1.gif (1892 bytes)

Original Articles

                                                                                                                                                                        Indian Pediatrics 1998; 35:1071-1079


D.K. Agarwal, A. Agarwal, M. Singh, K. Satya, S. Agarwal and K.N. Agarwal*

From the Maternal and Child Health Unit, Department of Pediatrics, Institute of Medical Sciences, Varanasi-221 005, and University College of Medical Sciences, Delhi - 110 095, India.

Reprint requests: Prof K.N. Agarwal, D-115, Sector 36, Noida 201301. (I1.P.), India.

Manuscript received: January 15, 1998; Initial review completed: March 10, 1998;
Revision accepted: July
1, 1998



Objective: To study the relationship between pregnancy wastage and maternal undernutrition and other sociodemographic factors in rural Indian women. Setting: Rural community of Varanasi. Design: Longitudinal observation. Method: In 49 villages during 1988-92, 8111 pregnancies were registered to observe for wastage. The pregnancy outcome was correlated with various factors. Results: There were 1321 abortions and 141 still births, Women's income < Rs. 250 per month, education < 10th class and protein intake < 50 g/day had significantly higher relative risks (RR) (4.1, 2.9 and 2.8, respectively) for abortions. Poor maternal nutrition was an additional important risk factor, for still births (RR 5.1 and 4.2 for maternal weight and height, respectively). A pregnancy interval over 2 years reduced both. Conclusion: Low socioeconomic status, chronic undernutrition and illiteracy in rural India are associated with high pregnancy wastage.

Key words: Abortions, Rural, Still births, Undernutrition.

PREGNANCY outcome is influenced by hereditary and environmental factors including those which affect stature in early life, current health and nutritional status, inter-pregnancy interval, maternal age, genitourinary or general diseases in women and socioeconomic and educational status. There has been considerable reduction in perinatal mortality (late fetal deaths and early neonatal deaths) indicating improvement in perinatal care and general socioeconomic and educational levels in developed countries(1). Whitfield et al.(2) from data collected after sixteen weeks of conception to first week of early neonatal death on 13446 births documented the total perinatal related wastage (TPRW) of 22.8/1000 births. The deaths from 16-28 weeks were 94, still births were 67 and early neonatal deaths were 27; thus 50% losses occurred during 16-28 weeks of pregnancy (late abortions). Hogston and James(3) also showed that out of TPRW, 2/3rd losses were late abortions (16-28 weeks). In India, most of the studies are limited to still birth rates and early neo- natal deaths and the data for pregnancy losses classified as abortions is largely missing. The present study documents the spontaneous abortions and still births and their correlation with maternal nutrition and other sociodemographic factors from a rural area. An attempt was also made to compare the pregnancy wastage in areas with or without Integrated Child Development Services (ICDS).

Subjects and Methods

Profile of Study Area

The study was conducted in two adjoining rural areas of Varanasi district of Uttar Pradesh. Twenty eight villages out of 148 of Harahua (ICDS) and 21 villages out of 172 of Kashi Vidyapeeth (Non-ICDS) blocks were selected using random sampling technique. According to 1990 survey these blocks had birth rates of 37.7 and 37.6; general fertility rates of 179.9 and 186.5; death rates 10.2 and 10.4 and infant mortality rates 102 and 88, respectively. The literacy rates for men were 43.6 and 41.9% and for women 9.9 and 8.5%, respectively. Agriculture was the main occupation. Majority (89.1 %) of women were housewives. The selection of the study women is presented in Table I. The data for perinatal and infant mortality rates were similar in these blocks (Table I). Thus data on 8111 pregnancies ill rural Varanasi for abortions and still births are presented together for both the blocks (Table II; Fig. 1).

Fig. 1 Schematic Diagram Showing the selection of study women from 49 villages

Plan of Registration

Women were registered during January 1988 to December 1992. Early termination of pregnancy included women having loss of pregnancy before 28 weeks of gestation. Ectopic, molar pregnancies and induced abortions were excluded while only spontaneous abortions were included. The still births referred to fetal loss at or after 28 weeks of gestation.

Three outcome measures were evaluated, namely, abortion rates, abortion ratios and still birth rates defined as below:

Follow-up of Target Women

After excluding the non-eligible women, the target women lists were given to the selected trained female village workers (One per village; population around 1000). They regularly monitored the last menstrual period by home visits, women missing two consecutive periods were registered. The information of study women was collected by a team of nutritionist and social scientist having post graduate degree.

Data collection included:
(a) Maternal anthropometry weight, height and mid-arm circumference, measured as per standard techniques using Chattilon weighing scale (USA), anthropometric rod and fibre glass tape, respectively (4); (b) Hemoglobin estimation by cyan-methemoglobin method(S); (c) Dietary intake pattern (24 hours recall); and (d) Sociodemographic characteristics. The registered women were contacted every month and later at 16
2, 28 2 and 36 2 weeks of gestation. Women requiring medical attention were referred to the University Hospital, Institute of Medical Sciences, Varanasi. Due to social customs no attempt was made to collect the abortive or stillbirth material. The anthropometry, hemoglobin, dietary intake, sociodemographic data collected at registration were correlated with pregnancy loss. The number of women measured for anthropometry, dietary survey, hemoglobin was variable as all women could not be covered and some refused, particularly the blood examination. Target women (likely to conceive) were examined as planned; however our team did not provide any supplement, i.e., nutrition, hematinic or education. In the ICDS block also the supplementation as well as other inputs were erratic, although operative since 1983 in these villages. The only difference in the two blocks could be introduction of female village workers (ICDS in 1983-84 and non-ICDS in 1988). During the study period the visits by workers were similar.

For data analyses, the Z proportion test, relative risk (Odd's Ratio or RR) and their 95% confidence interval {CI), attributable risk (AR) and multiple regression analysis were performed.


Table II shows that out of 8111 pregnancies, 1321 ended in spontaneous abortion (16.3%), The incidence of abortion during the four study years ranged between 14.7 to 16.9% (abortion rate/1000 women of reproductive age ranged between 30.4 to
43.5). Out of 6790 births (excluding abortions), 141 terminated into still births; giving a rate of 20.8/1000 births. The abortion and still birth rates did not differ during the study years.

Table III summarizes the relationship between various factors and abortions. Women having weight <40 kg and height <145 cm, had abortion ratio of 290.0 and 234.0 as compared. to those with weight >50.1 kg and height >155 cm having ratios of 144.4 and 169.0 (p<0.001 and p<0.01, respectively). The abortion ratio was 231.8 if mid arm circumference was <18.0 cm and a significant reduction was observed in the group 22.1-24.0 cm, abortion ratio being 130.4 (p <0.01). An increase in hemoglobin showed consistent reduction in abortion ratios, risk being lower in those with hemoglobin level of 8.1 g/dl or more. A calorie intake around 2000 and protein intake of 55.0 g/day were associated with lower abortion ratio.

The still births showed significant relationship with maternal weight and height, values <40.0 kg and <145 cm had rates of 174.6 and 73.5, respectively. The corresponding values for >45 kg and> 155 cm were 7.0 and 2.8 (p <0.001 for both). This relationship was also significant with a dietary intake of protein >55.0 g/day; the stillbirth rate reduced to 11.5 as compared to 78.0 in those receiving <45.0 g/ day (p <0.001). The RR values were 5.1, 4.2 and 3.3 with AR of 80.2, 76.1 and 69.5% in pregnancies with lower weight, height and
protein intake, respectively (Table IV).

                       Total number of spontaneous abortion
Abortion rate = --------------------------------------------------------x 1000
                                Net target women

                         Total number of spontaneous abortion
Abortion ratio = ----------------------------------------------------------x 1000
                             Total number of pregnancies

                         Number of fetal deaths at 28 or more completed weeks gestation
Still birth rate = ------------------------------------------------------------------------------------------------  x 1000
                                                Number of total births


Characteristics of Study Women from Both the Blocks and Pregnancy Outcome Data (1988-92).

Kashi Vidhyapeeth
Number of villages     28            21    49
Population 33770 32307 66077
Households 4494 4582 9076
Eligible women. 6199 5943 12142
Women excluded: family planning acceptors,      
unmarried, post menopausal, divorced widows 1505 1617 3122
Net target women 4694 4326 9020
Total pregnancies 3922 4189 8111
Total birth 3273 3517 6790
Still birth 57 84 141
(Rate) 17.4 23.9  
Abortion 646 675 1321
Ratio 164.7 161.1  
Live births 3216 3433 6649
Perinatal mortality ratea 70.3 71.1  
Infant mortality rateb 101.1 96.1  

a Number of still births and infants death of less than 7 days per year per 1000 births.
b Number of infant deaths during the year per 1000 live births.



Outcome Measures and Population Estimates for 9020 Target Women [1988-1992]

of study

Number of

women of
per year
of births

Number of
still births

Still birth
births per
First 1859 274 30.4 147.4 1585    30 18.9
Second 1897 309 34.3 162.9 1588 34 21.5
Third 2036 346 38.4 169.9 1690 '36 21.3
Fourth 2319 392 43.5 169.0 1927 41 21.2
Total 8111 1321     6790 141 20.8



Relative Risk,
95 per cent Confidence Interval Limits and Attributable Risk of Maternal Nutritional and Other Factors on Abortions.

Risk factor

Cut off

Women with
Women without
Risk (95% CI)

Risk (%)

Height (em) < 147.5 400 1451 1.21 17.5
  > 147.6 771 3392 (1.06-1.38)  
Weight (kg) < 42.5 375 1705 1.21 17.2
  >42.6 785 4309 (1.05-1.38)  
Mid-arm- <22.0 856 3996 1.66 39.7
circumference (em) > 22.1 258 1998 (1.43-1.92)  
Hemoglobin (g/dl) < 8.0 196 900 1.37 26.9
  > 8.1 465 2919 (1.13-1.16)  
Calories < 1700 515 2710 1.12 10.9
(Kcall day) > 1701 327 1931 (0.96-1.30)  
Protein (g/day) <50.0 570 2349 2.83 64.7
  > 50.1 272 3175 (2.42-3.30)  
Parity <1 509 1433 2.51 52.7
  >1 808 5748 (2.38-2.64)  
Per capita <250 1180 2539 4.12 68.1
income (Rs.) >250 141 1251 (3.94-4.31)  
Women Illiterate 1199 6702 2.06 47.5
education Literate 122 1409 (1.88-2.27)  
  <10th class 1296 7686 2.87 61.5
  > 10th class 25 425 (2.46-3.27)  
Caste Lower* 1255 6247 1.65 35.3
  Upper 66 543 (1.39-1.91)  

Scheduled, others and backward castes.

Lower abortion ratio and still birth rate were observed for maternal age groups 26-30 and 31-35 years, respectively (p< 0.001) for abortions between < 20 vs 26-30 years and p < 0.001 for still births between < 20 vs 31-35 years of age). In parity 2-4, the abortion ratio was lower, for primipara values for RR, CI and AR were 2.5; 2.4-2.6 and 52.7%, respectively. The still ,birth rate was .lower in para 2; for primipara RR, CI and AR were 1.4; 1.1-1.8 and 30.2%,respectively. The abortion ratio and still birth rate significantly reduced if interpregnancy interval was over 2 years (p < 0.001 and < 0.005. respectively; Tables III and IV).

Scheduled caste women had significantly higher abortion ratio of 347.1 as compared to 112.4 and 108.4, observed for backward and upper castes (p < 0.001). The still birth rate was 27.8 as compared to 17.0 and 13.1, respectively (p < 0.05). Large family size (9-15 members) showed reduced abortion ratio and still birth rate from 187.7 and 28.7 to 144.8 and 13.9 as compared to small (< 5 member) family (p < 0.001), respectively. The observations on women's education and occupation are difficult to comment as around 90% were illiterate and housewives. Still it appears that education over junior high school had lesser risk. Sedentary/field/paid jobs enhancing their per capita income improved pregnancy out- come (Tables III & IV).



Relative Risk,
per cent Confidence Interval Limits and Attributable Risk of Maternal Nutritional and Other Factors on Still birth. "

  Risk factor
Cut off
Women with
still birth
Women without
still birth
Risk (95% CI)
Risk (%)
Height (cm) < 147.0 80 1416 4.18 76.1
  > 147.1 45 3327 (2.88-6.05)  
Weight (kg) <45.0 110 2871 5.07 80.2
  >45.0 15 1983 (2.94-8.73)  
Mid-arm- <22.0 85 2971 1.52 34.7
circumference (cm) > 22.1 35 1862 (1.02-2.26)  
Hemoglobin (g/ dl) <11.0 109 2530 1.74 42.7
  > 11.0 17 688 (1.03-2.92)  
Calories < 1700 64 2195 1.26 20.9
(Kcal/ day) > 1701 37 1604 (0.83-1.90)  
Protein (g/ day) <55.0 85 2349 3.28 "69.5
  >55.0 16 1450 (1.91-5.61)  
Parity <1 39 1252 1.44 30.1
  >1 102 4725 (1.07-1.82)  
Interpregnancy <2 105 2950 1.48 32.5
interval years 2 36 1499 (1.01-2.18)  
Per capita <250 134 6123 1.79 72.9
income (Rs.) > 250 7 526 (1.02-2.55)  
Women Illiterate 122 5503" 1.5 32.9
education Literate 19 1287 (1.01-1.99)  
  <10th class 137 6390 2.14 52.8
  > 10th class 4 400 (1.15-3.14)  
Caste Lower* 134 6123 1.65 38.8
  Upper 7 526 (0.89-2.49)  

* Lower castes included - scheduled + others + backwards

A RR 2 was observed for abortion ratio in relation to women's per capita income Rs.250/-)-4.1; for primipara-2.5 and in
illiterate women-2.1; the AR being 68.1, 47.5 and 52.7%, respectively. Women having education over 10th standard had lower risks for abortion and still birth. for those under 10th standard education RR values were 2.9 and 2.1; with AR of 61.5 and 52.8%, respectively (Tables III & IV). Husband's skilled employment as well as education (even at primary level) significantly reduced abortion ratio and still birth rates (data not presented).

Harahua block has been receiving inputs from the ICDS since 1984 providing nutrition supplement (irregular/insufficient) in late pregnancy and education for child care. Similar level workers were introduced in Kashi Vidhyapeeth block (non-ICDS) 1988 onwards to monitor pregnancy outcome. The number of visits in both blocks were similar. The still birth rate decreased by 29.0% in the ICDS (16.9) as compared to non-ICDS block (23.9). Comparison between the supplemented and non-supplemented pregnancies in the ICDS block had still birth rates of 10.6 and 20.0 (50% reduction), respectively. The supplemented women showed marginally higher weight gain during pregnancy and increased caloric intake during the 3rd trimester
(Table V). Multiple regression analysis showed that the supplementation in ICDS contributed to birth weight and final gestation age by 0.9 and 1.9% variance, only respectively. The abortion rate in both the blocks was similar.



ICDS and Pregnancy Outcome.

(After mid-pregnancy)
Births 1038 1864 2902 2063
Live births 1027 1826 2853  
Mortality rates  
Stillbirth rate 10.6 20.8 16.9 23.9
Perinatal mortality rate 46.2 81.4 70.3 71.1
Early neonatal mortality 36.0 62.1 53.7 48.4
Infant mortality 76.0 112.8 101.1 96.1
Maternal (Mean SE) 
Weight gain during 6.37 6.27   6.24
pregnancy (Kg) 0.03 0.03   0.03
Weight pre-pregnancy (Kg) 42.5 42.5   41.5
  0.12 0.09   0.00
Height (em) 149.0 149.4   149.8
  0.13 0.09   0.11
Mid-arm circumference 22.0 22.1   21.9
(Pre-pregnancy) (em) 0.05 0.04   0.04
Third trimester (Means + SE) 22.5 22.6   22.7
Mid arm circumference (em) 0.05 0.04   0.04
Hemoglobin value (g/ dl) 9.3 9.3   10.0
  0.05 0.04   0.04
Caloric intake 1854 1789   1746
(kcal/ d) 7.7 6.3   4.9
Protein (g/ day) 55.5 54.7   55.1
  0.19 0.16   0.16


A report(6) on perinatal mortality during 1977-79 in rural India (Pune) population 40,000, showed still birth rate of 25.6, perinatal mortality rate of 46.7, with early neonatal, death rate of 21.7, and low birth rate of 19.9%. The still birth rate in rural Pune was considerably higher as compared to rates for Burma (17.4), Indonesia (13.7) and Thailand (8.9). Rural Indian data from Palaghar(7), Vellore(8), Udaipur(9) and Ambala(10) showed still birth rates of 24.6, 27.4, 24.2 and 26.8, respectively. These values correspond to the still birth rate of 18.9 to 21.5 observed in the present rural study from Varanasi, during 1988 to 1992. An abortion ratio of 162.8 in the present rural study is also comparable to other reports, inspite of the fact that these rural Indian women were undernourished with poor
socioeconomic status, do physical labour and have limited access to health services. The abortion ratio has continued to remain high in developed countries-USA-North Carolina-116.1(1l), Utah 166(12) and in UK 12%(13) to 20%(14) by 20 weeks of pregnancy. The reported abortion ratios from Canada(15) and Australia(16) were 128.1 and 182.0, respectively. There are few studies on spontaneous abortion in India which documented values of 40-70(17) and 41.5(18) per 1000 women. The ICMR multicenter study data from Chandigarh, Delhi, Gwalior, Jaipur, Lucknow, Pune and Varanasj(19) on 4261 rural and 3052 urban slums women (total 7313 pregnancies) documented abortions in 5.4%, still births in 2.6% and perinatal mortality of 57.7/ 1000 live births (rural-61.8 and urban slums-52.0). The risk factors for increased perinatal mortality (stillbirths and neonatal deaths in first week) were illiteracy, birth interval < 24 months, previous still birth, previous preterm, untrained birth attendant and birth weight < 2000 g. Surprisingly, inspite of regular LMP monitoring, the abortions were 5.4% only. In contrast, the present rural study at Varanasi had abortion ratios of 147.4, 162.9, 169.9 and 169.0 for the respective years from 1988 to 1992. However, the still birth rate was closer being 18.9, 21.5, 21.3 and 21.2, respectively. In the ICMR multicentric hospital based study, of the total cases registered, 9.5% were abortions(20). The Registrar General of India for 1994 has reported still birth rates of 3.8 and 2.9 for rural UP and Bihar, respectively (rural India having rate of 7.3). In contrast the best state with lowest perinatal and infant mortality, Kerala has a still birth rate of 9.0. These figures indicate the degree of unreliability in collection of data(21).

As observed above maternal nutrition, education, per capital income and protein intake are interrelated factors. The 'Women with lower per capita income and consuming lower quantity of protein in diet had. about 65-73% attributable risk for higher rates of abortion as well as still births in the community. A comparable level of attributable risk (76-80%) was also observed for still births in short stature and under weight women. Illiterate women had higher attributable risk both for abortions as well as still births. The relationship of husband's occupation and education possibly brings in the importance he attaches to family's care. All these factors in turn may influence inter-pregnancy interval.

The Varanasi rural study results on the ICDS supplemented women achieving reduction in still birth rate by 50%, stresses need for food availability. Long-term benefits have been observed in WIC-USA studies(22) showing reduction in fetal loss by 5.7 /1000 births. In addition, the control of nutritional anemia an observed risk factor in pregnancy wastage will also improve the pregnancy outcome(23).


The authors are indebted to the financial and infrastructure supports provided by the Banaras Hindu University, Varanasi and the Indian Council Medical Research and US-AID, New Delhi, India.



1. Karlberg P. Ericcson A. Perinatal mortality in Sweden (Analysis with international aspects). Acta Paediatr Scand 1979; 275: (Suppl) 28-34.

2. Whitfield CR, Smith NC, Cockburn F, Gibson AAM. Total perinatally related wastage: A proposed classification of primary obstetric factors. Brit J Obstet Gynaecol1986; 93: 694-703.

3. Hogston P, James DK. Total perinatal wastage. A clarification for priorities. Brit J Obstet Gynaecol1990; 97: 999-1002.

4. Agarwal DK, Agarwal KN, Upadhyay SK, Mittal R, Prakash R, Rai S. Physical and sexual growth pattern of affluent Indian children from 5 to 18 years of age, Indian Pediatr 1992; 29: 1203-1282.

5. Crosby WH, MunnJG, Furth FW. Cyanmethemoglobin method for estimation of hemoglobin. US Armed forces Med J 1954; 5: 693-697.
6. Perera T, Lwin KM. Perinatal mortality and morbidity including low birth weight (A South-east Asia Region Profile). World Health Organization, SEARO Regional Health Paper No.3, 1984.

7. Shah PM, Udani PM. Analysis of the vital statistics from the rural community, Palghar-II Perinatal, neonatal and infant mortalities. Indian Pediatr 1969; 6: 651- 668.

8. Rao PSS, Inbaraj SG. Extent of perinatal loss of south Indian urban and rural populations. Indian Pediatr 1975; 12: 231- 237.

9. Damodhar MD, Mathur HN, Sharma PN. Some observations on perinatal mortality in rural health centre. Indian J Pediatr 1983; 50: 629-633.

10. Kumar R, Singh S. Risk factors for still births in a rural community. Indian J Pediatr 1992; 59: 455-461.

11. Wilcox AJ, Weinberg CR, O'Connor JF, Baird DD, Schlallerer JP, Carifield RE. Incidence of early pregnancy loss. New Engl J Med 1988; 319: 189-194.

12. Scott JR. Recurrent miscarriage overview and recommendations. Clinical Obstet Gynaecol1994; 37: 768-773.

13. Lesley R, Brande R, Trembath PL. Influence of past reproductive performance on risk of spontaneous abortion. Brit Med J 1989;299:541-545.

14. Cunningham FG. Macdonald PC, Gant NF, Leveno KJ, Gilltrap LC. Abortion. In: Williams Obstetrics, 19th edn. Prentice Hall International Inc., USA, 1993; pp 665-666.

15. Risch Ha, Weiss NS, Clarke EA, Miller AB. Risk factors for spontaneous abortion and its recurrence. Amer J Epidemiol 1988; 128: 420-430.

16. Keeping JD, Najman JM, Morrison J, Western JS, Anderson MJ, Williams GM. A prospective longitudinal study of social, psychological and obstetric factors in pregnancy: Response rates and demo- graphic characteristics of the 8556 respondents. Brit J Obstet Gynaecol. 1989; 96: 289-297.

17. Menon MKK, Devi PK, Bhasker Rao K. Abortion. In: Postgraduate Obstetrics and Gynecology. Madras, Orient Longman, 1993; pp 89c105.

18. Mavlankar DV, Trivedi CR, Gray RH. Levels and risk factors for perinatal mortality in Ahmedabad, India. Bull WHO 1991; 69: 433-442.

19. A National Collaborative study of Identification. of High Risk Families Mothers and. Outcome of their Off springs with Particular Reference to the Problem of Maternal, Nutrition, Low Birth Weight, Perinatal and Infant morbidity and Mortality in Rural and Urban Slum Communities: A Task Force Study. New Delhi, Indian Council of Medical Research, 1990; pp 45-82.

20. Collaborative study on High Risk Pregnancies and Maternal Mortality (Institution based). New Delhi, Indian Council Medical Research, 1990; p 46.

21. Sample Registration System. Fertility and Mortality Indicators, 1994. New Delhi, Registrar General of India, 1996; p 72.

22. Rush D, Sloan NL, Leighton J, Alvir JM, Horvitz DG, Seaver B, et al. Longitudinal study of pregnant women. Am J Clin Nutr 1988; 48 (Suppl): 439-483.

23. Agarwal KN, Agarwal DK, Mishra KP. Impact of anemia prophylaxis in pregnancy on maternal hemoglobin, serum ferritin and birth weight. Indian J Med Res 1991; 94: (B) 277-280.



Past Issue

About IP

About IAP



 Author Info.