Are breastfeeding and complementary feeding of children of adolescent mothers different from those of adult mothers?

A amamentação e a alimentação complementar de filhos de mães adolescentes são diferentes das de filhos de mães adultas?
J Pediatr (Rio J) 2003;79(4):317-24


Adolescence is a critical period for humans during which intense biological and psychosocial transformations occur (1). The occurrence of pregnancy and maternity during this stage of life implies a number of different changes which, when added together, force a reevaluation of the future with a, generally unwanted, child. Pregnancy and maternity during adolescence are considered serious public health problems all over the world, primarily within low-income families; populations whose health is at risk (2).

In developing countries, including Brazil, research has been concerned with preventing infant malnutrition by means of promoting health and encouraging breastfeeding, primarily with adolescent mothers. The WHO, UNICEF and many research papers emphasize the necessity and importance of promoting exclusive maternal breastfeeding, as an efficient method of: preventing infectious diseases (gastrointestinal and respiratory complaints, otitis media), necrotizing enterocolitis, ulcerative rectocolitis, Crohn's disease; protecting against later allergy, insulin dependent diabetes mellitus, obesity and cognitive dysfunctions; and of favoring infant growth and development (3-7). Knowledge of the advantages of breastfeeding and human milk for the health, growth and development of children has already been well cascaded (5), however Gouvêa (4) also stresses the important contribution that breastfeeding has to make in speech development. In the best-organized health services in Brazil, multidisciplinary teams work on the self-esteem of pregnant adolescents and those in confinement, encouraging them to assume personal responsibility for their babies, strengthening the mother-child bond and favoring breastfeeding (4).

The Innocenti Declaration (8), in 1990, upheld exclusive breastfeeding during the first 4 to 6 months of life as the ideal objective for infant health and nutrition. The WHO currently recommends: 1) exclusive breastfeeding for the first 6 months of life; 2) adequate and opportune supplementary feeding; and 3) maintenance of supplementary maternal breastfeeding until 2 years of age or more (9,10). The latest National Demographic and Health Census (Pesquisa Nacional sobre Demografia e Saúde) 5, performed in 1996, showed that maternal breastfeeding among Brazilians is well below the WHO recommendations, with the median for exclusive breastfeeding at just one month, total breastfeeding at seven months and breastfeeding continuing for one year for 41.0 % of crianças (9).

A study by Almroth (7) et al. in Lesoto shows that exclusive breastfeeding is not practiced and that mothers, grandmothers and even health service nurses believe that offering water to children is important and necessary, promoting predominant breastfeeding. Cultural habits favor this practice from which springs the wide variety in predominant breastfeeding in Brazil (11-13).

Since March 2001, after an extensive review of the literature, the WHO adopted six months as the age at which to introduce supplementary feeding (14). In Honduras, randomized intervention studies did not reveal growth advantages in children who began with supplementary foodstuffs at four months when compared with children exclusively breastfed until the 6th month (15). Mehta (16) et al. claimed that infants consuming industrialized foodstuffs have a lower level of protein and fat ingestion, but that this had no effect on growth and that early introduction of solids to the diet does not alter growth or corporal composition during the first year. Burrows (17) et al. did not find any significant differences between the children of adolescents mothers and those of adults in terms of the ingestion of proteins, carbohydrates and lipids, while for Carruth (18) et al., the children of adolescents mothers ingested more fats and started to feed themselves with their own hands and ate cereals earlier than those of adult mothers although there was no difference for fruit, vegetables and meat.

This study took as its objectives the investigation of breastfeeding during the first year of life and the type of supplementary foods used at the end of the first year with the children of adolescent mothers; the comparison of exclusive, predominant, complete and total breastfeeding; the identification of the rate of breastfeeding for a complete year; the comparison of consumption of meat, offal, eggs, vegetables, cereals and milk based foods by the children of adolescents mothers with the consumption of the children of adult mothers at one year.

Materials and Methods

This is a bi-directional cohort study for which children were selected retrospectively and analyzed prospectively, when one year old. The research protocol was approved by the Committee for Ethics in Research of the Center for Integral Women's Health Care (CAISM- Centro de Atenção Integral à Saúde da Mulher) - /UNICAMP. This is part of a larger investigation of the growth and development of the children of adolescent mothers, for which the sample size calculation was based on the quantative variables of interest (weight, length, skull and upper arm circumference and skin-folds at the triceps and below the shoulder blade), using Frisancho as a reference (19), by means of the formula:

n = (2 (σ)2/(d)2).ƒ (α,β)

where is the standard deviation of the anthropometric measurements used, d is the difference to be considered significant between groups and is the type I and II error function.According to this calculation, 102 subjects from each group would be necessary, although it was actually possible to identify and assess 122 children of adolescent mothers and 123 children of adult mothers. The sample was sequential, that is the children were selected from the files of CAISM/UNICAMP by order of birth, according to the criteria: born at full term, single birth, weight > 2.500 grams, no congenital malformations, syndromes or diseases which interfere with growth and development; mothers less than 20 years old at delivery and mothers between 20 and 30 years old, all primaparous and without serious clinical or mental illnesses which could interfere with the care of their children. The selected children were located and were evaluated at home or at the Center for Pediatric Investigation (CIPED - Centro de Investigação em Pediatria -) of UNICAMP until the calculated sample size was achieved, and then passed by 20 % for both groups. A search was made to locate the mothers and their children in peripheral neighborhoods around greater Campinas-SP and nine adjacent towns. As losses were due to failure to find the addresses as registered at the CAISM or to change of address and/or city and/or State and were similar for the two groups, (four adult mothers and three adolescent mothers) there were not considered to be selection distortions, and the lost children were substituted by others who were selected later, according top the inclusion criteria. As a result of the difficulties in locating and accessing the addresses, the data collection phase extended from October 1999 to April 2001, performed by a single researcher/interviewer.

All of the mothers signed an Informed Consent Form, as recommended in Resolution nº 196 dated 10/10/1996 which deals with research involving humans of the National Health Council of the Health Ministry. The mothers chose the time and place for data collection and individual interviews.

The following were employed: 1) CAISM/UNICAMP record - maternal gynecological, obstetrical and neonatal data; 2) Record of child - information about neuropsychomotor development, in addition to the anthropometric data and details of immunization from birth to 1 year, transcribed from the child's "Caderneta da Criança" (child log-book); and 3) Home Visit Record - socioeconomic data, morbidity and nutritional investigation.

Education was defined as normal for age or not, taking into account that, at seven all children should have entered primary education, the eighth grade of which should be reached at 14/15 years of age, according to the Ministry of Education; and that adults, being over 20 years old, should have completed their secondary education. Conjugal status was defined as "united" for those who were married or had partners and "not united" when single, separated/divorced or widowed. Maternal occupation was categorized as: at home ("homemaker") and away from home (all paid employment). The family monthly income was divided by the number of people in the household and classified as: < 0.5; 0.5 a 1.0 (exclusive); 1.0 to 2.0 (exclusive) and > 2.0 minimum salaries per capita.

Exclusive breastfeeding signifies that the child ingests only mother's milk, without any supplements, while predominant breastfeeding is the ingestion of mother's milk and water and/or teas and/or juices and/or oral rehydration salts. Full breastfeeding = exclusive breastfeeding duration + predominant breastfeeding duration; and total breastfeeding during the 1st year of life = predominant and/or exclusive breastfeeding duration + breastfeeding supplemented by other foods. Mixed milk source = ingest mother's milk and cow's milk. Nutritional supplements were defined as: water, teas and/or substitutes for mother's milk, consumed during the first months of life; while nutritional complements are foods used to reinforce human milk from the sixth month onwards (9,13).

The nutritional part of this study investigated breastfeeding during the 1st year of life and nutritional complements consumed towards the end of the 1st year. The mothers were asked: 1) whether they breastfed, with a positive being registered if the child left the maternity unit suckling at the breast and continued to do so for a minimum of 20 days; 2) for how long (in days) exclusive, complete and total breastfeeding lasted; 3) at what age (in months) water and/or tea were introduced; 4) if their children still suckled at the breast at the time of interview; 5) what type of milk sources their children had at one year of age: maternal, mixed or cows' milk; and 6) if their children, by twelve months, had never ingested, ingested daily or ingested occasionally: meat (bovine, fish and chicken), offal (chicken or beef liver), eggs, cereals, vegetables and fruits. Daily being when the child ingested the nutritional complement every day and occasionally when the frequency of ingestion was irregular.

The following tests were used: McNemar before and after test (conjugal status of the adolescent mothers when they got pregnant and when the child had reached a year); Chi-square or Fisher's exact test, for the comparison of proportions; and the Wilcoxon (Breslow) test to compare breastfeeding survival curves. The analysis of survival time of exclusive breastfeeding, performed by means of the Kaplan-Meier method, considering only the interruption of breastfeeding with mother's milk as an event. The acceptance level was set at 5%.


Table 1 contains certain characteristics of both the adolescent and adult mothers, highlighting the conjugal status, type of employment and per capita income as they presented differently for the two groups. More than 50% of the adolescent mothers had a per capita income less than one minimum salary, when compared with the adult mothers (p = 0.014). The adolescents worked away from home less than the adults (22.1 %(27) v 35.8 %(44); p = 0.019). At the point of getting pregnant, the conjugal status "not united" was more common among the adolescents group (52.5 %(64) v 24.4 %(30); p = 0,001). During the first year of their children's' lives, 25.4 %(31) of these mothers got married or entered into a relationship, resulting in a significant shift towards the "united" status.

Table 1 -
Characteristics of adolescent and adult mothers

Table 2 describes the characteristics of the children in the sample. Length and weight at birth and ponderal index after one year were lower for the children of adolescent mothers, giving p = 0.027, p = 0.019 and p = 0.041 respectively, when compared to the children of adult mothers, applying ANCOVA and correcting for sex.

Table 2 -
Descriptive analysis of children of adolescent mothers versus children of adult mothers

Table 3 shows that there is no difference between the children of adolescent and adult mothers in terms of breastfeeding, breastfeeding for a whole year, types of milk source and ingestion of offal, fruit, vegetables and cereals at one year. Seven of the children of adolescent mothers were defined as not having been breastfed: one whose mother was HIV positive and six others who suckled for less than twenty days; whilst among the children of adult mothers, two did not breastfeed due to severe mastitis and three others were breastfed for less than twenty days. The mothers who breastfed for less than twenty days had personal and family problems which de-motivated them from maintaining breastfeeding. It was detected that the children of adolescent mothers ingested less meat and tended to ingest more eggs than those of adult mothers.

Table 3 -
Feeding pattern in the first year of life of children of adolescent and adult mothers

Employing the same methods of analysis as for exclusive breastfeeding, the following were analyzed: the point at which predominant breastfeeding began, full breastfeeding continued and total breastfeeding during the first year of life, the mean and median averages and 95 % confidence intervals (Kaplan-Meier survival method) of which are given in Table 4. The median for exclusive breastfeeding was 90 days for both groups, while the median for total breastfeeding was 6 months for adolescent mothers and 8 for the adults.

Table 4 -
Breastfeeding pattern of children of adolescent and adult mothers during the first year of life

Comparative curves for exclusive breastfeeding, start of predominant breastfeeding, full breastfeeding and total breastfeeding during the first year are given in Figures 1, 2, 3 and 4, having no statistically significant differences, when the Wilcoxon or Breslow tests were applied.

Figure 1 -
Análise do tempo de amamentação exclusiva entre os grupos (Teste de Wilcoxon ou Breslow: p = 0,632)

Figure 2 -
Análise da idade de início da amamentação predominante (Teste de Wilcoxon ou Breslow: p =0,077)

Figure 3 -
Análise do tempo de amamentação completa (Teste de Wilcoxon ou Breslow: p = 0,245)

Figure 4 -
Análise do tempo total da amamentação durante o primeiro ano de vida (Teste de Wilcoxon ou Breslow: p=0,269)


Breastfeeding patterns were not found to be different between the two groups, supporting Valenzuela (20) and negating other works which have shown adolescent mothers to have greater adherence to breastfeeding than adults (21,22). Recent studies show that the improved rates of breastfeeding are a result of the intervention of health professionals during prenatal, immediate postpartum and postnatal (23,24); and further show that expectant mothers and adolescent mothers and also women having their first child benefit most from educational activities aimed at promoting breastfeeding (23,24), reinforcing the importance of these educational activities which can be developed within primary care services. A positive influence promoting breastfeeding is observed from grandparents (11,25), while a study by Giugliani et al. (26) suggests that their interference has a negative effect on the duration of breastfeeding.

The median for exclusive breastfeeding was 90 days for both groups. This is better than that observed in Montes Claros-MG which was only 27 days (27). There the median for total maternal breastfeeding was 8.7 months whereas in this study it was 6 months for the adolescent mothers and 8 for the adults. Although this is a large difference (two months) this result was not statistically significant (Wilcoxon or Breslow test: p = 0.269), which means that chance is not ruled out. It is however important since the adolescent mothers presented a total breastfeeding median lower than the adults. A qualitative study may be able to explain why this group breastfed longer. Researchers have found exclusive and predominant breastfeeding rates of: 0.0 % and 62.9 % in Belém-PA (11) in 1997, 10.3 % and 32.3 % in Embu-SP (12) in 1999 and 37.8 % and 17.8 % in São Carlos-SP (13) in 2000, respectively. This diversity in exclusive and predominant breastfeeding prevalencies shows the influence of regional cultural habits and the need to promote breastfeeding incentive programs tailored to each region, taking into account the fact that Brazil is a continental country, rich in different cultures requiring regionalized approaches to promote exclusive breastfeeding. A single model for the promotion of maternal breastfeeding cannot work in a country with such social and cultural diversity.

The rates for breastfeeding lasting a full year were 35.3 % and 28.5 % for the children of adolescent and adult mothers, respectively, while, in Belém-PA (11) they were 62.6 % and 45.4 % in Feira de Santana-BA (28), 27.6 % in Embu-SP (12) and 41.0 % in Montes Claros-MG (27), maintaining regional diversity.

Comparing the educational histories of the adolescent and adult mothers no difference was observed. Care was taken to correlate years of education with maternal age, avoiding distortions from equating adolescents with few years' schooling with adults who had completed their education. This is an important variable since a return to breastfeeding has been observed among better educated mothers after a phase in which it declined due to the introduction of women into the workplace (29).

At the point of getting pregnant, there was a higher proportion of single or separated adolescents which had changed by the time of the interview with a significant shift in status, entering into a relationship or marrying, generally with the father of the child or an older man. The change in conjugal status is important to the emotional and financial stability of these mothers and is a positive factor for their children's development.

In both groups the number of mothers who smoked was small and there was no statistically significant difference between the groups. This is in contrast to a recent study which showed that during the breastfeeding period, smoking decreases in proportion to education and income (30). Another study indicates the ill effects of smoking, which affects the production of the mother's milk and the development of breastfed children of smoking mothers (9).

It was found that the adult mothers more often worked away from home, in order to help with the family budget, achieving a greater per capita income than the adolescent mothers, which intervenes in the acquisition of foodstuffs which are important for the adequate growth and development of their children, a fact also observed by Lima (31) et al. Research shows an association between a higher maternal socio-economic and educational level and longer duration breastfeeding (25), but, despite the adolescent mothers in this sample being poorer than the adults, there was no difference in terms of education and breastfeeding.

Families are responsible for the nutritional behavior of children by means of social learning influenced by psycho-social and cultural factors. In this sample the socio-economic factor had an effect on nutritional practices during the first year in that the children of the adolescents, who had a lower income, ingested significantly less meat than the adults' children, with no differences in terms of other, cheaper, nutritional complements. The adolescent mothers' lower buying power was probably responsible for the substitution of meat by eggs.

This study has helped to demystify the relationship, which has been described as unfavorable in the literature, between adolescence/breastfeeding/nutrition during the first year of life, in that no statistically significant differences were observed between the groups, other than greater meat consumption by the adult mothers' children and increased ingestion of eggs by the children of adolescent mothers whose lower buying power meant they could acquire less meat.