SEARCH

SEARCH BY CITATION

Keywords:

  • (very) low birthweight;
  • infants;
  • follow-up;
  • growth potential of (V)LBW infants;
  • Ghana

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Summary This prospective and descriptive study was conducted to evaluate the growth and survival of 105 low birthweight (LBW, 1000–2000 g) infants discharged during a 4-year period from Agogo Hospital, Ghana, and followed from birth until the age of 4–9 years. Thirty-two babies were very low birthweight children (VLBW, 1000–1500 g) and 73 (70%) were of moderately low birthweight (MLBW, 1501–2000 g). At the age of 3 years, 15 children (14.2%) had not come for follow-up; of the remaining 90 children, nine (10%) had died, five during the first 3 months of life. At follow-up from 4 to 9 years of age, two more children could not be traced and another two had died. Compared to a local reference population and the WHO standard, growth of survivors lagged behind and caught up only slightly during the first 3 years of life. From 3 to 9 years of age, median growth impairment increased, which either suggests impaired growth potential or poor health and inadequate nutrition. This long-term study confirms that LBW (1000–2000 g) infants, particularly VLBW children, are at high risk.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

In developed countries, low birthweight infants (LBW, generally defined by a weight of ≤ 2500 g), particularly those with very low birthweight (VLBW, ≤ 1500 g) are considered to be at high risk due to their higher neonatal mortality rate and incidence of handicaps among survivors. After the introduction of intensive care for LBW infants had lowered neonatal mortality, the quality of life of survivors became a matter of growing concern ( Verloove-Vanhorick & Verwey 1987). However, in developing countries facilities for intensive care for LBW infants are often minimal due to lack of qualified staff and equipment. Hence neonatal and postneonatal mortality rates remained high. Almost nothing is known about the incidence of handicapped children among survivors, and to our knowledge there is no information on prognosis and long-term growth and development of LBW children beyond 12 months. This is not surprising because follow-up is difficult, often due to irregular attendance after discharge and insufficient home-visit services. Therefore it seemed worthwhile to report our experience of follow-up on the survival and growth of 105 children with birthweights of 1000–2000 g over 4–9 years. The children were discharged from hospital over a 4-year period and living in Agogo and nearby communities within a 30 km radius.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Only children born in or admitted to hospital with a birthweight of 1000–2000 g were included. We did not include LBW children with a birthweight of 2000–2500 g, as they are not considered a high-risk group in most developing countries ( Mbise 1977; Boersma 1979). Infants with a birthweight < 1000 g were excluded because of their small number (= 2). There were 3 pairs of twins and two sets of triplets in the study population.

After birth, selected LBW infants were cared for in locally made wooden incubators in a neonatal unit (nursery) without sophisticated equipment or intensive care provisions ( van der Mei 1994). Staff members had undergone a simple training program and mothers fully participated in the care for their children. If the condition of mother and child permitted, all children were breastfed from birth and discharged as soon as weight gain on exclusive breastfeeding had been established (range 3–30 days), irrespective of the individual weight. In the study area most children are breastfed for at least 18–24 months. Supplementary feeding usually starts at the age of 6 months or earlier and consists of porridge mostly made of boiled, fermented corn dough with sugar.

Study population

Of 161 LBW children discharged from the Nursery of Agogo Hospital during the study period (1984–88), 105 children were included. There were 48 boys and 57 girls; 32 children were very low birthweight (VLBW, 1000–1500 g); 73 moderately low birthweight (MLBW, 1501–2000 g). Fifty-six children were excluded because they did not live in Agogo or nearby communities. Since average birthweight (1615 g) and incidence of VLBW (30%) in the study group were similar to those of the whole group (1609 g and 32%) we felt justified in omitting these 56 infants. Before discharge, all children were provided with the ‘Road to Health’ card; a copy of this card, with the names of child, mother, father and their address remained at the Nursery.

Mothers were asked to regularly bring their children to the Nursery for check-up: weekly at first, every fortnight or month after 6–8 weeks depending on clinical condition. At their visit the children were weighed on a beam-balance scale by a member of the Nursery staff. Weights were recorded on the Road to Health card as well as on its copy; the mothers received nutrition advice and necessary vaccinations were given. From the age of 6–12 months the children were seen at the Under Fives Clinic (UFC) of the hospital. There they were weighed on a beam-balance or a hanging scale under supervision of the paediatrician in charge (v.d.M.). Weights were recorded bimonthly during the first year of life and 6-monthly at the age of 12–36 months. Table 1 shows the attendance rate at various measuring points; unfortunately, many children did not come regularly for follow-up. In order to obtain sufficient numbers, weight and height of all children of 12 ± 1, 18 ± 1, 24 ± 1, 30 ± 1, 36 ± 1 months and 48 ± 6, 60 ± 6, 72 ± 6, 84 ± 6, 96 ± 6 and 108 ± 6 months of age were combined to establish measuring points.

Table 1.  Attendance of low birthweight children (1000–2000 g) at measuring points Thumbnail image of

Incremental weight gain was calculated by dividing the differences of weight at two measurement points by the number of months between these points. For the incremental weight gain calculations of LBW infants, only those children (= 38) were included who reported at least five times or at four consecutive times during the first 3 years.

The second part of the study consists of follow-up visits at home by medical and nursing students in the period 1992–94, when the 105 LBW children had reached the age of 4–9 years. Thirty boys and 30 girls could be examined and measured; most once, some twice. When a child was not at home but the occupants knew that he/she was alive, the child was recorded as ‘alive, not seen’. This happened on 17 occasions.

Reference populations

The reference populations consisted of:

  • 1217 children, including LBW (603 boys and 614 girls), born at Agogo Hospital in 1988 ('local birthweight data');

  • 694 children (336 boys and 358 girls), 2–37 months old, who visited the UFC at Agogo Hospital or Child Health Clinics in the Agogo area from 1992 to 1994 ('local growth data 2–36 months').

  • 601 apparently healthy children (321 boys and 280 girls) attending nurseries and primary schools in Agogo in 1998 ('local growth data 4–9 years');

  • international WHO standard

The Agogo reference groups represent a cross-section of the local 0–9 years-old population.

Statistics

Median birthweights of LBW infants were compared with local birthweight data. Median bodyweights from 2 to 36 months were compared with cross-sectional local growth data 2–36 months. Median cross-sectional weight and height measurements at the ages of 4, 5, 6, 7, 8 and 9 years were compared with cross-sectional local growth data 4–9 years. All anthropometric measurements were also compared to the WHO standard ( Figures 1, 2, 3).

image

Figure 1. Weights of LBW (1000–2000g) children vs. reference groups.

Download figure to PowerPoint

image

Figure 2. Heights of LBW (1000–2000g) children vs. reference groups.

Download figure to PowerPoint

image

Figure 3. LBW (1000–2000g) children vs. reference groups. Incremental weight gain (kg/month).

Download figure to PowerPoint

Differences between male/female distributions of the VLBW and MLBW groups were tested with the Fisher exact test. Differences in attendance rate (traced/not traced) between VLBW and MLBW groups, as well as the survival rate of VLBW and MLBW groups, were tested by means of an odds ratio. Different survival rates and differences in attendance (traced/not traced) between the sexes of all LBW children were tested with the homogeneity of odds test.

Differences in weights, heights and incremental weight gain between the study group and Agogo's local growth data 2–108 months were tested with the Mann–Whitney U-test. Weights were tested at the measuring points 2, 6, 18, 48 and 96 months; heights at the ages of 48 and 96 months and incremental weight gain at the age periods 0–2, 2–4 and 4–6 months, respectively. In all cases P-values < 0.05 were considered statistically significant. Z-scores for weight/ height-for-age were calculated by subtracting the median weight/height of the reference population at the child's age from the child's weight/height, and dividing by the standard deviation (SD) of the weight/height of the reference population at that age ( Shann 1993).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Survival of LBW infants from 0 to 108 months

Table 1 shows the children's gradually decreasing attendance through the first 3 years of life from 63% to 17%. Table 2 compares characteristics of the two LBW subgroups of VLBW and MLBW infants. There was no statistically significant difference between the sex distributions for VLBW and MLBW children (Fisher exact test, = 0.0576). Table 3 shows the outcome of VLBW and MLBW boys and girls at the age of 3 years. Fifteen children (14.2%) could not be traced. There was no statistically significant difference between the sexes for traced and not traced (homogeneity of odds: = 0.1627). However, there was a significant difference between the VLBW and MLBW male and female children according to traceability: VLBW could be traced more frequently (odds ratio = 7.261 (95% CI = 1.012–321.8), = 0.047) ( Table 3a).

Table 2.  Some characteristics of very low birthweight (1000–1500 g) and moderately low birthweight infants (1501–2000 g) Thumbnail image of
Table 3.  Distribution of VLBW, MLBW and sex in children alive, dead, traced and not traced at age 3 years Thumbnail image of

Two boys and seven girls (10%) of the LBW children followed (= 90) died during the first 3 years; five, of whom four were VLBW, during the first 3 months of life. In three children anaemia was a contributing cause of death; the others died of various or unknown causes. There was no significant difference in survival between male and female children (homogeneity of odds: = 0.9909). MLBW children had a better chance of survival (95%) than VLBW children (81%; odds ratio = 4.48; 95% CI = 0.86–29.32; = 0.04 one-sided) ( Table 3b).

Of 81 LBW children followed at 4–9 years of age, another two could not be traced. Of the remaining 79 children two had died, both girls. Seventy-seven LBW children (88%) of those followed since birth up to the age of 4–9 years (= 88) were still alive, 37 boys and 40 girls.

Growth of LBW infants from 0 to 108 months

Figure 1 compares weights of LBW children (VLBW and MLBW) from the age of 0–108 months to local growth data 2–36 months and 4–9 years and to the WHO standard. At the ages of 2, 6, 18, 48 and 96 months LBW children were significantly lighter than Agogo reference groups (Mann–Whitney U-test; exact < 0.05, 2-tailed). LBW boys and girls doubled their birthweight at the age of 2.5 months and tripled it at 5 and 5.5 months, respectively. From birth to 36 months of age, Z-scores to the Agogo reference population increased from  −2.1 ± 0.44 to −0.73 ± 1.48 and Z-scores to WHO standard rose from −3.6 ± 0.55 to −1.89 ± 0.72. At the age of 4–9 years Z-scores to Agogo reference population decreased from −0.75 ± 1.22 to −1.57 ± 0.78; Z-scores to the WHO standard decreased from −1.1 ± 1.4 to −1.61 ± 0.51.

Figure 2 compares heights of LBW children aged 48–108 months with local growth data 4–9 years and the WHO standard. At the ages of 48 and 96 months, LBW children were significantly shorter (Mann–Whitney U-test, exact < 0.05, 2-tailed). Z-scores for height to the Agogo reference population decreased from −1.29 ± 0.69 to −1.94 ± 1.49 from 4 to 9 years; Z-scores to the WHO standard decreased from −1.6 ± 0.87 to −2.24 ± 1.35 for the same age groups.

Table 4 and Figure 3 show the incremental weight gain during the first 3 years among the study population against local growth data 0–36 months and the WHO standard. The incremental weight gain of LBW infants from 0 to 2 months is significantly lower and the incremental weight gain from 2 to 4 months is significantly higher than the local growth data 0–36 months (Mann–Whitney U-test, < 0.05). From 4 months onwards, incremental weight gain was similar between the LBW children and the local reference group. However, if the incremental weight gain per month is expressed in percentages of weights at various measuring points, it decreased from 20% to 2.2% for the age span of 0–2 months to 30–36 months in LBW children and from 22.0% to 1.7%, respectively, in the reference population ( Table 4).

Table 4.  Median incremental weight gain in kg/month and percentage/month in LBW infants and Agogo reference population Thumbnail image of

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

A previous study ( van der Mei 1994) described postnatal survival chances (0–6 days of life) of 567 LBW (≤ 2000 g) infants born in or admitted to Agogo Hospital from 1982 to 1988 and compared the death rates of these children with 15 studies from different parts of the world. The outcome was similar to that in various developed countries in the 1960s and 1970s and better than most recent reports in developing countries. It seemed that LBW infants in the developing world, particularly those with a birthweight > 1500 g, have a reasonable chance of surviving the first weeks of life with only simple conservative care. This study aimed to assess survival and growth of these children during infancy and childhood.

VLBW children are at high risk especially during the first 3 months of life; by comparison MLBW children have a fair chance (odds ratio 4.48; CI = 0.86–29.32) of surviving childhood after discharge from hospital once they are gaining weight on exclusive breastfeeding irrespective of their individual bodyweight. In Africa LBW babies are discharged from hospital with much lower bodyweight than in Europe ( Boersma 1979; Ransome-Kuti 1985), as there is good evidence that they are better at adapting to extra-uterine conditions than Caucasians of the same gestational age ( Olowe & Akinkugbe 1978). Early discharge of these infants makes it easier to establish full breast-feeding, can avoid hospital infection and reduces the clinical workload. In Agogo Hospital, the average bodyweight on discharge of surviving LBW children was 1680 g with an average length of stay in hospital of 11.6 days ( van der Mei 1994). In western countries LBW infants are usually discharged when they have reached the weight of 2000–2500 g.

At the age of 3 years, 14.2% of children were lost to follow-up. This relatively low percentage compares favourably with reports on LBW infants from India ( Kadam et al. 1991 ; 41%, = 109), Nigeria ( Twaithes et al. 1974 ; 26%, = 50) and Brazil ( Victora et al. 1987 ; 7.5%, = 48), all followed for 12 months only.

Remarkably, we lost only 1 of 32 VLBW children to follow-up vs. 14 of 73 MLBW children. Mothers may care more for a ‘weaker’ child, or their involvement in daily hospital care in the neonatal ward may motivate them to come regularly for follow-up. This better compliance of VLBW infants was also reported from India ( Kadam et al. 1991 ).

Survival of those followed until the age of 3 years was 90%; most deaths (five out of nine) occurred in the first 3 months; 4 of them were VLBW infants. There was no significant difference in survival between male and female children. The significant difference in survival between VLBW and MLBW children confirms that VLBW infants are at increased risk, particularly during the first months of life. They should receive special attention during that period with emphasis on nutrition and general health care. Our results concerning LBW survival are similar to those from Brazil ( Victora et al. 1987 ; 87.5%, = 48) and Nigeria ( Twaithes et al. 1974 ; 82%, = 50). Comparison with data from developed countries is not feasible as circumstances and criteria vary considerably.

Growth

Unfortunately we did not know the gestational age of the LBW infants in this study; therefore a potential difference in growth between preterm and small-for-gestational-age children could not be studied. From birth until the age of 24 months the difference in median weight between LBW children and local growth data 0–36 months remained similar (± 1.5 kg, Figure 1). At 30 and 36 months the weight difference to the reference group had diminished to 0.7 and 0.5 kg, respectively ( Figure 1); whereas over time the difference in weight between LBW children and local growth data had increased from 1.5 kg at 4 years to 5 kg at 9 years of age.

LBW children aged 4–9 years were stunted compared to local growth data 4–9 years; difference in height increased from 6 to 11.5 cm during this time span ( Figure 2). The improvement of Z-scores for weight from birth to the age of 36 months to the Agogo reference population and the WHO standard indicate a certain catching-up; however, from 4 to 9 years the Z-scores for height and weight to the reference populations are worsening again, indicating increasing growth impairment. However, these findings should be considered with caution as evaluation of growth of the children at the age of 4–9 years is hampered by relatively small numbers of children (varying from three to 11 at different measuring points).

Remarkably, the incremental weight gain of LBW infants during the first two months was lower and from 2 to 4 months significantly higher than that of the reference groups, whereas the pattern was similar from 4 months onward; it seems that LBW infants have a slower start, catch up a little, but then lag behind during the rest of the first 3 years of life.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

With dedicated health personnel, a follow-up of LBW children in a rural area of Africa is possible at minimal cost and with relatively good compliance. Parents in a rural area of Africa can be motivated to bring their children regularly for follow-up. LBW children have a fair chance of surviving childhood once they reach the age of 3 months. We saw slight catch-up growth during the first 3 years; growth impairment increased during childhood. Possible long-term implications for ultimate weight, height and morbidity at adult age of this prolonged postnatal growth retardation remains to be investigated. LBW infants in rural areas in Africa, in particular VLBW infants, are at high risk and require close supervision, e.g. by frequent home visits, particularly in the first 3 months of life. These visits should serve to stress the importance of breastfeeding, of paying special attention to signs of undernutrition and anaemia and of monitoring growth. The help of Traditional Birth Attendants (TBAs) could be considered for this purpose.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

We thank the Director of Health Services in Ghana for permission to conduct and publish this study. Without the help of several medical students from Groningen and nursing students of the Agogo Nurses Training College, this follow-up study would not have been possible; they often spent hours finding the homes of the children, and their great efforts are much appreciated. In the smaller communities the help of TBAs was invaluable. The assistance of several Agogo Hospital staff members (particularly Dr Blijleven as well as Nursery, Labour Ward and Children's Ward staff) in tracing children is much appreciated. We thank Dr H.W.A. Voorhoeve and Dr H.B.P.E. Gernaat for their valuable advice and helpful contributions to the writing of this paper. We are much obliged to Mr J.H.M. van Raay for his kind assistance with computing.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • Boersma ER (1979) Perinatal circumstances in Dar es Salaam, Tanzania. Studies on some physiological aspects in the tropics. PhD thesis. Erasmus University, Rotterdam.
  • Kadam SB, Daga SR, Daga AS (1991) Quality of survivals on conservative neonatal care. Journal of Tropical Pediatrics 37, 250 253.
  • Mbise RM (1977) Neonatal morbidity and mortality of low birthweight babies in Dar es Salaam, Tanzania: a prospective study. PhD thesis. University of Dar es Salaam, Tanzania.
  • van der Mei J (1994) Survival chances of low birthweight infants in a rural hospital in Ghana. Tropical and Geographical Medicine 46, 313 317.
  • Olowe SA & Akinkugbe B (1978) Amniotic fluid lecithin/ sphingomyelin ratios: comparison between an African and a North American community. Pediatrics 62, 39 41.
  • Ransome-Kuti O (1985) Intra uterine growth, birthweights and maturity of the African newborn. Acta Paediatrica Scandinavica 319, 95 97.
  • Shann F (1993) Nutrition. Nutritional indices: Z, centile or percent? Lancet 341, 526 .
  • Twaithes M, McGucken RB, le Jossec M (1974) Developmental progress of Nigerian infants of low birthweight in the first year of life. Nigerian Journal of Pediatrics 1, 10 13.
  • Verloove-Vanhorick SP & Verwey RA (1987) Project on preterm and small-for-gestational-age infants in the Netherlands 1983. PhD thesis, University of Leiden, The Netherlands.
  • Victora CG, Barros FC, Vaughan JP, Teixeira AMB (1987) Birthweight and infant mortality: a longitudinal study of 5914 Brazilian children. International Journal of Epidemiology 16, 239 245.