Dietary intake of pregnant women and their infants in a poor black South African community

Tel: 021-9380242 Fax:021-9335519 Email: nelia.stevn@mrc.ac.za Abstract: Curationis 28(4): 12-19 The aims of this study were (i) to determine the dietary intake of women in a poor rural area during pregnancy and lactation, and (ii) to determine the nutritional status and dietary intake of their infants at age 6 months. We recruited 46 women, below 40 years old, in their 2nd trimester of pregnancy. The subjects were living in a rural area of Limpopo Province. Their heights and weights were recorded, as were their diets during pregnancy and for the first 6 months after delivery. We also recorded weights, lengths, and dietary intake of the infants at 3 and 6 months after birth. The subjects were living in severe poverty: none had running water and almost all did their cooking over an open fire. None of the subjects smoked and only one consumed alcohol. The diets of the subjects consisted mainly of maize, brown bread, sweetened beverages (cold drink and tea), and small amounts of vegetables and chicken. The diets were adequate in protein but were marginal in energy and in dietary fibre, and may be deficient in numerous micronutrients, particularly calcium, iron, zinc, niacin, folate, and vitamins A, C, E, and B6. This was seen during pregnancy and lactation. Blood analysis 6 months after birth revealed normal levels of vitamins A and E and an absence of anaemia. Body mass index (BMI) of the women was 23.9 ± 5.3 kg/m2 (mean ± SD) when measured 6 months after birth. Those above 25 years old had a higher BMI than did younger subjects (25.5 vs. 22.2; p= 0.028). Overall, 24% were overweight (BMI 25-30) while 9% were obese (BMI > 30). Most infants (93%) were breastfed for at least 6 months but exclusive breastfeeding was only done by 65% of mothers. One-third of breastfed infants also received formula. The use of formula while breastfeeding was twice as common among mothers aged above 25 years (46% vs. 23%). Early introduction of solid foods was very common in this group. Younger mothers introduced solids in the first month (51%) more often compared with older mothers who tended to introduce solids at 2-3 months (64%). The most common solid food given was maize meal porridge (by 78% of all mothers). Mean BMI was low at birth (< 15), but this reached a normal value by 6 months. A significant fraction of the infants was underweight or short (i.e., stunted), based on being below the 3rd percentile compared with NCHS standards. Underweight was seen in 15% at birth, falling to 10% at 6 months, whereas 15% had a short length at birth, rising to 35% at 6 months. Underweight tended to be more common in boys and

The aims of this study were (i) to determine the dietary intake of women in a poor rural area during pregnancy and lactation, and (ii) to determine the nutritional status and dietary intake of their infants at age 6 months.We recruited 46 women, below 40 years old, in their 2nd trimester of pregnancy.The subjects were living in a rural area of Limpopo Province.Their heights and weights were recorded, as were their diets during pregnancy and for the first 6 months after delivery.We also recorded weights, lengths, and dietary intake of the infants at 3 and 6 months after birth.The subjects were living in severe poverty: none had running water and almost all did their cooking over an open fire.None of the subjects smoked and only one consumed alcohol.The diets of the subjects consisted mainly of maize, brown bread, sweetened beverages (cold drink and tea), and small amounts of vegetables and chicken.The diets were adequate in protein but were marginal in energy and in dietary fibre, and may be deficient in numerous micronutrients, particularly calcium, iron, zinc, niacin, folate, and vitamins A, C, E, and B6.This was seen during pregnancy and lactation.Blood analysis 6 months after birth revealed normal levels of vitamins A and E and an absence of anaemia.Body mass index (BMI) of the women was 23.9 ± 5.3 kg/m2 (mean ± SD) when measured 6 months after birth.Those above 25 years old had a higher BMI than did younger subjects (25.5 vs. 22.2; p= 0.028).Overall, 24% were overweight (BMI 25-30) while 9% were obese (BMI > 30).Most infants (93%) were breastfed for at least 6 months but exclusive breastfeeding was only done by 65% of mothers.One-third of breastfed infants also received formula.The use of formula while breastfeeding was twice as common among mothers aged above 25 years (46% vs. 23%).Early introduction of solid foods was very common in this group.Younger mothers introduced solids in the first month (51%) more often compared with older mothers who tended to introduce solids at 2-3 months (64%).The most common solid food given was maize meal porridge (by 78% of all mothers).Mean BMI was low at birth (< 15), but this reached a normal value by 6 months.A significant fraction of the infants was underweight or short (i.e., stunted), based on being below the 3rd percentile compared with NCHS standards.Underweight was seen in 15% at birth, falling to 10% at 6 months, whereas 15% had a short length at birth, rising to 35% at 6 months.Underweight tended to be more common in boys and short length in girls.This study found that pregnant and lactating women had diets low in energy and micronutrients as reflected by the high prevalence of underweight at birth.Most infants were exclusively breastfed, but the benefit of this was offset by the early introduction of solid foods having a low energy and n u trient density.By 6 m onths, the prevalence of stunting had more than doubled.

Introduction
Several studies have furnished strong evidence that nutritional problems are frequently found in infants and children in South Africa.In particular stunting, as a manifestation of chronic malnutrition, is highly prevalent in children aged 1-3 years (26%) and decreases to 13% at ages 7-9 years (Labadarios et al., 2000:194).The higher prevalence at ages 1 -3 years was also found in a national anthropom etric study o f pre-school children in 1994 (SAVACG, 1995:106).Stunting was seen in 17% of those aged 6-11 months, and in 23% of children at both 1 -2 years and 2-3 years of age.These findings indicate, therefore, that stunting was already evident in many children by 6 months and that this number increased from there until 3 years where after it decreased in school-going children.The reasons for the high prevalence of stunting in children aged 1-3 years are not clear.However, it is postulated that this may be an outcome of poor diet 13 Curationis November 2005 during pregnancy, low birth weight of infants followed by an early and poor quality weaning diet.These factors have been described as determinants of poor n utritio n al status in o ther studies (Mamabolo etal., 2004:327;Steyn etal., 1993:10;Ladzani etal., 1998:63;Steyn, 2000:S11;Steyn etal., 1998:641).To shed light upon this problem , information is needed concerning the diets of women during pregnancy and of their infants.Unfortunately, the diets of pregnant women living in poor deep rural areas in South Africa have not been documented and neither has the diet of their infants during the first months following birth.Since complementary foods are commonly introduced during the first 6 months of life (Steyn et al., 1993:10: Steyn, 2000:S 11), this period may be critical in the causation of later stunting.
In this cohort study, we followed a group of women who were in their second semester of pregnancy at the start of the study.We recorded their diets, both during pregnancy and lactation.We also studied the diets and growth of their infants over a period of 6 months to determ ine whether dietary practices during this period could be responsible for the high levels of stunting seen from 6 months onwards.

Study area and participants
The study subjects were 46 women, aged below 40 years, who were in their second trim ester of pregnancy.They were randomly selected while visiting the Dikgale primary health-care clinic over a 3-m onth period.The subjects were divided into two groups: one of women aged below 25 years and another group above 25 years old.The reason for this was to determine whether infant feeding practices differ by m other's age.The Dikgale clinic is in a rural village in the Limpopo Province, South Africa.This area borders Zimbabwe, Botswana, and Mozambique.The area is known to be one of the poorest in the Country, with m ore that 50% o f adults being unemployed (Health Systems Trust, 2003:430).

Data collection and measurements
The data were collected either by nurses at the clinic or by trained fieldworkers, who were recruited from the community and trained for the required tasks.
In terv iew s were conducted in the subjects' language, namely Northern Sotho.Heights of women (without shoes) were measured using a stadiometer.Weight was taken with minimal or light clothing using a p o rtab le, electro n ic scale.Lengths of the infants were measured with the infant lying down stretched on a flat surface, as described by Lee and Nieman (1993:122).Weights of the infants were measured using a portable, baby scale, with the infant naked.The height and weight of the infants were compared with the reference data of the National Center for Health Statistics (NCHS), which are based on a population of North American children and is recommended by the World Health Organization (WHO) for international use (NCHS et al., 1977:50).Blood pressure of the women was m easured during pregnancy by trained nursing sisters using a mercury sp h y g m om anom eter according to recommended procedures (American Heart Association, 1988:502A-508A).Blood samples were taken from infants and mothers who allowed this (n=20) and samples were analyzed for retinal and átocopherol (Milne & Botnen, 1986:874).The small number of blood samples obtained is because this is difficult in rural areas; there are many superstitions associated with this as well as a fear of being found to be HIV positive.Dietary intake of the pregnant women (2nd and 3rd trimesters) was obtained through a repeated 24-hour dietary recall method, one weekday and one weekend day, through personal interview s by the fieldworkers, as recommended (Nelson & Bingham, 1997:143).The dietary intake of the lactating women was also obtained by means of repeated 24-hour recalls during post-natal visits.Dietary intake o f the infants was obtained using a d ietary history, through personal interviews by the fieldworkers with the mother or caregivers at 6 months after the birth of the baby.The fieldworkers reco rd ed food q u an tities w ith the assistan ce o f m easuring cups and various food models.The dietary data were analysed for nutrient content using the Medical Research Council's Food Composition Tables (Langenhoven et al., 1991a) and Q uantity Tables (Langenhoven et al., 1991b).Quality control checks was done on several occasions where different fieldworkers conducted interview s on the sam e subjects and the results were compared.No significant differences were seen when this was done.
A pilot study was undertaken with 12 pregnant women before the main study to test the q u estio n n aires and the accuracy of the data collection by the fieldworkers.

Statistical analysis
Statistical analysis was performed using SPSS (Statistical Products and Service Solutions) software 11.0 for Windows.Student's t-test was used to compare mean body mass index (BMI) in the older and younger age groups.The effect of age on formula use while breastfeeding was analysed using the Chi-square test.

Ethical approval
Ethical approval was obtained from the Ethics Committee of the University of the North.

Observations during pregnancy
Of the 46 women in the study, 22 (48%) were aged between 13 and 24 ("younger group"), and 24 (52%) were aged between 25 and 40 ("older group").Most of the women (88%) in the older group were married but only 12% of the younger group.Only one-third of the women had received an education beyond grade 9. Most of the fathers (59%), however, had received an education at least to grade 11.The m others were mostly (63%) unemployed, while 33% were students.With fathers, however, a much higher proportion (52%) was employed, while 22% were unemployed and 26% were students.There was an average of 8.1 persons living in each household.The m others had had an average of 2.1 p rev io u s p reg n an cies.H ousehold income was mostly in the range R500-R2000 per month (equivalent to about 170-680 U S $/m onth).None o f the subjects had running water in their house and 94% did their cooking over an open fire.
No subjects were smokers, and no one reported being HIV positive although HIV/AIDS has a high prevalence in South Africa.Measurements at the clinic during pregnancy indicated that 17% had an elevated blood pressure (BP) (systolic BP > 140mmHg and diastolic BP> 90 mmHg).However, 37% had an elevation of either systolic or diastolic pressure.BMI was determined at the time of the subjects' first visit to the clinic when they were about 4 -6 months pregnant (Table 1).The mean BMI was 27.2 ± 5.5 kg/m2 (mean ± SD).It was higher in the older group than in the younger group (28.8 vs. 25.5;p = 0.03).The proportion with a BMI = 25 was 61%.Analysis of the diets during pregnancy revealed a number of inadequacies (Table 2).The diets were marginal in energy and dietary fibre, though the content of protein was adequate.Intake of several minerals and vitamins was well below the Recom m ended D ietary A llow ances (R D A ) (M onsen, 1998;Food and Nutrition Board, 2001).
Most subjects (74%) ate three meals daily during pregnancy, while 85% also had between-meal snacks.The most common snack was fruit, which was eaten by 76% of subjects.
The beverages m ost com m only consumed were cold drinks (94%) and tea (91 %), followed by coffee (34%); all contained sugar.O nly one subject consumed an alcoholic beverage.Overall, the most common foods eaten were maize, brown bread, cold drink, and tea, followed by vegetables (most commonly spinach, pum pkin, carrots, cabbage, tom atoes, and onion), and chicken (usually once a week; typically the cheaper cuts such as gizzards, and feet).
The most common type of food cravings were for fruit and vegetables, being reported by 67%, followed by cold drinks (20%), and milk and milk products (15%).

Observations during lactation
When measured 6 months after birth, the BMI of the women was 23.9 ± 5.3 kg/m: (mean ± SD).As during pregnancy the   Health, 2002:247).The dietary assessments were repeated during the period of lactation (Table 3).This was done at every clinic visit.Energy intake was higher than that seen during pregnancy.However, a similar pattern of sufficiency and inadequacy was seen for most nutrients.The most noteworthy changes were that intake of calcium, vitamin A, and vitamin C was even more sub-optimal than before.This can be largely explained by the decreased consum ption of milk, in the case of calcium, and of fruit, in the case of vitamin C. Blood samples, which were obtained from 20 subjects 6 months after birth, were analysed for vitamins A and E. There was no indication of subnormal serum levels of either vitamin (< 1.05 mmol/L for vitamin A or < 11.6 mmol/L for vitamin E).
The normal serum level of vitamin A is consistent with the fact that the women had been given a supplement of the vitamin post-natally.Blood analysis also indicated that none of the subjects had anaemia, consistent with the fact that they had been given iron supplements at visits to the antenatal clinic.

Infant Feeding
Most infants (89%) were cared for by their mother.The great majority of infants (93%) were breastfed for at least 6 months.However, exclusive breastfeeding was only done by 67% of mothers.Exclusive breastfeeding in this context means that no other milk was given, yet solids were often introduced from an early age.Onethird of breastfed infants also received form ula. Use o f form ula w hile breastfeeding was twice as common among subjects in the older group (46% vs. 23%; p=0.029).Mothers cited the following reasons forgiving formula milk in addition to breast milk: had to go back to school (53%), own milk too weak or too little (27%), and m others' advice (13%).Nearly half of infants (43%) were given solid food within their first month, rising to 56% by 3 months.The most common solid food given was maize meal porridge (given by 78% of all mothers).Blood samples were obtained from 20 infants and analysed for serum levels of vitamins A and E. Two of the infants had a m arginal vitam in A d eficien cy (< 0.7 mmol/L) but none had a sub-normal level of vitamin E (< 11.6 mmol/L).As with the mothers the adequacy of the vitamin A blood levels reflects the supplements of the vitamin that the infants had been given.
The weights and lengths of the infants are shown in Curationis November 2005 Dewey et al., 1992Dewey et al., :1035;;Hop et al., 2000Hop et al., :2684)).A significant fraction of the infants was underweight or short (i.e.below the 3rd percentile compared with NCHS standards) (NCHS, 1977:50).Underweight was seen in 15% at birth, falling to 10% at 6 months, whereas 15% had a short length at birth, rising to 35% at 6 months.Underweight tended to be more common in boys and short length in girls.

Discussion
This study, carried out in a poor rural community in the Limpopo Province of no rth ern S outh A frica, p resen ts a snapshot o f the nutrition and health status of women during pregnancy and lactation.Some aspects of the findings were encouraging: based on self-reported information none of the women were smokers or HIV positive and only one consumed alcohol.These findings are consistent with the fact that there are strong taboos against sm oking and alcohol consumption by women in this culture.However, these observations are o v ersh ad o w ed by serio u s dietary problems.The data reveal that the women ty p ically consum e a diet w hich is adequate in protein but m arginal in energy and dietary fibre, yet may be deficient in numerous micronutrients, particularly calcium, iron, zinc, niacin, folate, and vitamins A, C, E, and B,.This  Health. 2002:244).B reastfeed in g was seen to be the dom inant feeding practice (93% of mothers were breastfeeding at 6 months).However, the benefits of this were partly neutralised by the fact that one-third of breastfed infants also received formula.
A more serious problem is the early introduction of solid food, especially maize meal porridge.Half of infants were given solid food within their first month.Set against this nutritional background it is unsurprising that a significant fraction of the infants were underweight.A m ore serio u s problem was the prevalence of stunting: 35% had a length that was below the 3rd NCHS percentile at 6 months.
An earlier South African study reported high rates of stunting in infants aged 12-23 months (SAVACG, 1995:106).The rates were highest in rural areas, including Limpopo Province.These observations were confirmed on children aged 1 -3 years (as well as at ages 4-6 years) in the National Food C onsum ption Survey (Labadarios etal., 2000:194).Our results may also be compared with another study carried out at about the same time in an adjacent community (Mamabolo et al., 2004:327).The latter study included more subjects (276) but only considered the post-natal period.In that study, similar observations were made concerning infant feeding practices.As seen in this study many infants had poor weight gain while a much more common problem was a high prevalence of stunting (36% had a length that was < 3rd percentile of the NCHS median at 6 months and this continued unchanged until an age of one year).That study also revealed that factors predictive o f a better growth pattern were electricity in the house, and a mother who was employed and was better educated.These three parameters are probably indicators of a lesser degree of poverty.
Low birth weight has been associated with adult developm ent o f obesity, cardiovascular disease, type 2 diabetes, and hypertension (Barker & Godfrey, 2001:255).M uch less is know n concerning the long-term consequences of being short in infancy.However, since the nineties there have been a number of stu d ies, w hich have show n an association between early stunting and overweight at a later stage, particularly in studies of populations undergoing the nutrition transition (Doak etal., 2005:129;Popkin etal., 1996:3009).
In agreement with our findings, other studies carried out in South Africa during the 1990s also reported high levels of b reastfeed in g as w ell as the early

Curationis November 2005
introduction of solid food (Steyn et al.. 1993:10;Ladzani et al., 1998:63;Steyn, 2000:SI 1;Steyn et al.. 1998:641).The W orld H ealth O rganizatio n has recommended that infants be exclusively breastfed for the first 6 months (WHO/ UNICEF, 1990).This appears to lower the risk of stunting, especially among those living in poverty (Dewey et al., 1992(Dewey et al., :1041;;Hop et al., 2000Hop et al., :2688;;Villalpando & Lopez-Alarcon, 2000:546).Accordingly, efforts should be m ade to educate mothers better in this regard.However, this is unlikely to be more than a partial answer to the problem of stunting; the clearly deficient diets of the women described in this study are also an urgent and important target for intervention.In this regard, poverty is the dominant obstacle to progress.One approach to improve the nutrition and health status of the infants would be the expanded use of supplements, both for the women during pregnancy and lactation as well as for their infants.For example, it has been shown that giving vitamin A supplements to infants and young children can significantly reduce mortality rates (Rahmathullah, 2003:254;Ross, 2002Herrera, 1992:267).The fact that fortification of maize and flour has been introduced since October 2003 should assist in alleviating this problem.

Conclusion
This study found that both pregnant and lactating women had a diet low in energy and micronutrients.This was reflected by the high prevalence of underweight at birth.The m ajority o f infants were exclusively breastfed but the benefit of this was offset by the early introduction of solid foods.The prevalence of stunting was 15% at birth but increased to 35% at 6 months.

Table 1 . Mean body mass index (BMI) of pregnant (4-6months) and lactating women attending the antenatal clinic at Dikgale
* Student t-test is significant at p < 0.05

Table 2 . Nutrient intakes of the study population during pregnancy and lactation as analysed from repeated 24-hour recalls
* RDA: Recommended dietary allowances are the 1989 American RDA, but values for vitamin B6, niacin, and folate are based on recent updates.