 1 . 1 . 1 L . . . . ... 1 . . . . 1 .... . .... I. 9 .9 1 1 . .. .1....99a. . 9 99 09.9 997999994171... 0&4“. .9U9‘99lfnvil/n 9.4... I. 9....
‘1 .9 1 9 911.1 ...0 1.0.. .0 . . 9 . . 1 1 .. .9 1.. .....91999....0.90 .....01...9.99.J9.— 9.9- 9.199 0,001.31“! ~09/I '..'(J'
. ... . _. . 1 . 1 9 1 1. 19.1911... . . . . . 1 . . . . . . . 1 . 9.1.1.11 .. ......1 .1. 31h”... 1.... 9999 . 9 5999.49.91.94 91.» . .9!“N.1.9.. “99"” 9’... 1'9! 9/.999ru
. 1 .1 1 . .. . .90 1 9. 1 . . 9 9 9 11 1 . . . . . .. 1 ‘. 9.1 t a 9 .
. . . . . 1. . . 1 9 . .1 1 . ... .1 1 1 . . 1 1. 11 . 9 . . . . .1 . . 9 9 a. ’99. .“m
.. . . . . .9 9 . 9 . .1 ... 1. s . . _ 1 9 1 .1 . .0... . 91 1 .1 . .1 . . 1 1 . . . 1. 1 . 9 9 _ 9. . . 1 1 . .3. .1. 9 1.. . .1 9 9 9 9 0 . I 9 .9.»..9o.o’van.a’n..r. “9090MI‘ “an/1 9?.9’.“(.” 9““NH99999 -
1 .. 1 . 1 . 9 . 9 _ . 1 . . ... 1 1 .. 9 . . . . 9 I .
. . 1 .. .1 1. 9. . 9 1 1. . 1 1. . . 1. . .. . . . 1 .. 1 . . . . . ...9 .1 1 . 1 ..9. 9 1 . 1 11.19. .9 1.117.1119.9991..9.1... .9...” .1919. 9 ..9 3.197999 V1019”... x91 ’HiLgé 0 ”V9060
. 1 1 . 1. I . 1 . 9 . . 9 99 1 . . 9 1’ ‘9‘
.. 1 1 1 . 1 1 1 . . .1. . . 1 . .1 .. 1 . . 1 . .1 .. 1 . . 1 1. . 1 1 1 . 1. . . 11.111191. .1... . 1 . 4.1....1u19...79.9l99 1.1.4.1....1911911999W111. 4.1.9.9999...- . ”I",
. 19. 1. . . .1. . 9 . . . 1 . 9 .. .9 9 . 9 . . . .... 199.9 forfwyxl 1h... .II.I
1 . 11 1. 1. 1 1 19. 9.9. .1 . . 1 9 1 9 1 . 1 9 . 1 . . . . . . . . 1 . . . _ .1 . 1 . 1 . . . 1 . . 1.... . . . 0989 ..19Hwek9m9191u11’l09Wd6w1JJ.’ 90C ".91.. 1
. . .. . . . ......1 . . .1 .. 1 ... .. .... .. . I1. 111 .. ....
. 99 9911.19.99 . 1 ...0 1 1.. . 1 . 9 . . . . . .. . .1 9 . 1 . 9 . J. 1 1. 1. . . . 1../..rflaw ”Hug-991.91,..fl” AA”.19 90I.V1’.. 99J
.1 1 1 1 1. . .. . . . . . . . .. . . . 1 . . 1 . 9
0 9 .1 9 1 .1 9 19 . . 1. . . . . 1 . 9 . ..1 . 9 9 9 ._ 19 . 19.9 . .1.. 9.... . . .99. oul’t m", .“1f’99urufuwr/0/OIfV0 9'“ ..91‘0'.19Fri.19.”.9' “It! .f "Ann“
1 9 .1 1. 9 9 ..9 1 9 . . _ . . . J . . . . O V .6 999 1
. . . .. . \ ... .. .1 .. . .. 1 . . . . 1 . . .. . . . . 1... . .. 1.1 . 1. ....9~.1..“‘19.1”.19./.9.d.b9..1 11 90.709.119-7Eo1 . 9V
. 1.. 1 . .. . 1 . . . _ 1 . . . .9 . . . 1 . .. .. . .9....1..1.... 9 .9 1.1.99 .9 9.1.9.9”.u..9 “..9.
. . 1 . 1. . . 1 . 1. 1 . . . . 1 1 . . 1. 1 . . . .999.9..99,99 f. 99.3fl919...79.&9 09. («’91 . J‘f‘
.. . . . . .. . . . . . . 1 1... . , . .9...J.’94. 999’? .1 99.9.r1 f. .
_ . .. .. .. .. 1 . . . , . . 1 . .. ..... .f. . . . . 99.99 99.199996 4. 70.....4. 1'09,” ”r .
. . .1 .. .11.... ..1 . . .. ... . . 1 .. . ......9..J.. i1-9l9919. 99““) 1.9 9 filo.
. ...10 .1. 9 . . .. 99 9 9. . 1. .. . 1 9 1 1 9 ... . 99. . .. . 119 1 . I 091909. a. .9199” ' .... v .
.. 1 . 1 1 . . . . 1. 19.9 f 9.. 9/91- 29.99 ’9109.9. .7 W99.
. 1 .1. . . . . . . . . . . 1. .9 9 .9109)? .f f :49... 999 9“”; ”:1, 9 99 1 11
. . 9.. .1 .. .. . . . . . . . . . . 1. . 1. 1 .199u9 9199 1.» .19 99.99.91. . 41:19.. I. 1939. 7 fly.
1 1 9 9 1 1 . . . 01. l 9. u . 9 . 9 .
. 1. 9 9 1 . . 1 1 . . .. 1 . . . 1 . 9 9919. ..OI.r/..9.'9.R1799!".9999914 I’. 9 pfrP'l’.”
91... . .1..<9..11199o . 991.9. I’m/19.91999 OK .9919 4.791999.
9 . 1 9 1 1 . . . 9 _ . . 9 . 1 . q 9 O .11999991.9.o.19r11.99.919’0091019 9 9.91.19”! 90IIIUV'71ve'19t‘99.
. 1 9 1. 1 . 1 .. 9 . 1. . 1 .1 9 .99 I . 1 1 . .
.v 9 . v 1 9 9 1 . 1 . 1 1 . 1 . . 9 . .. 997 . 9 .91 . a .1.”...9 “9’... 9. 9’9, 0“,... . 9 I 9 919u9 9 f 9’90 ”999'” .0 9199“,? CIV’J/’¢§:'
1 1 1 . . 1 .1 . ... . . .9. . . 99 . . b 1 .
1. . v. v . . 1 1 91 . 1 . 1 1. 1 . .9. 1 9 1. . .. 1. . 19. 1. ”399.. . _ 1......f...19-.1991..1_l991‘9»1w0991.4?9/9f1l9rf9c. '9 U ” Oirviw.
. .. 1. 1 1 . . . . ... . . 9.... ....1....991...1.1 11.....91.-. r/9r47199199.111.’.19..
_. ..9 1. . 1 1 . . . 1 .1.. 9 .....L...919. .9....91t1...1.9 199.14....971fv .' 9 4.9.I99'Ir911'95
1 . 9 . . 1 .. 9. .9. 1 a 1 1 . 1994.9.099999 0.90.9091 9.99....» .94 'M #13111. '9 .19,
v . . .1 . . . . .... .. ..9. ........1.1.. .1 11.. ..1...11...1I.11.9.1 r¢fr ’99. 9.091
. 1 . .. . . .. . . . 1 R .. .1 1 . .919 .f . . tau
9 . 9 . . . 9 9 9 . 1 1 . . .. 9.. 9. .9 . 9 1 1. .90... 9 1.19;. 9.0.9.9109991.1I99..v.’9a91.9.19'0 '0 . 0.999'fl04f 419.419.990./Ir’tf’O'/9/¢4 IOIIQ'TUQl g 'l
. 1 . . . . . .1 .. .... ... . . .. ... 71911.9. 9.11 1.....1191f91./..1999..~ .e’.».0..t9/19.919H11.1/flr99.
_ .9 . 1. . 1 1 . . 1 _ . ..119 . .. .... ...r..1.l..99’..1l.../..99.9rt11.199.1 9011919r/19.199.1/ I!
. 1 1 . . . _ . 9 . 9 1 1 . 9. . I. 0 0 '11!
. . 1 . . 1 1 9 . . . . . . . 1 _ . . . . . 1 . . . . o. 1 .1 , .... . . .9. 1. ......99...1.J1.9.9.1f1.9_.9r9.1.v“.t.-.91.¢9...19’90dgf1rm191;9.1’k. pvzffifrla..b
. . . . . . .. . . .. . 1. . 1. 1 1 . .
fi . .9 1 . . 1 1 . .1 . . . 1 1 . . . . 1 .. .. 1 ..1’119191.91....9. ’9 f.99’y19..9’¢r-J”/.1 91999 1.. 091.9%” .14..9.9\9i1r19..l gr J’ry1h9 ....900941
. 1 . . . 9. . .1.0 . J o 9. . 1 1 ‘.O.r.91 9 1.11 I
. 1 . .. . 1 . . .1 v . . . .9.. .. 91.1 1.1.4111... . t . .1.! 12.9.1199 .9’./vl/w«y1,”10910..1 I 9
w . . 1 . .9 1. 1. r..-.....9..1...l..4 .11... 1......1l..191.f./9.f1.n:9 ”19.9.1.9... 999.01 9'1
. 1 . . . . . ..- .1 9 1.1.9... 9. 9.... 1— 1. . ’1'..11.991
. . v . . . ... . 9.. . 1 1.1. . ./ I... 1.1.11.1...5. (.../... ....(49 ..Qv.’7...71...r..r9r191h'g(
. . . . . . . _ . . . . 1 .. .. . a . 991919.... . 9. 1.... .9991rl ..l9..\. . I99
— 1 . . .1 . . 1 . 1 1 . 1. .1 1 1 1 .1 1
. 1 . . 1 . . . . . 1 .9 . 9 . . 11. . K9..99.9.._'..1..111 1.9.1.19. 191.9... 9 .9. . ......ttr ruuu1‘M19'le99I'J91990171991109‘ 99!
v _ . . . . . . . . . . 1 . . . ... . . .. . . ... .119. . . I ..O. . ..1. 11:11.1. (9.11.1109. 9.119.11'r01911310 In.
. . . .. . . . . .. 1. 1 1. .... .1... .1.. .141
V . 1 . . . 1 . 1 9 1 1.. 1 .9. . 1......911'. .9..9991!9 ..Ln'i'huty'lgnvdwvg
1 .199 . 1 99 . 91 999'...— .1909 191%91. r .9 9... 1111.11 9’;
9 t. . . 9 9 . 1 1 9 . . 1 .10. 1 ”1.9 9.9019 1 1 999 '9 9 .‘Jfi’i‘."..3}t,
. .19 9 1 . 9. I 91 .1 1.
1 . 1 1 . 11 .1 9 9 1 9 . 0 9.1 99 911
. . . 1 . 1 1 . . . . . 1 0 O . .l .9. A _ . 1 9 9 ..999 9199.91. .09‘ 9.03994109l If, W'I!€9”l9l
.0 . i. N . . . ... 11 1 .. .1. (9../11.9 1....19. ...9/l9’1u'11191.999 1.9919..1O9'.9
. . . 1 . . 1 .. 1 11 1 9 9f (01" .
v . 9 .i: . o . 9 1 . . 9 919‘ 1 I. . 1 01110 1 1. 9.1.9.9.. 9 19911 999 9 1.9.9., 0"...999
v . .. .1 . J 1 . . . 1. 9 _ . 1 . ..1 1‘1.91).-.99l90991919.r.9.l"1
. . . 1.. .1 . . 1 / 11. 91.1.19. 1 . 9. ...99..99199.19 49'. ‘19119091.99.
m . . . . 1 . . 9 9 . 9 I. .99 1.19.19 . .09 90 19’01 99.10.9..l.”lo1.l'.l90190
9 1 1 19 1 1 ..9 1 . ... 1. 1. . . .99 9 9r 11.9“! 01999? 1Jllr.1...
1 1 . 9 .. . 9 1 1 .9 9990!] ..9 .1. .099! l1 9 1 . 911
u f . . 9 . 1 9 . . 1 9 9 11 9 1 f 99 I . .1 . 0 .9 0 u . . 9.991090 fl9'f0 1.99.1019”... '-
fl .0 1 9 . 1 . .1 1 1 . .1 . 1 .. 1.9.. I. . I .1 .9 . .v .r .
. . . . . 1. . . 1 . 1 1 . . _. 1 1 n. ... 111.14 .1....31111h
“ . 9 1 9 .1 1 09.. 1 9. . 1 9 1 1. 1 9 . . J1. 1‘ . 9.0 90u1 990197.. F I»: ,9 .1
. . . . . _ 1 9 1 1 9 O. . l 099 1’9 1 r I. I1 '11 1'
q 1 . ,. . e . 1 .1 1 1 .9... 9... 1’9. 11. Ira..ln1f1l . I
. . , III . . . .. .. . T... .. .. .. 1 1 .1.1m..... 011-32.“. . .
1 . . 1 . N ." 9 9 . 1 . 1 . . . 1... 1.1.1.1.111 1 .1. 99 1v 9 . 919...! ff 1H../1I... .1.19 0)....1w.v199;\..w1l
O . . . 1 ’1 ... . 1 91...... 9 9 . . .r . 9.1.9H14r_0911;9.u. 1....919... .’
. 1 1 . 1 . . . 1 ..9 1 1 9 .99. 1 .91999 . 1.1.! 001llafll# 41(91919r1r0’91d1.1.1r
1 . . . . . .9 1 . . 1 1 ..I 9 . 1 . .
_ . U 1 1 ..9 .1. .9 1 .. .9 1... ..1..11.9 ...9.......9' .9 J: 91(‘1111'1991I
. 1 . . 1 9. 1 9 1 1 . . 9 9 9 ’91... 1 #9. r0
1 1 . . .. 1 1 1 .1 ...9 . 19 1 1 9 91 . 1 1 1| 9019.9..O111 . 9.01.11 .1
. e E .91....- .. .. .. 1.9.1.1...“
. D . . .. . . ., 1... -. ...... 1..“. 1...... .
. . _ . . 1 9119r'91 99.9901 1 1‘
. 1 . .9 91 1.19 91 I .19 1 915.1101‘DI1-V. '
_ e . .. . 1.1 1 . 1 91 . .929. 1 . It 9 .1.91'..119.r.(...f1/1'..9111 .I. 1141.19.0’1J.’.9.9...,1 .
. ..nn . 1. .

 

1 . 1 .9.. .1 1 .1 1 .. 19.99....t11.1.911
.1 .1l9.1. . 9. . .9 119.. 1....99.19.llf....9.91.991’1 1
I. . 9 v . 1 1 11 9 . . .1 . 19 1 1 9’0!" . 190991....
' . 9 9 1991.1 1 1 .9.. 1111’.9.. 1.9.119 1.911991. 1 1
o s . 9 1. I . 1 19 1 991.1 1 1.9191 1910999009 99-
‘. 1 91. . 1 I991 .. 1 . 1 9 11,99! . 1 1.9.... 10911.. 01
9 . .9 1 11 1 1 . 9 II. 1 917/199 919911n1r9" 99
I. 1 1 I 9 I1 9 1 d 1/1 .1... 1.4.0O .... .9 19.199119 1111. '
9 1 . . . 1. . ,l . I . 09 It... '9 1)I9 . ‘. "0.' ..
. 9. 9 1 .1 ..9) r I .1 1 I 9019 1.91 9 .'..l:'101 9’1
.. 1 1. . 9 1 o. ... .11 1 9(I1119Jtr.
. 1 11 1 9 1. r . 9”) . . 911/. 11 (A
. . .I: 1 Il1v1 1’ .9 1.91. 91'.
1 9 .9 1 9 0.. 1 17.1 . .11 1 J9 911.00
. .I .. . _ . .. 11.... 1- -9. .
v 9 1 9 J V .9 1 19 '3'. v
. 1 1 99 r 1 1 , .9191. f1 1
. 1 .. 1 . 9 . . . . 9 ...u110’r .9 ..
1 . . .. 1 .. 1 11 .9. 1 9 .../..1‘.’..' 9'9 |1 9

EU)“ FEEDING AND. TOD

. .. . .1911.4 1

 

 

. 1 9 I9. 9" 0 I I I 1
H 1 1 9 1 9 o 9 '7
7. . 9 1 11 1 1 9. .9 I 1! 1 9.
. 1 1 91 9 . 1 r . 1 .1 1 I
1 1 1 1 .91 .. .19 9.. 11
v 1 . 1 1 1 l .91 1
A. 1 1. 1 1 9. 1
v . . 1 . . 1
. 1. r. 1 1..
w 0 I 0 '
1 1 1 1.1 1 I I.
~ I 9
1 1 9 99 I
1 9 9 1 n
.1 9 1 . 10' 0
w 9. '1 9.9 9 l
9 9 9 . 1 0' 9
“0 1 9 . 91" 1
. 1' 9.
I . 90.1 1 I 1 1 01 9'
w 9 1 l9 1’
w 1 9 . 1. 1 1 .. 0 19. 9
1 1 t 1 . ' I
I I ll . .1 O 1 .1 1 1
_ . . . . 1
9 1 9 1 . 1 1 9 1 1 9-191 1 . . .I ..1 .10
v 9 99 I 1 9
V 9 9 9 9. 1 94 9. 1. .9 9"- 1
. .
v 1 .. 9 1 1 1 99 11’ 1 ..9 9 1
fl 1 1 9 . 1
9 . 19 1 1 ..9 I 1 9.?9119. .0. .
1 . 9’.
1 19 9 9. 1 1.1 .9 99) 199
h . . - -
W .1 1 1 I. 9' 91 1 '1‘?
. . . . .11 . 9
0 0.
. 1 9 . 1 1 . .9
v ’ 1 ’1
1 1 11 1» 1 .11 . 1!. .
‘.. I 1 . 91 1 91 O 11.n’ 1.
. . 1 91 . 1 . .1 . 1 ..l 1
_ . . . . . . . - .1.. .1 -.1 1
. 11 99 It
_. '1 1. 9 1. . .1 1 1 1 r. 1
A . . 1 1 9| -.1'ftb 1 |
w . . . .. . 11-
1 . 1 11 1.1 19
A; 9 I- vlo 1. .1
1 9 9. fl. 99. 119
W 1 . 19.. 1 C
1 A 1"-
w .. . . . ,1. . 1.9 . .
1 1 . 99
l 1. 1 '1' . r1 1 .9
A 9 1 . 1 r 09 9
. .. 1 9 ...99 . . 1 . . 1
9'. 9
. . _ 1 a 1 .9 1 1
1 1
. .

 

   
 

D 119 I 1 . 1 1 .1 .I 0 I ‘1 ll.
1 . . . . . . ...-1.91. .12.... .9 .1 1 1.4....1...11.1I.w¢. 1 . . . . a 9 1.. .9 .. 1 91. . .u .uudf...b.0)!91..s.091.11 . .71199. ..9». ..99. .1 9.9! 9199-éotfifla.‘99‘1. .14..9 990 .9‘.11u1.4.§‘w01‘10...A 99.1.0.0 599 9.111. ‘09.”.Y‘%0s9lV919¢9INH09MANO§9NQVL31IITI

.1\1.1

       

1 1 9 1 . 91 9 O1. 1 O19 . 91 9109'
. . . 1 . 9 1 I DID 09.9.9.‘9c01.'9.110“.§'0ugon.§

9.

W . . .1.

1 . . . 909 g.
$11901. \1 .... 11.1.1 .0.. . . 1.. 9 . .91.- .1 .1....1 “1.5.1 1.911.910.3111}. '1 91.9 111191.91. r 91. 91...v.‘.11.1.199. .091”... 191 101......» 1 .rr a 1 1 .... 11‘ 1.. r1... .1\. 919.9999 9 1. 9 1 1 ..9 09% 9 1.1 9 11.113.031.91 \‘19.91.99.9 9.9 :9’I9r91l0999‘99’r9. 990.9 9.099.991ukzkafl .9 .-

9 '

a la}! mug IHZIIQIJIIJIIII Lu: 1;! m 1H: m In W! ll

 

tW-h-W/

 

ABSTRACT

CHILD FEEDING AND TODDLER MORTALITY
IN SELECTED YORUBA VILLAGES

BY

Patricia Ladipo

Malnutrition of children resulting in high toddler
mortality rates and suboptimal development is one of the
major problems in developing nations today. Several studies
in Yoruba villages have pointed out high child mortality
rates and low nutrient intakes of children.

In this study of child care and nutrition in rural
Yoruba villages, present child feeding practises, household
dietary intakes, attitudes towards foods for infants, and
child mortality rates were uncovered. Child feeding practi-
ces were related to child mortality rates. The relationships
between the ages of mothers and child mortality rates and
child feeding practices were studied. Responses given in
two sub-cultures were compared. Variables relevant for
further studies in child nutrition were uncovered and possible
approaches for an applied nutrition program in the areas
studied were suggested. ‘

Literature is reviewed on the topics of malnutrition
and toddler mortality, child mortality rates, the age of the
mother as a variable, child feeding practices, intakes, and
related cultural aspects.

Eleven villages, located close to the cities of Iwo

and Ife in the rain forest belt of the Western State of

Patricia Ladipo
Nigeria, were selected. All households in each village were
included in the sample. Within each household, one wife was
interviewed according to a pre-set schedule of questions.
One hundred and seventy-four women were interviewed. Data
were analysed using percentages, the chi-square test, and
the Pearson product-moment correlation.

Most of the women were married to farmers, were of
child bearing age, were illiterate, and worked as petty,
traders. Very few had access to radio programs and no one
had ever been contacted by a home economics extension agent.

Breast feeding is the preferred feeding method and
goes on for more than two years when most babies are weaned
of their own free will. Supplementary feeding usually
begins by six months and, at first, consists of cornpap.
Very few superstitious beliefs relating to foods for babies
were found. Mothers lost almost one-third of all live
births. Over half of the children who died were toddlers.

The following hypotheses were supported at the .05
level of significance: '

- One-to-four-year-old mortality rates for mothers
who give only starchy foods as the main supplementation will
be higher than for mothers who give infants a variety of
foods. I

- One-to-four-year-old mortality rates are positively
related to the ages at which children are first introduced to

stew .

Patricia Ladipo

0 There is no significant difference in the propor-
tion of children lost as one—to-four-year-olds by literate
Vand illiterate mothers.

Hypotheses predicting no relationship between age of
respondent and toddler mortality rates and child feeding
practices and attitudes were all supported. The hypothesis
stating that Iwo and Ife mothers would not differ signifi-
cantly as to the type of supplementary food given to infants
was supported.

However, hypotheses predicting relationships between
.toddler mortality and time of weaning, distribution of animal
'protein, and number of foods thought appropriate for babies
were not supported. Likewise, hypotheses predicting no
significant differences between responses of Ife and lwo
women on number of foods thought appropriate for babies, the ,
age at which stew is introduced to children, and weaning
before, after, or at the same time the child learns to eat

an adult diet were not supported.

CHILD FEEDING AND TODDLER MORTALITY
IN SELECTED YORUBA.VILLAGES

BY

Patricia Ladipo

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Institute for Extension Personnel Development

1971

ACKNOWLEDGEMENTS

I wish to express my thanks to Dr. Mason Miller, my
academic advisor, for the encouragement and guidance he gave
throughout my study program. Special appreciation is also
expressed towards Dr. Portia Morris and Dr. Irving wyeth who
gave their time generously to guide the writing of the thesis.

Mrs. Helen Gifft and Diva Sanjur of Cornell University
and Miss Jean Ritchie of F.A.O. gave invaluable advice and
encouragement regarding the content of the questionnaire.
Thanks are due to Mr. D. O. Oyeyemi, Mrs. J. Jaiyebo, Mr.

A. Adebayo, Mr. M. Olajide and Mr. Mojeed of Ife University
for their help in the translation, typing, and mimeographing .
of the interview schedule. The cooperation of the Department
Of Extension Education and Rural Sociology at Ife University
in introducing interviewers to the Ife villagers is greatly
appreciated. The interviewers, Miss Idowu Asajuegbe, Miss
Ayoni Ladipo, and Mr. Olasupo Fabiyi, were tactful and cheer-
ful in their work. Special thanks go to them and to Mrs.
Michael Abkin for her help in typing the thesis.

Without the cooperation of the village leaders and
women, this study would not have been posSible.

Most of all, I would like to thank my husband,
Olasupo, who encouraged me in this study from the time it

was a half-formed idea to the final proofreading.

ii

TABLE OF CONTENTS

LIST OF TABLES
INTRODUCTION

REVIEW OF THE LITERATURE
Malnutrition and Mortality
Mortality Rates in Other Areas
Child Feeding Practices, Attitudes, and Intakes
Age of the Mother
Yoruba Culture

METHODOLOGY
The Questionnaire
‘Sampling
Data Collection
Statistical'Analysis

RESULTS
Description of Respondents
Feeding of Children
Dietary Intakes
Attitudes Towards Foods for Babies
Mortality of Children
Testing of Hypotheses

DISCUSSION, IMPLICATIONS, CONCLUSIONS
Discussion and Implications
Implications for Nutrition Education Programs
Conclusions .

BIBLIOGRAPHY

APPENDIX
Questionnaire

iii

Table

 

\IO‘U! .8

m

10
n
12
13
14
15
16
17
18

19
20

21

LIST OF TABLES

Mortality Rates in Rural West Africa
Child Mortality in Osagere
Child Mortality in Ilesha Villages

Percentage of Ibaden Children Given
Supplementary Foods

'Supplementary Foods Given to Ibaden Children

Ages of Respondents

Occupations

Education of Respondents .
Exposure to Outside Sources of Information
Preferred Feedinngethods

Types of Liquids Fed by Bottle

Reasons for Using a Feeding Bottle

Age of Weaning

Age of Introduction to Adults' Foods

Age of First Supplementary Feeding
Supplementary Foods Mentioned

Age of Introduction to Stew

13
13
29
3o
31
31
32
33
34
35
36
37
37
33

Women who Felt Foods Were Inappropriate because

Babies Have no Teeth
Women who Felt Some Foods Give Babies Worms

Women who Felt Some Foods Cause Piles and
Loose Stools

Mortality Figures

iv

40
4O

41
42'

Table

22
23
24

25
- 26
27

28
29
30
31

Oneeto-Four-Year-Old Mortality in Relation
to Time of Weaning

One-to-Four-Year-Old Mortality in Relation
to Distribution of Animal Protein

One-to-Four-Year-Old Mortality in Relation
to Type of Supplementary Food

'Age of Respondent and Toddler Mortality

Age of Respondent and Infant Mortality

Number of Foods Thought Appropriate for
Babies in Two Communities '

Age of Introduction to Stew in Two Communities
Types of Supplementary Food in Two Communities
Time of Weaning in Two Communities

Summary of Hypotheses Testing

44

45
46
47

48

49
so
51

INTRODUCTION

PrOper nutrition is basic to optimal health and, as
such, is a subject of world wide concern. Coursin writes,
"Undernutrition is the major problem affecting the well
being of mankind in the developing countries of the world
today." (10:1)

Malnutrition is especially detrimental to humans in
the early stages of development, in early childhood. ‘Many
do not survive early malnutrition. The high mortality rates
of one-to-four—year-olds in less developed countries are
thought to be related to poor nutritional status. For those
who do survive, deficiencies during early development may
have lasting effects in terms of height, stamina, and brain
.function. In a review of research on undernutrition and
brain function, it was concluded that, "If present estimates
of the prevalence of undernutrition and possibly related sub-
normal brain function are valid, the upgrading of mental
ability by 10% to 20% could be the critical factor of bringing
children in the developing countries to a level of performance
that would immeasurably advance their productive capacities
for self and nation." (10:13)

Child feeding habits are complex, the results of

economic, sociological and cultural factors working together.

Knowledge about the background of the mother, the availabi-
lity of foods, traditions, and food related beliefs and
taboos helps build an understanding of child feeding prac-
tices.

In the Western State of Nigeria, as in other less
developed parts of the world, child deaths are common occur-
rences. Everywhere, one sees children wearing bracelets
and amulets and bearing names like, ”We live here on earth."
These children are 32153, believed to have died and returned
to the womb time and again to torment their parents. One
cannot be in contact with traditional Yoruba families without
observing the loss of many children and particularly toddlers.
Malnutrition is known to be a major contributor to toddler I
mortality (4, 28, 29, 17, 42). Even casual observation of
Yoruba mothers pouring corn starch down their babies' throats
suggests nutritional causes of mortality and points to the
need for more applied nutrition programs.

- FAO and WHO have emphasized the need for collecting
adequate baseline data prior to the planning and evaluation
of applied nutrition programs (50). This study of child
care and nutrition in rural Yoruba villages was undertaken
with a view to: t

1.. Uncover present child feeding practices, family
dietary intakes, attitudes towards foods for infants, and
child mortality rates.

2. .Determine which variables and particularly which

child feeding practices, if any, are closely related to

child mortality.

3. See whether mortality rates and child feeding
practices differ according to the mothers' ages.

4. 3Compare the responses given in two sub-cultures.

5. Determine which of the variables are relevant
for further studies in child nutrition.

6. Determine possible approaches a nutrition pro-
gram might take in the areas studied.

In line with the objectives stated above, data were
collected on the ages, educations, and occupations of res-
pondents, their exposure to sources of information outside
the village, the history of their child bearing, infant and
child feeding practices, their attitudes to baby foods, and

the types of foods eaten by the respondents' families.

REVIEW OF THE LITERATURE

There has been no study in Yoruba land which system-
atically relates the variables chosen for this thesis. There
has been work in the Western State of Nigeria and other less
developed areas which relates to some of the variables
studied here. These variables include: malnutrition and
mortality, child mortality rates, age of the mother, child

feeding practices, attitudes, and intakes of children.

Malnutrition and Mortality

 

Williams describes malnutrition as an iceberg, with
only its tip being apparent in mortality rates (52:5). A
high child mortality means there is high morbidity. ”Mor-,
bidity is a physical, mental, and economic burden on the
.community" (53:346). Others support the view that mortality
of infants and young children is an indirect indicator of
the general level of health and nutrition in a community
(12:307, 24:374). Ahmed and Van veen found that people
consuming more complex diets lost a lower percentage of
their children than did those with simpler diets (1).

While malnutrition may have some bearing on neo-
natal and infant mortality, it becomes a more important con-
tributor to death after the first year (17:376, 42:112).

In less developed countries, mortality of one-to-four-year-
olds is often more than half of the infant mortality.

4

5

Deaths of one-to-four-year-olds or toddlers may equal or even
exceed infant mortality (6:199, 17:358). Wills and Waterlow
explain that infants and toddlers are equally exposed to
disease; the difference in their mortality rates is due to
malnutrition of the toddlers. They have proposed a mortality
ratio of one-to-four-year-olds to zero-to-one-year-olds as

an index of malnutrition (54). Using this index, Gordon .
writes that if the ratio is less than 5% it indicates good
nutritional conditions. A 5% to 10% ratio indicates inter-
mediate status and a ratio from 10% to 40% indicates a
deficient nutritional state for a community (17:362).

Looking at the second-year death rate and food intake
at the national level, Gordon showed that protein intake was
negatively related.to the second-year death rates up to
intakes of 40 grams/capita/day. Fat intake was even more
strongly correlated with second-year mOrtality, perhaps be-
cause fat intake is related to standard of living. There
was relatively little relation between caloric intakes and
second—year mortality (17:375).

Many others accept a high toddler mortality rate as
being indicative of a high incidence of malnutrition (29).
'Malcolm reports a dramatic fall in the mortality of toddlers
in New Guinea after a nutrition education program taught
earlier.supplementary feeding (28).

Although high toddler mortality rates and the pre-
sence of overt nutritional deficiencies often occur together

(4, 17:372, 28), not all of the toddler mortality is directly

due to malnutrition.

‘Hendrickse writes, "After the neonatal period, mor-
bidity and mortality occur principally as a result of in-
fective disorders interacting with malnutrition." (20:253)
Abundant evidence for the interrelatedness of nutritiOnal
state and infectious diseases in Africa is reviewed by
. Lowenstein (24).

Scrimshaw et_al in their field study of nutrition and
infection in Guatemala found that giving supplementary food
to young children in one village reduced disease incidence
and severity as judged by duration of illness. In another
village where medical care and improved hygiene were provided
without a nutrition program, disease incidence was not res-
tricted, although case fatality was lowered (41). They found
that the incidence of disease was directly related to the
nature of weaning and the adequacy of the food proVided (41).
Furthermore, the incidence of major diseases reached its
peak in the second year of life at the time of weaning (16:436).
Clinical examination of children showed that case fatality
and complications were related to the poor nutritional state
of the patients (ibid.) Diseases where malnutrition was seen
to play a part included diarrheal disease, measles, chicken-
pox, and particularly, malaria (17:372).

While poor nutrition seems to lower resistance to
infection, infectious disease has been shown to precipitate
nutritional diseases. Collis randomly selected Kwashiorkor

patients in Western Nigeria and found that in almost all

7

cases, the actual symptoms of acute Kwashiorkor had been

 

precipitated by infection (9:218). In other areas of West
Africa, child mortality rates are high even where overt mal-
~nutrition is rare (15:49, 9:224). In Gambia, McGregor found
that poor care leading to infection was likely to be the
major cause of toddler mortality. He felt that disease pro-
bably caused malnutrition, but he did not look closely at
dietary intakes (26:1661-1666).

Whether malnutrition or infection is more important
as a contributor to toddler mortality is not an important
issue. They have been shown to be interrelated. Two pro-
grams in West Africa, using a team approach, have demonstrated
remarkable reduction of toddler mortality. Wilkinson in
Sierra Leone, used a combined program of prophylactic drugs
and vaccines and prevention and treatment of malnutrition to
reduce the mortality rate of children up to the age of five
from 332 to 156 per 1000 live births (50:172). A similar
program in a Western Nigerian village reduced the infant
mortality rate frOm 295 to 72 per 1000 live births and the
toddler mortality from 277 to 43 per thousand living children
aged one to four (31:82). In this program, mothers were
taught to introduce supplementary foods by six months at the
latest, and to introduce protein-rich foods by nine months.
Bean dishes were to be given first, then once accepted, to
be followed by meat in the same quantity as that taken by
adults (31:85).

Seasonal factors have been shown to have some bearing

8

on toddler mortality. McGregor found that two-thirds of the-
deaths of young children occurred in the wet months (26:1662).4
when malaria is more prevalent. Thompson felt that seasonal
factors may have been more important than weaning, or that
weaning in the rainy season may have tipped the balance

against a sick child's recovery and reduced resistence to
disease (46:132). In Nigeria, the rainy season comes direct-I
1y after the ”hungry season" in April and May. In Eastern
Nigeria, Dema and Osama found that fresh meat consumption in
the hungry months is only 1% of what it is during the harvest
months. Consumption of dry fish, dry beans and seeds, and
leafy vegetables is also reduced (13:60). It may be that
increased prevalence of disease following a period of nutri-
tional deficiency is responsible for high infant and child
mortality rates in the rainy season.

Overall lack of care of toddlers may be responsible

'for both poor nutrition and increased exposure to disease.
Wills and Waterlow felt that infants and toddlers are equally
exposed to disease (54) but this may not be so. Babies are
usually carried, but toddlers are allowed to crawl and sit
on the ground, sometimes eating dust as described by Williams
in Ghana (51). This would make them more exposed to worms
than babies are. In Gambia it was noted that toddlers are
often left home when their mothers go to work in the fields.
During the time mothers are away, they may not be properly
fed or cared for (26:1665).

9

Gordon writes that more emphasis should be given to
mortality in the second year than to infant mortality. By
the time a child reaches his second year, he has survived
the hazards of infancy, he is loved, and he represents an
investment. Furthermore, he responds well to preventive

measure5~(l7:377).

Mortality Rates in Other Areas

 

Before presenting child mortality rates for the
study Villages, it might be well to have a look at rates for
some other West African areas.

Gilles presented the following table (15:73):

TABLE 1

MORTALITY RATES IN RURAL WEST AFRICA

 

 

 

 

Rural Areas of Neonatal/ Infantile/ Child

West Africa 1000 1000 (3 mo.-4 yrs.)/
‘ 1000

Akufo (W. Nigeria) 50 109 430

Imesi (W. Nigeria) 78 114' 572

Gambia ' 54 134 400

Senegal - 172 344

Guinea ' - 217 41s

This shows that most of the child deaths occur after infancy.
A closer look at Gilles' work in Akufo shows that most of the
‘ deaths in the first two years of life occurred from 15 to 20
months of age (15:72). I

A slightly different picture, though still with a
high ratio of 1-4 year mortality was shown by Edozien in

Osagere, Western Nigeria (14):

10

TABLE 2
CHILD MORTALITY IN OSAGERE

Age' Number of Deaths per 1000 Live Births

Neonatal 67.6
1 month-l year 67.6
1-2 years 27.0
,3-5 years . 40.6
‘Total - ‘ 202.8

Also in Western Nigeria, Collis, in 1962, found the

following for women in Ilesha villages (8:154):

TABLE 3

CHILD MORTALITY IN ILESHA VILLAGES

Average number of children per mother ' 4.2
Total live children 135.0
‘ Total dead children 119.0
Child (up to puberty) death rate per 1000 births ‘ 468.0

In Lagos, in 1968, there were 38 infant deaths re-
ported for every 1000 live births (44:575). This rate is low,
but it may be seen by looking at the 1967 figures that the
mortality of 1-4 year olds almost approaches the infant mor-
tality: there were 1796 infant deaths and 1155 deaths from
1-4 years of age (44:596). .
Child Feeding Practices, Attitudes, and Intakes

Most of the studies of infant and child feeding
practices found were devised as a baseline for further work,
and did not attempt to relate practices to mortality. For
example, Rao interviewed nursing mothers in order to see

what supplementary diets would be acceptable and how best to

11

formulate a program (38). .

Whiteman studied attitudes towards breast and bottle
feeding in New Guinea and found that 58% of the mothers
thought breast feeding was better, 19% preferred the bottle,
16% didn't know which was better and 8% felt both were equalé
1y good. Those who preferred breast feeding mentioned that
it was less expensive and more convenient than bottle feeding.
that it allows the baby to benefit from the food the mother
.eats, and that it was their custom to breast feed (48:161).
Wennen writes of the decline of breast feeding in Nigeria,
pointing out that 12 of 27 mothers at one clinic were carry-
ing feeding bottles. However, he did not ask what was in
'the bottles or how frequently they were used (47).

Earlier, Jelliffe reported the long duration of
breast feeding in Nigeria and the copious supply of milk
(22:63). Since then, evidence has been accumulating to
indicate that in fact, milk secretion may be inadequate in
Nigeria. Edozien reports very low outputs of from 250-600 ml.
per day with a mean of 500 ml. per day.' Bassir found a mean
of 500 mm. per.day as compared to 800 mm. for European
standards. He points out that this shortage of volume would
mean a loss of 9% of the potential protein nitrogen and would
make mothers' milk alone inadequate for a four-month-old baby.
Furthermore, there seems to be a gradual decrease in milk pro-
:duction after six months. This is about the time Nigerian ba-
bies experience a lag in growth as compared with North American
and European babies (37:37-39). Morley reported that in
Imesi, there was insufficient breast milk for 20t of the

12

children studied (32: 190).

Beliefs and taboos related to foods have been studied
in some areas. FOr example, among the Chimbu in New Guinea,
-Whiteman found that food taboos were not as important an
obstacle to good nutrition as was the lack of understanding
of the relationship between_diet and health (49). In Kenya,
Ojiambo uncovered superstitions such as: feeding babies
meat makes them greedy; eggs retard babies' development: and
vegetables, sweet potatoes, maize and beans lead to bowel
trouble. He found that the list of foods forbidden to
babies far exceeded those allowed (36:521). Hauck reported
the beliefs that bananas give babies worms and eggs make I
children steal in Eastern Nigeria, but she said these beliefs
were not universally held (19:328).

Age of introduction of supplementary foods tends to
be late in many less developed areas. Guthrie reports 12-18
months as the median age for giving solid foods in the
Philippines. This was about the time of weaning (18:167,168).
Ojiambo found a wide range of ages for introducing solids in
Kenya, the most popular ages being 6 months and 1 year. In
Buganda it was reported that 63% of the mothers sampled
weaned directly to an adult diet without using any special
food, and that 25% of the mothers weaned babies without
having introduced them to any supplementary foods (33:159).

4 .In Eastern Nigeria, Hauck found that babies were
generally introduced to hot corn pap by 4—6 months. This

was followed by thicker pap which was followed by yam and

‘13

sometimes palm oil by 12-18 months (19:328). Breast milk
would have been the main source of protein.

A more quantitative study of child feeding practices
was done by Jelliffe in Ibadan, Western Nigeria (22:64). He
recorded the percentage of children in various age groups
who were being given semi-solid food in addition to breast

milk. The results for 208 children were:

TABLE 4

PERCENTAGE OF IBADAN CHILDREN GIVEN
SUPPLEMENTARY FOODS

 

Age Percent Given Food
4-6 months 14
7-9 months 45
10-15 months 80
16-24 months 100_

The foods most commonly given are shown below. In a few
cases, beans were given. Babies only rarely got fruits or

eggs, and none were given meat or fish.

TABLE 5
SUPPLEMENTARY FOODS GIVEN TO IBADAN CHILDREN

Percent of Children

 

 

F°°d TakingFoods
Maize gruel (ogi) 100
Boiled yam and yam flour porridge 40
Bread 30
Boiled Rice 20
Cassava-meal gruel 12

Morely found that corn pap was introduced at about
five months at Imesi and that the next important food to be

introduced was bean pudding at 10 months (32:184). He

14

compared the age of introduction of solid foods for both
well nourished and malnourished children and found them to
be similar. He concluded that dietary factors are not the
only important causes of malnutrition (32:191). .

Qgi or hot pap was the most important supplementary
food reported by Hauck, Jelliffe, and Morely. It is pre-
pared as follows: maize seeds are soaked one to two days
in tepid water, wet-milled, and sieved to remOve fiber and
hulls. The filtrate is allowed to settle andbecome sour.’
The sediment is then diluted to about 10% solids in water
and then boiled into a pap or gel (2). During this process,
about 40% of the protein, 50% of the calcium.and 75% of the.
phosphorus are lost (37:30). Fermenting the ggi does I
increase the protein and B vitamin content, but not enough
so as to make up for initial losses. Although digestibility
is increased slightly, Net Protein utilization, Protein
Efficiency Ratio, and Biological Value are decreased (2).

The problems of child nutrition in Nigeria do not
.end when the child is finally weaned onto an adult diet.
Nicol reports a high incidence of positive signs of mal-l
nutrition in Western Nigeria. Most of his subjects were
children (35:124). Oke pointed out that it is not so much
the nutritive value of the food, as the distribution of food
within the family which is so very poor. Members of the
family whose bodies are already developed get more than they
need at the expense of members who are still growing (37:28,

29). McFie says that all over Nigeria, the nutrient intakes

15

of children, particularly in relation to their calorie
requirements, is poorer than that of adults (25). Earlier,
Nicol said, ”The parents in most of rural Nigeria do not
realize the needs of growing children." (34).

Without clinical investigations and longitudinal
studies of growth, it‘is impossible to evaluate the health
status of children. However, mortality rates do indicate~
a definite failure to keep children healthy although they
do not reflect marginal ill-health. Therefore, it is ex-
pected that mortality rates for one-to-four-year-olds will
be related to child feeding practices and attitudes. Thus,
the following hypotheses were set and considered.

H1: Women who wean their babies before or at the
same time the children learn to eat an adult diet will lose
more of their children as one-to-four-year-olds than women
who wean after the introduction of a complete diet.

H : Women who give children as much animal protein
as they give themselves will lose fewer one-to-four-year-old
children than women who give a relatively small share to the
children.

H :_ One-to—four-year-old mortality rates for mothers
who give 3nly starchy foods as the main supplementation will
be higher than for mothers who give infants a variety of foods.

H : One-to-four-year-old mortality rates are inverse-
ly related to the number of foods a mother feels are appro-
priate for babies. '

H : One-to-four-year-old mortality_rates are inverse-
ly relatea to the ages at which children are first introduced
to stew.

(Age of the Mother'

Coale writes, "The past two decades constitute a

period of extremely rapid declines in mortality in many of

16

the less developed countries of the world, as low cost but
effective techniques of medicine and public health have been
exported to these areas...” (7:185). Where changes in health
care have occurred, one would expect young women to lose_
fewer of their children than older women.

Couvee found in New Guinea that menopausal women had
lost a slightly higher percent of their children than had
fertile women, and that they lost 10% more children after
they had walked than had the younger women (11:331). Ardener
reports that the percentage of a woman's children born that
died under two years of age increased with the age of the
woman from 10% for 15-19 year old mothers to 29% for women
50 or more years old. In a Western Nigerian village,
Monekosso reports that the percentage of a woman's children
surviving decreased with the age of the woman (30:50).

However, where health care has not changed or where
improvements have not reached people, one would not expect
to find changes in mortality rates over time. Wilkinson
reports that even fairly large hospitals do not radically
alter disease patterns of young children in rural areas of
Sierra Leone, because the hospitals are not easily accessible
(50:172).

Changes towards modernization have been found to
bring changes in diet (5). Where such changes are going On
in the Andes, it was found that older families tend to main-
tain simple, traditional diets whereas younger families are

adopting more complex diets. However, the younger families

17

also have more education, and education was thought to be
more important than age in terms of dietary changes of the
whole family (1:94). In Mexico, age of the mother was found
to be positively correlated with age of weaning, and level

of education was negatively correlated. Of the two variables,
Vage was more related to weaning period than was education

- (40:446).

Because Child care and nutrition are not taught in
most Nigerian schools, it was expected that there would be
no difference in child mortality rates for literate and
illiterate mothers, and so the following hypothesis was set:

Hi: There is no significant difference in the pro-
portion 02 children lost as one-to-four-year olds by literate
and illiterate mothers.

Although there has been a great deal of progress in
Nigeria in terms of education, economics, and health faci-
lities, there is reason to believe that little of this pro-
gress has had any impact on rural women. A comparison of
responses for old and young women would indicate whether or
not this is true, and whether important changes have occurred
over time.

H : There is no relationship between the ages of'
the respozdents and the proportion of children lost as one-
to—four-year olds. -

H : There is no relationship between the age of
the respogdent and whether she weans her child before, after,'
or at the same time the child learns to eat an adult diet.

H : There is no relationship between the age of the
respondeng and the proportion of animal protein given to her

children.

H 0: There is no relationship between the age of the
respondent and the type of supplementary food given to infants.

18

- H1 : ‘There is no relationship between the age of
the responéent and the number of foods thought apprOpriate
for babies. '

H13: There is no relationship between the age of
the respon ent and the time at which she introduces her
infant to stew. .

Yoruba Culture

 

Leighton and others have described some of the salient
perceptions, beliefs, and feelings of the Yoruba in the Abe-
okuta area. Many of these apply to the Iwo and Ife villagers,
and several are pertinent to a consideration of child care
and nutrition programs.

"Success is counted in children, wives, houses, money,
land, titles, and good standing with one's neighbors. All of
us--adherents to traditional beliefs, Muslims, and Christ--
ians—-be1ieve in an after-life in which people are judged as
good or bad. The here and now is, however, important in its
own right.” (23:37) (Of all the things which are valued,
children are most important.

"Our world as women Centers on children.“ (23:47)
Children provide women with fulfillment, insure their status
and their husbands' houses, and give security for old age.
"Since our world as women is centered on children, it fo-
_cuses on the future, but we can help them by continuing to
practice many of our old ways." (23:49)

There are other sentiments which should be considered
in terms of planning a change program. "Health is precious

to all of us.” (23:51) "There are many changes coming which

19

promise a better life for us. But we had better not become
so enchanted with the new as to depart radically from the
old customs and beliefs.” (23:37) This is particularly
true as regards health practices. Modern medicine is not
so available as to allow follow-through and complete treat-
. ment, and so people must return to old remedies.

Iwo and Ife people belong to two distinct sub-
cultures. If there are no significant differences between
responses from the two sets of villages, then the data on
practices and beliefs may be generalizable to other Yoruba
areas.

H13: There are no significant differences between
responses of Ife and Iwo women on the following variables:
a) number of foods thought appropriate for babies: b) the
age at which stew is introduced to children: c) the feeding
of starch alone or a variety of supplementary foods: d) wean-

ing before, after, or at the same time the Child learns to
eat an adult-diet.

METHODOLOGY

The Questionnaire

 

The author developed a questionnaire which three
people independently translated into Yoruba. The three
translations were then combined by a fourth party into one '
version which was translated back into English to check the
closeness of meaning. The combined Yoruba version was then
pretested on ten illiterate volunteers in the cities of Iwo
and Ife. City rather than village women were interviewed
for the pretest so as not to contaminate the sample. It was
known that they do not differ from village women in terms
of language and culture. Several alterations had to be made
after the pretest. For example, it was found that when asked,
”How old were your children when you weaned them?', mothers
gave very vague, general responses. This was corrected by
asking specifically about the last child weaned. The order-
ing of the questions was improved during the pretest so as
to maintain the respondents' interest and cooperation.

The final questionnaire is shown in the appendix.
When age was not known, respondents were asked to tell who
were the kings of their home towns at various stages of their
lives. This information was later checked with local histo-

rians to determine ages.

20

21

Questions about the history of the respondents'
children might be prone to all the shortcomings of a retro-
spective survey. Accuracy may be influenced by forgetfulness
as well as reluctance to disclose personal information (50:
168). However, mortality rates based on retrospective
reports have been found to be consistent with rates calcu-
lated from the age composition of a population (7:179).
Morley asked mothers about the outcomes of their last preg-
nancies. He questioned them more than once over intervals
of six menths and then checked the information with other
relatives. He found consistency among the responses (31:82).
For these reasons, it was thought that the method of gaining
demographic data on the children would be reliable. The 4
percentage of total live births lost as one-to-four-year
olds was used as an indication of malnutrition. .

Since it was known from quantitative dietary studies
(9, 25, 35) that large amounts of starchy foods were consumed
by all age groups, Section E on opinions and foods concentrated
on animal proteins, vegetable proteins, fruits and vegetables.

Information on dietary intakes was collected by the
twenty-fouréhour recall method. This method has been used
_by other workers where quantitative data are not necessary
(28:48, 40:443). A village study comparing intake measured
by observation and that measured by recall showed no signifi-
cant difference in results (38).

Most of the questions were open-ended. Though this

made analysis more difficult, it was hoped that it would

22

result in more accurate answers and suggest ways of working
with existing attitudes and beliefs in a nutrition‘program.q
Many of the open-ended questions were funnels, or sequences

leading from the general to the specific.

Sampling

The University of Ife had randomly selected a group
of villages near Ife for a socio-economic survey and com-
munity development project. It was thought that if part of
the child care survey were done in these same villages, the
two sets of data would complement each other. Both sets of
data were gathered independently but in the same villages.

A comparable set of villages was wanted in another
sub-culture. The Iwo area was chosen because no similar
surveys had been done there and because the cooperation of
the villagers was assured. Of those which were close enough
for the author to reach every day, three groups would have
been suitable for a community development project. The
group of villages most comparable to the Ife site held its
main market at Akinbami, which is located on a major road.
The other Iwo villages studied were Olowu, Magboyide, Obojona,
Akuru, Akayo, and Alabata. The Ife villages were Iyanforogi,
Erefe, Ladin, and Aroko. Ife villages were larger, and so,
fewer of them were needed to provide an approximately equal
sample from each area.

In each village, all households were included in the

sample. In each houSehold, one woman was chosen. The only

23

criterion for being chosen was having been married at some
time. Thus, in addition to mothers of young children, the
sample included new brides who had not given birth, older
married women who had no children, and widows. The reason‘
for including these women as well as women in the latter
group was that they were likely to be responsible for the
care of some relative's children.

With one exception (the result of language barrier),
the first married woman encountered by the interviewer in
each household was the one chosen. Interviewers went to the
villages at all times of day, so that they would not always
meet women of the same ages and occupations. For example,
at nine o'clock in the morning, when petty traders and farm
women are at work, most of the women at home would be old or
sick or engaged in occupations other than trading and farming.
By the same token, since many occupations are seasonal and
since there are seasonal variations in dietary intakes,

interviews were spread over the twelve months of 1969.

Ninety women in four Ife villages and eighty-four
women in seven Iwo villages were interviewed. All of the
I villages are located in the rain forest belt of the Western
State of Nigeria.

The Yoruba are city residents. People who have
farms too far away from the city for daily commuting built
villages near the farms. Those who spend most of their time
in farm villages consider their farms to be extensions of

.their city compounds. Usually all the houses in a village

24

belong to people from the same city. Thus, the traditions
peculiar to a given Yoruba city are also found in its I
villages.

Modern amenities such as pipe borne water, electri-
city, and hospital services are lacking in the villages
studied, though they are available in the two cities, and so
village people are aware of them. There is a lot of commuting
.from the villages to the cities. People go "home” to sell
farm produce, to visit relatives, the attend the hospital,
and to observe religious and other celebrations. During
important Muslim holidays, some Iwo villages may be completely
deserted.

. The village of Olowu appears to be typical of the
villages studied. Approaching Olowu from the road, one is
first struck by the brownness of the village in contrast to
Ithe lush green of the surrounding cecoa plantations and
forest. The houses, made of mud, are the same color as the
ground which is kept clear of vegetation.- The roofs are
mostly corrugated metal, but rust makes them brown,too. (Only
occasionally does one see a plastered house.

Towards evening, the air is full of the scent of wood'
cooking fires and palm oil. Here and there one hears the dull
thud of pestle on mortar, pounding yam. Men come back from
their farms, carrying their cutlasses and short-handled hoes.
Women and children come home carrying baskets of harvested
food on their heads. Women traders come back from market

with their wares on their heads and babies on their backs.

25

Throughout the village, those who have stayed home greet those
who are returning, "Welcome, did you sell? I hope everything
is alright.” Small groups gather to chat at the tailor's and

carpenter's houses while supper is being prepared.

Data Collection

 

Three people besides the author worked as interviewers.
They included one secondary school student and two university
students. They received training, practice, and supervision
before they began collecting data on their own. Interviews
were conducted in Yoruba and the secondary school student
recorded in Yoruba.

Women were interviewed in their homes. Reddy found
that one can get more information from mothers in the relaxed
atmosphere of their own homes (39). The main problem with
this approach was one of securing privacy, since the presence
of interviewers in the village always attracted a large crowd.
Due to the personal nature of the questions, the interviewers
had to repeatedly ask onlookers to leave, promising that once
the interviews were completed, they would read the questions

to anyone interested.

Statistical Ana1ysis .

As a result of the kind of sampling we did within the
household, the statistical analysis has been restricted to
non-parametric statistics. The hypotheses were tested using
the chi-square test (27:212-242). This test procedure does

not require any distribution-—that is, it is distribution-free.

26

Consequently, there is no need for the sampling procedure to
conform absolutely to the books. It is claimed, however, that
the sampling procedure is a nested random sample: Within a
village, a complete enumeration of households was undertaken,
and within a household a woman was chosen in such a way as

to introduce chance elements by varying the time of visit.
This claim was, however, not necessary for our analysis which
used chi-square--a non-parametric statistic--which can be

used to study relationships and differences (27:357).

.In order to use the chi-square test, the data were
tabulated in a frequency distribution. This frequency dis-
tribution then formed the observed data. The marginal dis-
tributions (row and column totals) were then used to calcu-
late the expected frequency. For example, if the following

distribution were observed,

1 _2 Total

1 n11 n12 “1-

2 n21 n22 n2-

3 n31 n32 n3-
Total n.1 :1.3 n..

the expected distribution would be computed by using the

formula,

where i refers to rows and j refers to columns, so that for

n one has n1: n:1 and for n one has ”3° n-2 .
11 ‘13—— 32 T—

27

After the expected distribution had been computed,

2
2 = {£9%§1—-, was used to compute

the chi-square formula, x
the various chi-square values shown in column 2 of Table 31.
This formula merely says that one subtracts each expected
value from the corresponding observed value, squares the
difference and then divides the square by the expected value,’
repeats for each expected value and adds all the results.
This computed chi-square is then compared with the tabulated
ones (43:28-29) using the appropriate levels of significance
and degrees of freedom which are computed as the product of
number of rows minus one and number of columns minus one.
In Table 31, three different levels of significance were
used for testing each of the hypotheses. I

The rationale behind using the chi-square is as
follows: If there is no relationship between the row vari-
able and the column variable, the expected frequency distri-
bution generated by using the marginals will not be signi-
ficantly different from the observed frequency distribution,
and one will therefore expect the computed chi+square to be
close to zero. If the computed chi-square is less than the
tabulated one, it is concluded that there is no significant
relationship between the row variable and the column vari-
able. But if the computed chi-square is larger than the
tabulated one, then there is a significant relationship.

Since the chi-square test does not tell us the

direction of relationship, the Pearson product moment cor-

relation coefficient (27:115-l43) was computed in cases where

28

the direction of relationship was wanted.
Data for purely descriptive purposes were hand-
tabulated and the numbers and sometimes the percentages

were presented.

RESULTS'.

Description of Respondents

Most Of the women interviewed were of child-bearing
age. The Iwo sample contained a higher proportion of very
young and very old women than the Ife sample, as shown in

the table below.

TABLE 6
AGES OE RESPONDENTS

 

 

 

 

Age in Years Iwo Ife ‘ Both
No. % No. % No. %’

16-25 ' 18 21 6 7 24 13.79 .
26-35 29 35 40 44 69 39.65
36-45 f 17 20 35 39 52 29.90
46-55 10 12 6 7 16 (9.20
56-65 ‘ 5‘ 6 2 2 7~ 4.02
66+~ 3 4 0 ' 0 3 1.72
Not known __3. __2_ __1 ‘__l ' '_;3 1.72;

90 100 174 100.00

Totals 84 100

Most of the wives worked.' The most common occupation
-was petty trading which included fecd vending, selling native
soap,_and selling odds and ends such as matches. A large '
egroup of women were farmers.‘ Some‘were engaged in unskilled
labor, jobs needing no apprenticeship, such as collecting
palm fruits. Some of the women were big traders who traveled
a lot in their work. This group included produce buyers._ A

few women combined farming with other work and a few were

29.

30

engaged in skilled labor such as palm oil processing. Only
four respondents were housewives alone.

The largest majority of the respondents' husbands
were farmers. They grew both cash crops such as cocoa and
kola nuts and food crops. Several of them combined farming
with large trading operations. A few were salary earners.
This group included soldiers, teachers, and clerks. Other

occupations may be seen in Table 7.

 

 

TABLE 7
OCCUPATIONS
Occupations Number in Each Category
- Wife Husband
Petty trader 79 --
Farmer . 39 117
Unskilled laborer 18 1
Big trader 13 --
Farmer and big trader .-- 10
Farmer and laborer or petty trader ll 10
Skilled laborer 10 ‘ 8
Regular employee of government or
business -- 6
None . 4 2
Totals , 174 167

Only thirty-two women interviewed had ever gone to
school. Twenty-two of these were literate. Table 8 shows‘
that several women did not attend school for very long. The
highest level of school completed was Class VI, the equivalent

of twelfth grade.

31

TABLE 8

EDUCATION OF RESPONDENTS

 

 

Years in School Number of Women
1-3 11
4-6 11
7-9 7
10+ 3

As shown in Table 9, very few women had access to
radio programs in the villages. However, a few mentioned
that they had rediffusion at home in the cities. A larger
proportion of women said they had watched a film at least
once. These films were either Indian dramas shown in the
cities or educational programs brought to the villages by
the Ministry of Information. Only 16 women had ever been
contacted by change agents on the subject of child care and
feeding. These agents were mostly nurses, and not home

economics extension agents.

TABLE 9
EXPOSURE TO OUTSIDE SOURCES OF INFORMATION

Number _Percent

Had radio in the village 12 7
Had ever seen a cinema 78 45
Had been contacted by agent 16 9

Feeding of Children

Many women felt that they had never been taught to
care for their children. In the Ife villages, thirty-nine
perCent of the women said either that God or no one had

taught them child care practices. Seventeen percent of the

32

Iwo respondents gave that answer. Iwo women seemed to have
a clearer perception of child care as something to be learned
and taught. Seventy-three percent said they were taught by
their mothers, mothers-in-law, and other elders, as compared'
with forty-seven percentin the Ife villages- Only seven
percent in the Ife villages and eight percent in the Iwo
villages said they received training from medical personnel.
Almost all of the respondents said they would like to learn
new ways of feeding and caring for children.

At the time of the interviews, many of the women were
nursing babies: fifty-one percent in Ife and 36% in Iwo.
The difference between the two sites reflects the difference
in age distribution.

Table 10 shows the respondents' preferences for
bottle or breast feeding. Considering that most of the women
who offered no opinion did not know what a feeding bottle was,

there is a marked preference for breast feeding.

TABLE 10

PREFERRED FEEDING METHODS

Response Iwo Ife
Number Percent Number Percent
Preferred breast feeding 36 43 34 38
Preferred bottle feeding 6 7 3 3
Both equally good 6 7 l9 , 21
No answer or can't say 36 43 34 38

The main reason given for preferring breast feeding
was that it is the natural method, what God provided for babies.

Other advantages cited were lower cost, better growth rate,

33

convenience, and tradition. Some women were aware that breast
feeding allows the baby to benefit from the food the mother
eats, and that it strengthens affective ties between mother
and child.

Those who preferred bottle feeding said that it pro-
vides "real food" prepared by machine; it allows for easier
regulation of the number of feedings: it is more hygienic:
and it does not limit the child to the mother's supply of
milk. Those who felt that both methods were equally good
said that occasional bottle feeding gives the mother some
freedom; it can supplement insufficient breast milk; and it
allows them to "follow the advice of educated people" and
at the same time to use the natural method.

Thirty percent of the respondents in the Iwo villages
and fifty percent in the Ife villages had at some time used
a feeding bottle. Thus, it is seen that in both areas, less
than half who had used one thought it was as good as or-
better than breast feeding.

Table 11 shows the types of liquids fed by bottle.

TABLE 11

TYPES OF LIQUIDS FED BY BOTTLE

 

 

Liquid Given by Bottle Number of WOmen Mentioning_£t
Lactogen and other milk products 49
Hot corn pap 33
Water 16
Glucose 9
Commercial baby cereal 8
Native medicine ;

Egg

34

-The most common reason for use of the bottle was to
avoid the dangers of forced feeding whereby hot pap and ~
native medicine are usually poured down the baby's throat.
The next most important reason was instructions from a nurse
or doctor at the hospital or maternity clinic. Other reasons

may be seen in Table 12.

TABLE 12
REASONS FOR USING A FEEDING BOTTLE

Reason Number of Respondents

Easier feeding 31
Maternity instructions 14
It is good, or they say it is good 13
Saw others do it 8
Freedom, travel 5
Hygiene 3
. To give something hot 2
Variety from breast 1
Pregnancy 1

Most of the babies were weaned between the ages of
two and four years. Table 13 shows that only one woman,
the one who became pregnant very early, weaned her baby
before he was a year old. Most Ife women weaned their
babies after the second birthday, but Iwo women tended to _

wean after the third birthday.

35

TABLE 13

AGE OF WEANING

 

Iwo Ife
Age of Weaning Number Percent Number Percent
Less than 1 year 1 l - -
l year-l year, 11 months 5 6 12 13
2 years-2 years, 11 months 23 27 49 54
3 years-3 years, 11 months 37 44 l7 l9
4 years + 3 4 2 2
Not applicable 15 18 10 12.
Totals - 84 100 90 100

The most common method of weaning was to make snacks
available to the child and to allow him to give up the breast
of his own free will. Fifty-two percent of all respondents
who had weaned a child said they used this method. Rubbing
bitter leaf on the breasts was the next most popular method,
used by twenty-seven percent. Refusal was used by nineteen
percent. Only two percent of those who had weaned a baby did
so by replacing breast milk with Lactogen or some other milk
product. Two of the respondents in the Ife villages and ten
in the Iwo villages sent their children away to their grand-
mothers during the time of weaning. .

The most frequently mentioned weaning food was bread.
Others were eggs, ground nuts, sweets, rice, beans, and corn
pap. Milk was mentioned in Ife but not in Iwo.

Table 14 shows the ages at which children of Iwo and
Ife mothers were able to eat a complete adult diet. It may
be seen that the Ife children were given full diets much

earlier than those in the Iwo villages.

36

TABLE 14

AGE OF INTRODUCTION TO ADULTS' FOODS

Iwo ' Ife
Age Number Percent Number Percent
Less than 1 year 0 0 ' l l
l year-l year, 11 months 6 7 44 49
2 years-2 years, 11 months 15 18 24 27
3 years-3 years, 11 months 18 21 6 7
4 years + 16 19 1 1
Not applicable 29 35 14 15
Totals 84 100 90 100

On the whole, Iwo mothers tended to wean before the
child was given a complete diet, but Ife mothers weaned after.
This can_be seen from Tables 13 and 14 above. In Ife, 50%
of the 1 year to 1 year, 11 months group were given adult
food, but only 13% of them were weaned compared with 6% and
7% respectively in Iwo.

Most mothers began supplementary feeding by the time
their babies were six months old. Table 15 shows that the
ages at which supplementation was first introduced were
about the same for both groups. Those who said they began
feeding before one month actually gave native medicine, not
food.

Foods which were mentioned as having been given
before the child could eat a full adult diet are shown in
Table 16. Corn pap was the most frequently mentioned, whereas
meat was not an important item. Ife women had a greater
tendency to mention commercial products such as milk and

baby cereals.

37

TABLE 15

AGE OF FIRST SUPPLEMENTARY FEEDING

 

 

 

Age Number of Mothers Percent for Both
in months Iwo Ife Both Ife and Iwo
Less than 1 28 25 i 53 30
1-3 7 10 17 10
4-6 23 10 33 19
7-9 8 9 .17 10
10-12 1 l 2 1
13+ 10 13 23 13
Not applicable 13 16 29 17

, Total ‘, 90 84 174 100
TABLE 16

SUPPLEMENTARY FOODS MENTIONED

Number of WOmen
Foods Iwo Ife

Corn pap 53 69
Lactogen and other milk products 12 30
All soft foods 6 10
Yams 13
Eggs

Beans

Bread

Commercial Baby Cereals
Glucose ‘
Fruit

Rice

Pepper sauce

Meat

Cottage Cheese
Plantain

«7:069:01»:anme
I I qu I umuwul

Since the question on supplementary foods was vague
and answers depended on what the mothers thought was important,
a specific question on the age at which the child was intro-
duced to stew was included. The median interval for Iwo

mothers was 10-12 months, for Ife mothers it was 4-6 months.

38

Thus, the Iwo mothers began to introduce their babies to what
may be considered the staple diet about six months later than

did the Ife mothers.

TABLE 17

AGE OF INTRODUCTION TO STEW

 

Age of baby Number of Mothers

in months Iwo Ife

Less than 1 0 5
1-3 5 19
4-6 17 22
7-9 11 15
10-12 14 4
13+ 26 15

Dietary Intakes

 

The evening meal was the heaviest, the one requiring
the most preparation in about two-thirds of the Ife house-
holds and three-fourths of the Iwo households.

Some form of animal or pulse protein was served at
almost every meal. At breakfast, about ninety percent (Ife
90%, Iwo 91%) of the families consumed protein, usually in
the form of beans. In most households, the afternoon and
evening meals were usually based on animal protein cooked in
a stew and supplemented with melon seeds and locust beans.
At lunch, fifty-six percent of the Ife households and seventy-
seven percent of the Iwo households had some meat, fish, or
poultry. In the evening, sixty-six percent of the Ife
families and seventy-six percent of the Iwo families served
animal protein. A much larger number of families in Ife

did without animal protein on the day before the interview.

39

Of those respondents who had children living with
them, seventy-one percent at Ife and 72% at Iwo said they
gave the children less meat than the adults. The amount of
meat, fish, or poultry eaten by adults was very small, usual-
1y just one piece. 1 3

Almost all of the respondents served green vegetables
during the day, and most of them served vegetables twice or
more. Most of the vegetables eaten came from the consumers'
farms.

Only ten percent of all respondents said they had
eaten fruits. Seven percent mentioned the use of milk
products in the home, usually Laotogen for babies or milk for
their husbands' tea. Eggs were mentioned by four women in
the Iwo area, but no one in the Ife sample had eaten eggs.

Vendors sold breakfast foods to sixty percent of the
Ife respondents and to forty-two percent of those in the Iwo 3
area. In the afternoon, only twenty-four percent at Ife
and five percent at Iwo villages bought food from vendors.

In the evening, eleven percent of the Ife women and two
percent in Iwo villages bought food from vendors. In each
area, only two percent of the women bought snacks from

vendors.

Attitudes Towards Foods for Babies

Many high protein foods were considered inappropriate
for babies on the grounds that the children have no teeth,
they do not know how to eat the foods and there is no need

for them to eat the foods. Thus, all types of meats,

4O

ground nuts, crayfish, groundnut balls, and bean cakes were
not thought suitable for babies. A few people felt that
these same foods plus okra and eggs cause piles and loose
stools. Twenty-five respondents felt that meat, poultry,
fish, and snails cause worms in babies but not in older
people. Three people felt that feeding fish, meat, and eggs
to babies causes them to be greedy, or even to steal. How-
ever, it was explained that this was because those foods
.are luxuries and a baby who is fed them may develop expensive
tastes which cannot easily be satisfied.

Tables 18, 19, and 20 show some of the reasons

certain foods were thought unsuitable for babies.

TABLE 18

WOMEN WHO FELT FOODS WERE INAPPROPRIATE
BECAUSE BABIES HAVE NO TEETH

 

Foods Number Percent
Meat, Fish, Poultry 49 28
Snails 60 34
Crayfish 46 26
Groundnuts 45 26
Melon Seed Cakes 54 31
Beans 12 7
Bean Cakes 21 12
Papaya 31 18

TABLE 19

WOMEN WHO FELT SOME FOODS GIVE BABIES WORMS‘

Foods ' Number Percent
Meat, Fish, Poultry 24 14
Snails 3 2
Okra , l 1

Eggs 1 1

41

TABLE 20

WOMEN WHO FELT SOME FOODS CAUSE
PILES AND LOOSE STOOLS

Foodse .Number Percent
Peanuts l4 8
Meat, Fish, Poultry 5 3
Okra 2 1

Vegetables, locust seed, melon seed in stew, beans
and bean pudding were not widely objected to, providing the
baby is able to take the pepper in them.

Most fruits were thought to be good for babies, though
it was often expressed that guavas are only goOd for lite-
rates. Several respondents did not know what local cottage

cheese (wara fulani) was.

 

With the exception of the idea that animal protein
foods cause worms, it may be said that there were almost no
superstitious beliefs related to foods for babies. On the
other hand, there was a widespread attitude that if God
meant babies to eat certain foods, He would have provided

the necessary equipment (teeth).

Mortality of Children

 

Table 21 gives mortality figures as reported by res-
pondents for the Iwo and Ife areas separately and for the
total sample. Ten women in the sample had never given
birth. For the 164 women who had, the average number of
live births was 3.2. On the average, each woman lost one

child (.96 deaths per woman). About a third of these deaths

42

occurred in the first year of life, but over half occurred

from the ages of one to four years.

Total child mortality was much higher in the Iwo

villages than in the Ife villages, but the ratio of one-to-

four-year-old mortality to infant mortality was only slightly

higher in the Iwo area.

TABLE 21

MORTALITY FIGURES

 

 

 

 

 

 

Iwo Ife Both
Total Live Births 264 275 539
Number of Deaths by Age Group ‘
0-1 year 30 23 53
l-4 years 53 36 89
5+ years 8 8 16
Total Child Deaths 91 67 158
Mortality Rates/1000 Live Births
by Age Group
0-1 year 114 84 98
1-4 years 201 131 165
All ages 382 244 293
Age Groups as Percent of
Total Mortality
0-1 year 33 34 34
1-4 years 58 54 56
Ratio of 1-4 year old mortality
to 0-1 year old mortality 1.76 1.56 1.68

 

43

Testing of Hypotheses

 

Relationships Between One-to-Four-Year-Old Mortality and
Child Feeding. 1

H1: Women who wean their babies before or at the
same time the children learn to eat an adult diet will lose
more of their children as one-to-four-year-olds than women
who wean after the introduction of a complete diet.

Comparisons were made among toddler mortality rates
of women who weaned their children before, after, and at the
same time their children could eat full adult diets. Table

22 shows the proportions of deaths in the one-to-four-year-

old group for mothers in each category

TABLE 22

ONE-TO-FOUR-YEAR-OLD MORTALITY IN RELATION
TO TIME OF WEANING

 

 

 

Time of Percentage
Weaning Relative Number of , of
to Time of Total Live Deaths Ages Total Live Births
Introduction of Births 1-4 years Lost as
Adult Diet 5 1-4 year olds
Before 123 18 14.63
Same 129 31 24.04

After (227 46 20.26

Women who weaned before the child could eat a full diet lost
fewer of their toddlers than those who weaned after, but those
who introduced a full diet at the time of weaning lost a
higher proportion than those in either of the other catego-
ries. The differences among the three categories were small.
The computed x2 of 3.5472 was not significant at either the .

5% or 10% level. Thus, hypothesis 1 is rejected.

44

HI: Women who give children as much animal protein

as they 9 ve themselves will lose fewer one-to-four-year-old
children than women who give a relatively small share to the
children. -

This hypothesis was not supported. In fact, looking
at Table 23, one can see that in households where animal
protein was served on the day prior to the interview, those
who gave children the same amount as they ate had actually

lost more of their toddlers.

TABLE 23

ONE-TO-FOUR-YEAR-OLD MORTALITY IN RELATION
TO DISTRIBUTION OF ANIMAL PROTEIN

 

 

 

 

. prgigififsgfltfigl Number of Percentage of
to Children in Total Live Piazza TOtaIL:;:eaglrth3
Comparison with Births _ _
Mother's Portion 1 4 Years 1 4 year 0168
Less 262 32 12.21‘
Same 90 16 17.78
None served in
Household 94 19 20.21

Although this finding is the opposite of what was
expected, the relationship between proportion of meat served
and proportion of toddler deaths was not significant at the

5% or 10% levels.

H : One-to-four-year-old mortality rates for mothers
who give only starchy foods as the main supplementation will
be higher than for mothers who give infants a variety of

foods.
Table 24 shows that a higher proportion of toddlers

was lost by mothers who fed only starchy foods for supple-

mentation.

45

TABLE 24

ONE-TO-FOUR-YEAR-OLD MORTALITY IN RELATION
TO TYPE OF SUPPLEMENTARY FOOD

 

 

 

 

Number of Percentage of
Type of Total Live Deaths Total Live Births
Supplementation Births Ages Lost as
1-4 Years 1-4 year olds
Starchy 330 65 , 19.70
Variety . 188 _ 22 11.70

2 was 89.5491 which was significant at

The computed x
the 5% level. Thus, hypothesis three was supported.

H : One-to-four-year-old mortality rates are inverse-
ly relates to the number of foods a mother feels are appro-
priate for babies.

No relationship was seen between the two variables.
While the Pearson product moment correlation showed a slight
tendency for proportion of deaths to increase with number of

2 of

foods thought appropriate (r=.l41639), the computed x
21.7349 was not significant even at the 25% level.

H : One-to—four-year-old mortality rates are posi-
tively reIated to the ages at which children are first.in-
troduced to stew.

- This hypothesis was found to be supported. The
computed x2 of 96.3489 was significant at the 5% level. The
Pearson product moment correlation, though small (r=.18114l),
was positive.

In summary, of the predicted relationships between
toddler mortality and child feeding practices, only those

concerned with the type of supplementary food and the age

of introduction to stew were supported.

46

Relationship between Education of Mother and Toddler
Mortality. O I

H : There is no significant difference in the pro-
portion 0? children lost as one-to—four-year-olds by literate
and illiterate mothers. ’

As predicted, the relationship between literacy
and toddler mortality was not significant even at the 25%
level.. ’
Relationships between Age of Respondent and Toddler
Mortality and Child Feeding. I

H : There is no relationship between the ages of-

the respondents and the proportion of children lost as one-
to-four-year olds.

2 of 28.2678 was not Significant even

The computed x
at the 25% level. However, Table 25 shows that the very
young and the very old lost more of their toddlers than did

those in the ages 26-55.

TABLE 25
AGE OF RESPONDENT AND TODDLER MORTALITY

Number of Percentage of
Age of Respondent Total Live Deaths of Total Live Births

 

 

 

in Years Births 1-4 year Lost as
olds 1-4 year olds
16-25 ' 46 11 23.91
26-35 179 26 . 14.53
36-45 ‘ 182 27 14.84
46-55 68 7 10.29
56-65 37 8 21.62
66+ _ 17 5 29.41

In contrast to toddler mortality, infant mortality
tended to increase steadily with the age of the respondents,

as shown in Table 26.

47

TABLE 26
’AGE or RESPONDENT AND INFANT MORTALITY

Number of Percentage of
Age of Respondent Total Live Deaths of Total Live Births

 

 

 

in Years Births 0-1 year Lost as

‘ olds 0-1 year olds
16-25 - 46 4 8.70
26-35 179 15 6.15
36-45 182 17 9.34
46-55 68 9 13.24
56-65 ‘ 37 5 13.51
66+ 17 - -

There is no relationship between the age of the
respondent and whether she weans her child before, after, or
at the same time the child learns to eat an adult diet.

This hypothesis was supported. The computed x2 was
4.3660 which was not significant even at the 25% level.

H : There is no relationship between the age of
the respondent and the proportion of animal protein given
to her children.

The computed x2 of 6.2561 was not significant at any
of the three levels tested for. So, hypotheSis nine is
supported.

. There is no relationship between the age of
the respongent and the type of supplementary food given to
infants.‘

This hypothesis was supported. The computed x2 of

2.9090 was not significant, even at the 25% level.

. There is no relationship between the age of
the responéent and the number of foods thought appropriate
for.babies.

2

The computed x of 17.3866 was not significant at

any level tested, so this hypothesis was supported.
There is no relationship between the age of

the responéent and the time at which she introduced her
infant to stew. -

48

2 of 38.6592

This hypothesis is accepted because the x
was not significant at the 5% or 10% levels. However, it was
significant at the 25% level. The Pearson product moment
correlation showed a light tendency for older women to intro-
duce stew later than younger women (r=.0234).

In summary, all of the hypotheses which stated that
age is not related to toddler mortality and child feeding
were upheld.

Differences in Child Feeding between Iwo and Ife Villages.

H13: There are no significant differences between

responses of Ife and Iwo women on the following variables:
a) number of foods thought appropriate for babies; b) the
age at which stew is introduced to children; c) the feeding
of starch or a variety of supplementary foods: d) weaning
before, after, or at the same time the child learns to eat
an adult diet.

Table 27 shows the number of foods thought appro-

priate for babies by Iwo and Ife village mothers.

TABLE 27

NUMBER OF FOODS THOUGHT APPROPRIATE FOR
BABIES IN TWO COMMUNITIES

Iwo Ife

Number Of Foods Number Percent Number Percent

 

 

0-5 8 10 18 20
6-10 19 23 19 21
11-15 22 26 25 28
16-20 9 11 .17 , 19

More Ife than Iwo women felt that very few foods are
appropriate for babies. Conversely, more Iwo than Ife women
felt that many foods are suitable for babies. The differences

in response for the two communities are significant at the SS

 

49

2 = 11.1929).

level (x
Table 28 shows the ages at which mothers in the two

sets of villages introduced their babies to stew.

TABLE 28
AGE OF INTRODUCTION TO STEW IN TWO COMMUNITIES

Ife Iwo

Age Of Chlld 1n Months Number Percent. Number Percent9

 

 

0-6 31 39 22 30
7-12 37 46 25 35
13-18 12 15 17 23
19-24 - - S 7 ‘
25-30 - - 1 1
31-36 - - 2 3
37+ __: __: __1 __1
Totals 80 100 73 100

It is seen that all Ife women introduced stew no
later than a year and a half, but several Iwo women waited
for more than two years. The differences between responses
of Ife and Iwo women were significant at the 5% level
(x2 = 15.0387).

Ife women had a greater tendency than Iwo women to
feed starch alone as the supplementary food. This is shown
in Table 29.

TABLE 29

TYPES OF SUPPLEMENTARY FOOD IN TWO COMMUNITIES

 

 

 

Food Ife Iwo
Number Percent Number Percent
Starch 57 66 39 57
Variety . 29 34 30 43

However, the difference of response on this variable is very

slight. The computed x2 was 1.7739 and was significant only

 

50

Aeat the 25% level.
Table 30 shows that Ife women had a marked tendency

to wean their babies after the introduction of a Complete diet.

TABLE 30
TIME OF WEANING IN TWO COMMUNITIES
Time of Weaning Relative

to Time of Introduction
of Adult Diet

Ife Iwo‘
Number Percent Number Percent

 

 

 

Before 9 12 29 42
Same 13 17 25 36
After 54 71 15 22

The differences between the two groups on this ques-

2 = 36.7625).

tion was significant at the 5% level (x
In summary, it has been shown that there are signi-
ficant differences between the two communities on the follow-
ing variables: number of foods thought appropriate for
babies, the age at which stew is introduced, and time of
weaning in relation to the time of introduction of an adult
diet. The type of supplementary food given was not signi-
ficantly different for the two communities. .
Table 31 on the following page summarizes the testing
of hypotheses. (It may be seen that toddler mortality rates
were significantly higher for mothers who fed infants only
starch foods as supplements and for those who introduced stew
at a later age. However, other variables concerned with child
feeding were not significantly related to toddler mortality

rates. Literacy and toddler mortality rates were not found to

be related. The age of the respondent was not related to

51

 

 

, Husbandomaw are DGOOHmasmHm uoz «a Cameos momozuomhm a
.m .m .n hh.m ao.¢ mm.m N mwoh.om end
.m .m.: .m.c mm.H ah.~ <m.m H mmnn.a «Una
.m .m .m ,vh.h vo.oa am.~a o hmmo.md Ema
.m .m .m mm.m .mh.h me.m e mmma.aa Ema
.m .m.: .m.: om.vm o~.ov hn.me om ~mmm.mm sea
.m.: .m.: .m.: mm.m~ av.m~ H¢.Hm om owmm.ha «Ha
.m.o .m.o .m.o mo.m vm.m ho.HH m omom.~ roa
.m.c .m.s .m.: mm.~a mm.ma Hm.ma ca Homm.o rm
.m.c .m.o .m.s mm.~H mm.ma Hm.ma 0H comm.v «m
.m.s .m.s .m.s em.m~ mm.em me.em mm mee~.e~ «e
.m.: .m.c .m.s mm.w v~.m no.Ha m mmma.m «m
.m .m .m om.vm m~.ov uh.mv om mmem.mm «m
.m.s .e.s .m.s me.m~ He.e~ He.Hm em meme.a~ 4
4m .m .m mm.a anew em.m .H Hmem.mm «m
.m .m.: .m.: hh.N Howe mm.m N. «moo.m N
««¢.m .m.c ««.m.c 5n.~ Ho.v mm.m N mnem.m H
,emN wee mm «mm wee mm someone ax
c0wmsaosoo Nx oouoasoma mo mooummo oousmfioo e mwmonuommm

OZHBmmB mummmfiomwm ho NMdZSDm

Hm Mdmda

52

toddler mortality or to child feeding practices. Some
differences in child feeding practices were found between

the two groups of women.

 

DISCUSSION, IMPLICATIONS, CONCLUSIONS

Discussion and Implications

 

This study demonstrates the very high toddler mortal-
ity rates in the two communities studied and lends support to
the work of other researchers who found toddler mortality to
be related to nutritional factors. The total child mortality
rate found in this study was slightly higher than for Osagere
village (14), but much lower than for the Ilesha villages
(8:154). In Nigeria it may be that mothers forgot some of
the children who died or that they counted a child believed
to be abikg as only one death. In either case, mortality
would be higher than reported here. A future study of
mortality should ask for total number of pregnancies. Even
if the mortality rates are an underestimate, they point out
that the ratio of deaths of one-to—four-year-olds to infant
mortality is about four times as high as what Gordon says
indicates poor nutritional status of a community (17:362).

As discussed above, the relationship between toddler
mortality and time of introduction of an adult diet relative
to time of weaning was not as expected. Those who weaned
babies at the same time they introduced adult foods lost the
highest proportion of their toddlers. It may be that the
sudden loss of milk protein coupled with the increased con-
sumption of red pepper and the increased exposure to diarrhea

53

S4

and infection contributed to the higher mortality. Likewise,
the higher rate of toddler mortality for those who weaned
after the introduction of a complete diet might be related

to the type of food given plus hygienic factors. While the
probability that the relationship tested in hypothesis one was
due to chance is 25%, still, the relationship does bear con-
sideration. It may be that early introduction of the wrong
foods is harmful, and that one should not advise early intro-
duction of food without specifying what foods.

It was thought that distribution of animal protein
would reflect the respondents' attitudes towards the impor-
tance of animal protein for children, and as such wOuld be
related to toddler mortality. With the outcome of hypothesis
two, it is now realized that too many other factors besides
attitude affect meat distribution and that one cannot hope
to relate a single, recent event to the loss of children
which may have occurred twenty years ago. However, when
toddler mortality rates were compared for those who served
animal protein in the household and those who had not served
any, it was seen that those who had not served any had lost
more of their children as one-to—four-year-olds (20% as
compared to 14%). Meat consumption may be a reflection of
general economic factors which may be related to toddler
mortality.

Only a very simple categorization of supplementary
foods was made, the idea being that infants fed a variety of

foods would have higher protein intakes than those limited

 

55

to corn pap and yams. Toddler mortality rates may reflect
differences in protein intakes. The foods asked for were
those given in addition to breast milk before the introduc-
tion of an adult diet. Those getting a variety of foods may
have been given soft foods containing tolerable amounts of
pepper and therefore not precipitating a crisis as early
introduction of adult foods may have done.

The lack of relationship between the number of foods
thought appropriate for babies and toddler mortality rates
was probably due to two things. Almost no one felt that
animal protein foods were appropriate for babies and even
those foods thought appropriate were not often given as sup-
plementation.

The earlier a baby starts to dip into stew, the
sooner he is introduced to the main source of animal protein.
This may explain the relationship between early introduction
to stew and lower toddler mortality rates.

The age of the mother was not an important variable.
It was seen that those in the age groups 16-25 and 56+ lost
more of their toddlers than women ages 26-55. Most of the
women in the high and low age groups were from the Iwo
villages. It is not clear whether the age distribution may
have been responsible for differences in mortality rates
between Iwo and Ife or whether different child care practices
in Iwo were responsible for higher toddler mortality for Iwo

(older and younger) women.

56

Since for the total sample, there was no consistent
relationship between age of respondent and child feeding
practices, one can say that child feeding in the two areas
studied has not changed significantly in the last forty
years. This is probably related to the low levels of mass
media exposure and the fact that women learned to care for
their children from their elders, or as some said, from God.

The fact that infant mortality rates were lower for

'ff“*“r*-

younger women may be a result of increased use of hospitals gJ
and maternity clinics for deliveries. Since the proportion

of infant deaths has been decreasing, but the proportion of

toddler deaths has not decreased, a young woman is more

likely than her mother was to have a baby that lived through

infancy only to die in early childhood. This demonstrates

the urgent need for more programs aimed at young children.

Implications for Nutrition Education Programs

Because of the importance of children in Yoruba
culture and because of the widespread interest in learning
new child feeding practices, one would expect mothers to
be receptive to a well planned program. The child feeding
practices that this study revealed would have importance in
planning a nutrition education program. I

It appears that there is no danger that breast feed-
ing is declining in the two communities studied. However,
due to the poor sanitation of feeding bottles usually found
among poorer classes of less developed countries (21), the

use of feeding bottles should be discouraged, even if feeding

57'

bottles are not used to replace the breast. Since the main
reason for use of a bottle was to avoid forced feeding,
another alternative such as the use of spoons and cups might
be recommended.

The feeding Of native medicine is an important fea-
ture of infant care in both communities. These medicines are
made from quinine leaves and as such probably help prevent

malaria. However, more needs to be known about the other

1..—*7, ~

ingredients used and what physiolOgical effects they have, hi
particularly in regard to the babies' digestive processes.

The possibility of adding protein supplements to these tonics

should also be investigated.

Morley recommended that supplementary feeding should
begin by six months at the latest (31:85). However, forty
percent of our sample began supplementary feeding later than
this, suggesting the need for teaching earlier as well as
more appropriate supplementation.

It has been indicated that early introduction of a
complete diet may be harmful, so there is a need to develop
special foods for babies. The question on attitudes towards
foods for babies points out that the main objections to pro-
tein rich foods were the idea that babies do not need them
and the fact that such foods are hard. Other foods were
thought unsuitable because of pepper. Mothers could be
taught to remove a small portion of the stew they are pre-
paring before adding pepper. They could add small amounts

of ground and prepared meat, fish, peanuts, melon seeds,

58

\

locust seeds or crayfish. This high protein Sauce could be
fed to children along with soft vegetables such as okra and
any carbohydrate rich food the mother thinks suitable. There
are many other dishes, and especially bean dishes which a

mother could prepare easily for infants, just by altering a

 

small part of the food she is preparing for the family.
Mothers might also be encouraged to give their babies more 1.
fruit.

It is hoped that in time, and with the provision of Ea

alternatives, the practice of feeding corn pap may almost
disappear. Before that happens, though, supplementing the
pap with eggs, using only fermented ggi, and feeding a thick-
er, less watery pap would improve the calorie, protein and
vitamin content.

Since bread was the snack most frequently given to
children during the weaning period, it could be made more
nutritious by spreading peanut butter on it. Groundnut or
cocoa flour could be added before baking. A high protein
bread available at the same price as regular bread and
advertised as a special weaning food would probably sell very
well.

The qualitative study of dietary intakes points to
some probable nutritional deficiencies among older children
and adults. Certainly, children do not receive enough
animal protein as evidenced by the distribution of meat with-
in the family. Secondly, the large number of households
which did not consume any animal protein points to both a

lack of understanding of the importance of animal foods and

59

the need to teach budgeting so that money will be available.

‘for meat. -Although vegetables were consumed often, a pot

of vegetables could be cooked for lunch on one day, and then

reheated and served at every meal until supper on the follow-

ing day. In this way, even vegetables fresh from the farm

lose most of their vitamin value. It is not expected that

education on the proper handling of vegetables will have much ’
impact, since too great an expenditure of time would be 1
required. However, an increased consumption of fresh fruits
would go a long way towards improving vitamin intakes.

Convincing mothers that their children need special
foods will probably be much harder than teaching them how
to prepare the foods. However, this study pointed out that
most of the women are very eager to learn new child care
practices. The amount of respect given to ”authorities" such
as medical personnel is also encouraging. The survey showed
a slight tendency to emulate "educated” people, and other
people in the community, providing that what these people
do is not too incompatible with traditional ways. An
educational program might do well to emphasize the value of
traditional Yoruba foods and the idea that modern people
prepare special foods for babies.

When it was seen that literacy is not related to
toddler mortality, a small analysis was done of child feeding
practices by level of education. Feeding practices did not
vary according to years of shooling. Many girls are forced

to quit school after one or two years, sometimes before they

 

60

are literate. If one were to teach child care to school
girls, it should be in the first two years Of primary school.
Because this study has pointed out that few women
are literate and few have access to radio programs, an inter-
personal rather than mass media approach to change would be
more effective. I
The amount of extra time and energy needed to prepare
special baby foods may be prohibitive for many of the working
mothers, especially those who trade and spend much of their
time away from home. For this reason, it might be desirable
that special teaching emphasis be given to food vendors.
Vendors could prepare part of their recipes especially for
babies, or some vendors might even find it profitable to
.specialize in baby foods in an area that is educated as to
the importance of the foods. Vendors, especially those who
hawk, are in an excellent position to advertise and popularize
locally made baby foods. Furthermore, the progress of the
program.could easily be assessed through sales of the foods.
There were significant differences in child feeding
practices between the two communities. These differences
would not affect what one would teach in a nutrition program,
but they would influence teaching methods and standards for
evaluating the program.' For example, one would want to
teach both Iwo and Ife mothers that all the foods listed in
Section E of the questionnaire are good for babies. But in
the Ife villages, where fewer of the foods were thought ap-
propriate, more time would have to be spent on this aspect

of the program than in the Iwo villages. Because of the

61

differences cited, one could not generalize the findings of
this study to other Yoruba communities.

This study suggests that the most important data
needed for planning an applied child nutrition program are
those relevant to the type of supplementary food used and
the age of introduction to stew because of their relation-
ships to toddler mortality. In addition, one would want to

find out what foods are thought acceptable for babies and

 

the prevalent beliefs about unacceptable foods. Information E“
as to the occupations of women, their reliance on food
vendors, and their exposure to outside sources of information

would also be useful in planning details of the program.

Co