BIRTH WEIGHT, STILLBIRTH AND MATERNAL COMPLICATIONS IN SWAZILAND AS RELATED TO PARITY AND MOTHERS AGE

Department of Home Economics, University of Swaziland
Luyengo Campus, M205 Swaziland

ABSTRACT

A situation analysis of the condition of mother and child at birth in Swaziland was conducted in order to determine whether there is a problem which requires an intervention for improving the quality of life. The analysis of mothers age, birth weight and prevalence of stillbirths and maternal complications in Swaziland covered a five year period (1990 - 1994). Data was collected by reviewing birth registers in the four regional hospitals of Swaziland situated in Hhohho, Manzini, Lubombo and Shiselweni regions. The data was used to establish the relationship between parity and mothers age with stillbirth, birth weight and maternal complications. The results show that the rate of stillbirth is about 20 or more for every 1000 births. The stillbirth rate is 2 - 3% with high frequencies in teenage and advanced age groups. The proportion of underweight babies in Swaziland is high (rated at 10%). The average rate of maternal complications is 13% with the most common types being prolonged labour, premature delivery and caesarian section. Birth weight was found to increase with age and parity. There was no relationship between age and maternal complications. Birth weight and stillbirth indicated an increase with parity. There was no relationship between parity and maternal complication. The observed high values of stillbirth and under weight calls for the need to investigate the causes so as to identify interventions with the view to improve the quality of life of the mother and child.

INTRODUCTION

In Africa about half a million women die from maternal causes every year leaving behind at least a million motherless children (WHO, 1987). Complications of pregnancy also result in illness. The International Safe Motherhood conference revealed that 1400 women die everyday in the process of carrying or delivering their children (Barros et al., 1987). In Africa women produce as much as 80% of the food supply (WHO, 1987). This implies that maternal deaths which could be avoided waste not only precious lives but human resources too.

Over the last 20 years rates of maternal mortality and anaemia among pregnant women in the developing world have not improved. Over half of the pregnant women in Africa are anaemic (Maternal Health, 1994). This condition affects the health of the baby since it depends on the health of the mother during pregnancy and delivery. For example, the percentage of low-birth-weight babies went down in developed countries between 1979 and 1990 but it remained virtually the same in Africa (Maternal Health News, 1994). In Swaziland, the still birth rate for the four year period 1989 to 1992 was 20 per 1,000 births at one regional hospital (Shongwe, 1993). However, the information addressing maternal problems on a country-wide coverage is lacking. Swazi women and children suffer disproportionately in terms of poverty, access to social services, health and welfare (UNICEF, 1994). The need for continued attention to the welfare of health is in line with Swaziland's national priorities for development as recently articulated in the general policy statement presented during the opening of parliament in 1994. An early assessment of the situation has two advantages. First, the status, extent and possible causes of the problems could become apparent thereby forming a basis for intervention. Second, the risk of losing the producer i.e. the mother and the baby could be minimized.

Therefore, in view of the above needs this study was conducted in order to determine the extent of the problems as a basis for interventions that may be necessary. The objectives were :

Population and data collection

The target population of this study included all pregnant women who went to deliver in the four major hospitals of Swaziland, i.e. Mbabane Government Hospital, Raleigh Fitkin Memorial (RFM) Hospital, Hlatikhulu Government Hospital and Good Shepherd (GS) Hospital from the year 1990 to 1994. These hospitals are located, respectively, in the Hhohho, Manzini, Shiselweni and Lubombo regions. The birth registers in the maternity wards were reviewed to obtain the required information i.e. age, parity, home district, birth weight of baby, stillbirths and maternal complications. The population size of the investigated mothers was 71521.

Data Analysis

Data collected was summarised for appropriate statistical analysis. Means, range, regression, chi-Square and frequency analysis were used to describe the situation and relationships between each of the studied factors, age of mother, parity and the outcome of pregnancy.

Mothers age and birth weight

Results showing means of ages of the mothers, birth weight and the percentage of underweight babies in the four regional hospitals are presented in Table 1. The age range for mothers in the four hospitals was between 10 and 52 years. The age means for the mothers among the four hospitals varied between 23 and 25 years. The population of Swaziland is young, therefore, it is most likely that some of these young people are mothers yet themselves are dependant on their parents. According to UNICEF (1994) as quoted by Swazi News (1995), 27% of all deliveries are constituted by school children. These teenagers are most likely to give birth to babies of low birth weight. The mean birth weight for the four hospitals was close to each other and ranged from 3.05 kg to 3.11 kg. The means were above the 2.5 kg weight used as a cut-off point for determining underweight in babies (WHO, 1989) . However, the proportions of underweight babies was high with about 10% in three hospitals. The high proportion of underweight in Good Shepherd may be attributed to the type of data collected which was based on only maternal complication cases only. Generally maternal complication cases which also include premature births are expected to result in higher proportion of underweight (WHO, 1986).

The proportion of underweight babies is high for a country such as Swaziland with a high GNP and per capita income of USD-850 (UNICEF, 1994). This is of great concern since the birth weight of an infant is the single most important determinant of newborn survival, future growth and development. In Swaziland low birth weight contributes largely to neonatal deaths (WHO, 1994), a cause that could be prevented through better management during pregnancy. It is believed that this is due to inadequate nutritional supply and lack of prenatal and perinatal care since approximately half of the population is below the poverty line. Food insecurity affects nearly 40% of the total Swazi population (UNICEF, 1994).

Table 1. The means of ages of the mothers, birth weight and the percentages of underweight babies in the four regional hospitals

Stillbirth in the regional hospitals

Stillbirth is defined as an infant born after 28th week of gestation who did not breath after birth or show any sign of life. Results in Table 2 present proportions and rate of stillbirth in the hospitals. The stillbirth rate at RFM was as high as 31 per 1000 livebirths, followed by the other hospitals which both had 20 per 1000 livebirths. The figures were very high compared to developed countries such as England where the rate of stillbirth was 10 per 1000 births in 1978 (Social Services Committee, 1980). However, there was a slight improvement based on Mbabane Government hospital data compared to the 25 per 1000 births reported for the period of 1989-1991 by Ahmed, (unpublished data) as quoted by Shongwe (1993). The stillbirth rate recorded at RFM was high compared to other hospitals possibly due to its central location thus receiving more referrals from all over the country than the rest of the hospitals.

Table 2: Proportions and rate of stillbirth by hospital

The industries that surround the cities attract women of child bearing age and yet are responsible, to a large extent, for pollution which is believed to contribute to maternal complications and stillbirths. Generally, for more than half of the stillbirths, the mothers were admitted with the baby dead in the womb. This may be due to the fact that approximately 77% of the population lives far away in the rural areas (UNICEF, 1994). Few community based clinics within walking distances have maternity units capable of supervising normal deliveries, therefore, cannot provide obstetric first aid.

The Ministry of Health, Swaziland (1993) revealed that stillbirths are common in Swaziland and account for sixty percent of perinatal deaths. The causes have not been established since there is not yet any research done on this subject. The review of literature together with studies done elsewhere have documented a number of pregnancy complications that can eventually lead to stillbirth. According to WHO (1994) and Tahzib (1983), haemorrhage, hypertensive disorders of pregnancy, sexually transmitted diseases, prolonged labour and infection during pregnancy can result to stillbirth. Parents who smoke also increase the risk of prematurity and stillbirth to the fetus (Maternal Health, 1994).

Maternal complications

Data in Table 3 present the percentage frequency of the commonly occurring maternal complications reported in the four regional hospitals. The most prevalent complications for all the hospitals were prolonged labour, caesarian section and premature births. Prolonged labour is believed to be due to Cephalo-Pelvic Disproportion (CPD). Literature show an association of CPD with teenage girls and women who have suffered from chronic severe malnutrition from childhood (Shah, 1989). According to UNICEF (1994) 27% of all deliveries in Swaziland are constituted by school children in which case prolonged and obstructed labour are serious complications. If these complications do not receive timely management they may result into the death of the women, rupture of the uterus or infection (WHO, 1994). They can even result into severe disabilities such as obstetric fistulae. The consequences of such damage are urinary incontinence and faecal incontinence if the rectum is affected and the victim will also have given birth to a still born baby (WHO, 1994; WHO, 1991).

The high rate of premature births may be attributed to sexually transmitted diseases, rupture of uterine membranes and hypertension disorders which include toxaemia, pre-eclampsia or eclampsia (Milaat and Du Florey, 1992; WHO, 1989). The high rate of caesarian cases may be due to it being the technique employed to solve some of the maternal complications such as CPD.

Table 3: Percentage frequency of maternal complications

Relation of age with still birth

Table 4 shows the percentage of stillbirths that occurred among deliveries in increasing age groups reported in the four hospitals. The percentage stillbirth rate ranged from 1.1% to 4.9% for three hospitals excluding Good Shepherd. The RFM results indicated an increase in stillbirth with age yet Hlathikulu figures were high amongst teenagers and much higher with mothers above 35 years. The latter case conforms to the literature which explicitly indicated a high rate for mothers aged less than 20 years and over 35 years and a low rate for mothers whose age is in between (Milaat and Du Florey, 1992). Mbabane Government hospital recorded high stillbirth rate amongst the middle ages and low with the extreme ages. Good Shepherd hospital results on stillbirth were low in the middle ages compared to the extreme age groups. It should be noted that Good Shepherd hospital results

show high values as compared to other hospitals and that may be due to the fact that the results are based on data for mothers who had complications. Mbabane Government hospital results are not consistent with those for the three hospitals.

Relation of age with birth weight.

Regression analysis was performed to relate the data on age and birth weight in the four regional hospitals. The analysis for each hospital gave a regression equation of the general form:

Birth weight = A + (B x age)

where A is a regression constant
                B is a regression coefficient for age.

The corresponding coefficients of determination (R²) and probability, p, were also calculated. The values of the regression constant, regression coefficient, coefficient of determination and probability for various hospitals are given in Table 5. The values of the regression constant A, indicate the birth weight value which a baby would weigh if the mother had zero age. These values although real and non-zero have no independent interpretation since if the age of the mother is zero, there is no mother, hence no baby weight to talk about.

Table 5: Values of regression constant (A), regression coefficient (B), coefficient of determination (R²) and probability (_P) for the effect of age on birth weight.

The values of the regression coefficient B are all positive and non-zero indicating that for all hospitals the weight of a baby at birth is expected to increase by the values of the regression coefficient for every increase in the age of the mother. These results conform to literature which revealed that there is a higher percentage of infants born to mothers less than 20 years old weighing less than 2.5 kg than infants born to older women (WHO, 1989).

The values of probability, (p) are all about zero indicating that the effect of age on birth weight is very significant. However, the contribution of age to birth weight is very small as depicted by the small values of the coefficient of determination R². Therefore, it means that other factors are also responsible for the variation in birth weight. Factors affecting birth weight could among others be smoking, nutrition, parity, stress, and other social-economic environment (Maternal Health, 1994).

Relationship between age and maternal complications

A statistical analysis using chi-square, X², test was employed to determine the relationship between age and maternal complications. Results showing the calculated chi-square values of the age and maternal complications at 5% level are shown in Table 6. Based on the results from Mbabane, RFM and Hlatikulu hospitals there is no relationship between age and maternal complications. This suggests that maternal complications can occur at any age group. However, the results for Good Shepherd hospital indicate a significant difference. The complications tend to increase in the middle age group, however, the deviation in the teenagers and older age was very small. This deviation from observations in other hospitals may be due possibly to the fact that only complications cases were collected and a few cases (4%) were considered normal deliveries at GS hospital. This included those born before arrival (BBA) at hospital. The normal cases here are what had originally been recorded as complications but were later considered to be normal. Therefore the very small value may have resulted in a biased chi-square.

Table 6: The calculated chi-square values, X² of age and maternal complications parity and stillbirth, and parity and maternal complications for the mothers in the four hospitals at 5% level.

The results based on the three hospitals which show no significant relationship between mothers age and maternal complications are inconsistent with the literature possibly due to the fact that the present study is a retrospective analysis of hospital records; it is possible that some records may not have been detailed enough, leading to underestimation of maternal complications. Literature indicate an increased risk during adolescence and older age groups (Ministry of Health Swaziland, 1993).

Relationship between parity and stillbirth.

Data in Table 6 presents the calculated chi-square, X², values for parity of mothers and stillbirth. The results indicate dependency in three hospitals namely RFM, Hlathikulu and Good Shepherd since the calculated X² values are greater than the critical value of X² = 5.99. Hence, parity has an effect on stillbirth. However, Mbabane Government Hospital results do not show any relationship between stillbirth and parity. Stillbirth rate for the three hospitals was lowest among babies of mothers with parity one and two and highest for babies whose mothers had three or more previous births, thus conforming to literature. According to Barros et al. (1987) fetal mortality rate is lowest among babies of parity 2 and highest for babies whose mothers had 3 or more previous births. However, Milaat and Du Florey (1992) revealed that women of parity 1 to 7 have lower odds of perinatal death than women in their first pregnancy.

Relationship between parity and maternal complications.

Results showing the calculated chi-square of the interdependence between parity and maternal complications are included in Table 6. The chi-square values show that maternal complications have no direct relationship to parity in the three hospitals. These observations are inconsistent with literature. According to the Ministry of Health, Swaziland (1993) certain pregnancies are more vulnerable to complications such as the first pregnancy and those at the extremes of the reproductive age group. According to Shongwe (1983) and IMPACT (1986) a large percentage of women who die from haemorrhage are those with parity 5 and above as compared to those with lower parity. However, complications of Toxaemia are more likely for women having their first child.

Good Shepherd hospital results indicated a relationship between parity and maternal complications. However, this unique observation may be due to the fact that only the parity of the maternal complication cases was used due to the unavailability of the parity of the normals.

Effect of parity on birth weight

Regression analysis between parity and birth weight was performed for the four regional hospitals. The values of the constant A and coefficient B and their corresponding coefficient of determination (R²) and probability values are presented in Table 7. The results indicate that the relationship between parity and birth weight for the four hospitals was significant (p = 0.001). The regression constant reveal that a mother giving birth for the first time would be expected to get a baby of about 3 kg which is the average weight for babies born in the four hospitals.

Table 7: Values of regression constant (A), regression coefficient (B), coefficient of determination (R²) and probability (p) for the effect of parity on birth weight.

The regression coefficient was positive for all the hospitals indicating that the weight of each consecutive baby is expected to increase. The coefficient of determination was quite low indicating that other variables not investigated in this study are also responsible for variations in birth weight. The variables may include unfavourable socio-economic status, inadequate nutritional supply, lack of prenatal care and age (WHO, 1989).

Nutrition security and outcome of pregnancy

Malnutrition begins early for most African infants: in the woumb. The World Bank report on World Development quotes percentage low birth weights ranging from 8 to 16 out of all live births for Namibia, Zimbabwe, Zambia, Lesotho, Kenya, Malawi, Uganda and Tanzania. Low birth weight is a major cause of early malnutrition, that may begin in utero because of poor maternal nutrition. According to Siandazwi and Hansch (1993), it is also the leading risk of neonatal and infant death. Despite significant improvements in other areas of infant health such as measles immunization, the problem of low birth weight has not improved significantly. Supplementation of food and micronutrients (particularly iron) given to the mother is known to increase birth weights and, thereby to reduce the risk of child malnutrition thereafter. To achieve further reductions in infant mortality, attention needs to be shifted to this problem. Of the 2.5 million infant deaths that occur in Africa each year, nearly half take place during the first 14 days, and are due to the same complications that jeopardize the mother's life - management of labour and delivery, and the mother's general health and nutritional status prior to pregnancy (JSI, 1993).

According to UNICEF (1994) approximately half of the population in Swaziland is below the poverty line. As a result food insecurity both chronic and transitory affect nearly 40% of the total population. In rural areas women are mostly affected since culture does not allow them to eat certain protein foods such as meat, milk and eggs yet these foods are essential to pregnant women. Inadequate sanitation and hygiene and poor access to protected water continues to be a problem in rural areas. Only 45% of the rural population and 80% of the peri-urban dwellers have access to safe water; 40% of the rural population have no access to sanitary means of excreta disposal (UNICEF, 1994). The combined result may contribute to maternal complications, stillbirth and low birth weight due to inadequate nutrition security. Observations made by Sibandze (1994) indicate that this problem is mostly found in Lubombo and Shiselweni districts. However, the results from this study do not show a clear relationship between regional hospital location and the outcome of pregnancy. This may be due to the fact that mothers who gave birth at a particular hospital may not necessarily be from the region where the hospital is situated. Therefore, a study tracing the mothers' place of residence is proposed. The proposed study would assist to investigate the nutritional causes of maternal problems.

CONCLUSIONS AND RECOMMENDATIONS

The underweight babies for hospital born babies in Swaziland is high (about 10%) while the stillbirth in both teenage and advanced age groups is highly rated at 2-3%. The average maternal complications is 13% with the most common types being prolonged labour, premature birth and caesarian section.

Birth weight increase with both parity and mother's age. However, there is a varied relationship between stillbirth and mothers age. There is no direct relationship between maternal complications and parity. In view of the above results, there is a need for further research to investigate the causes of high rate of stillbirths in Swaziland and the reasons for the high percentage of underweight births. In order to obtain accurate information there is a need to ensure accurate full record keeping in hospitals.

ACKNOWLEDGEMENTS

Our sincere gratitudes are due to Mrs. S. Kabaija, Dr. S.V. Nkambule and Mr. W.O. Mukabwe for their assistance with statistical analysis and computer work. The Ministry of Health and the staff of the four regional hospitals are thanked for their cooperation.

REFERENCES.

Barros F.C., Victoria C.G., Vaughan J.P. and Estanislau H.J. (1987). Perinatal Mortality in Southern Brazil. Journal of World Health Organization. 65 (1): 95-104.

IMPACT (1996). Family planning saves lives (Nov. 1996)

JSI (1993). The Mother Care Project, Project Abstract.

Maternal Health. (1994). Global child health and review. 2(1) : 29 - 31.

Milaat W. A. and Du Florey C. V. (1992). Perinatal Mortality in Jedda, Saudi Arabia. Journal of Epidemiology. 21:82-90

Ministry of Health, Swaziland, WHO, UNFPA, UNDP, UNICEF and WFP (1993). Safe Motherhood Initiative in Swaziland.

Shah, K. P. (1989). Enquiry on the epidemiology and surgical repair of obstetric related fistula in South East Asia. WHO, Geneva.

Shongwe S. (1993). Maternal mortality at Raleigh Fitkin Memorial Hospital, Manzini, Swaziland. M.Sc. Thesis: London School of Hygiene and Tropical Medicine, University of London.

Siandazwi, C. and Hansch, S. (1993). The food and nutrition situation in East, Central and Southern Africa: Collaborative efforts in ECSA Report. The Commonwealth Regional Health Community Secretariat for ECSA.

Sibandze, B. (1994). Personal communication. Chairman of National Disaster Task Force, P.O. Box 432, Mbabane, Swaziland.

Social Services Committee of the House of Commons (1980). Perinatal and neonatal mortality. House of Commons second report from the Social Services Committee session 1979-1980: Volume 1, United Kingdom.

Swazi News. (1995). Jan.

Tahzib, F. (1983). Epidemiological determinants of vesicovaginal fistulas. British Journal of obstetrics and gynaecology. 90 : 387 - 391.

UNICEF (1994). Children and women in Swaziland : Situation Analysis, April 1994.

WHO. (1986). The Growth Chart.

WHO. (1987). Safe Motherhood.

WHO. (1989). The prevention and management of postpartum haemorrhage. Report No.7

WHO. (1991). Report of WHO on perinatal infections. Report No. 10

WHO (1994). Mother-baby package.