Adverse Reproductive Outcomes Associated with Teenage Pregnancy in three Maternity Hospitals in Asmara, Jan 01 – DEC 31, 2018

Background: The highest rate of teenage pregnancy in the world is in sub Saharan Africa. Thus, out of the twenty countries with the highest rate of teen marriage, Eritrea ranked fourteenth. This aim of this study was to assess adverse outcomes associated with teenage pregnancy. Methods: A retrospective cohort study was followed for 424 nulliparous pregnant women between the age of 14 and 24 years. The maternity clinical record was used to review data on maternal spontaneous abortions, preterm birth, method of delivery, postpartum hemorrhage, perineal laceration, baby weight, Apgar score at 5 minute and congenital anomalies were taken as our variables of interest. Moreover, face-to-face interview was conducted to enlighten the socio demographic characteristics, circumcision status of the mother and antenatal care visits during their last pregnancy. Descriptive and inferential data was analyzed using SPSS version 20. Chi-square test and bivariate logistic regression were also used. Results: The socio demographic characteristic of the study shows that the educational status of teenage mother was low compared to adult mothers; in addition unmarried teenage mothers were high compared to adult mothers. In this study teenage mothers had signicantly higher anemia, preterm delivery, Post partum hemorrhage and low birth weight babies as compared to the adult mothers. Also the study revealed that compared to adult mothers, teenage mothers had signicantly higher abnormal delivery mode with low Apgar score among teenage mothers. Conclusions: In this study adolescent pregnancy was found to be associated with increased outcomes of anemia, preterm birth, low birth weight, abnormal delivery and post-partum hemorrhage. Even though preeclampsia and Apgar score were not signicantly different between adolescents and adult mothers this should not be ignored and just bypass it not as adverse outcomes of teenage.

operative intervention rates, obstructed labor and obstetric stula are other additional outcomes faced by teenage mothers. The rates of preterm birth, low birth weight and asphyxia are higher among the children of adolescents, all of which increase the chance of death and future health problems of the baby (3) Methods Study Design A retrospective cohort study was carried among 424 nulliparous pregnant women between the age of 14 and 24 years, inclusively with a live singleton birth during January -December 2018 in Asmara, Eritrea.

Study Setting and Population
The study was conducted in three maternity hospitals of Asmara, Eritrea. The two hospitals namely Edaga-Hamus Maternity hospital and Betmekae Community Hospital are public hospitals whereas; Sembel Polyclinic Hospital is a private institution. All (165) mothers below the age of 20 delivered in these hospitals between January 2018 and December 2018, excluding women with chronic illness and mental problems, were selected as mothers exposed to teenage pregnancy. For each exposed mother two non-exposed mothers aged 20-24 were selected and matched for. Therefore, the initial sample size was 495 mothers. For these 495 mothers secondary data was collected from the maternity hospital delivery cards. In order to gather the primary data we had to contact each mother using the telephone number registered in the hospital. However the telephone of 46 mothers was not working or registered as a wrong number and 25 mothers were not in this country at the time of the study. This makes the nally sample size for the study to be 424.

Data collection
The maternity clinical record was reviewed and data on maternal hemoglobin and blood pressure level, spontaneous abortions, preterm birth, method of delivery, postpartum hemorrhage, perineal laceration, baby weight, Apgar score at 5 minute and congenital anomalies were taken as variables of interest.
Interview: Each mother was invited, using telephone, to come to the National Union of Eritrean Women (N.U.E.W) o ce. Face-to-face interview was conducted to collect information on socio demographic characteristics, circumcision status of the mother and antenatal care visits during their last pregnancy was gathered. Based on the nal sample size of 424 mothers the power of the study was calculated using this formula. Z 1-β= value of the standard normal distribution corresponding to the power of the study.
Overweight baby 0(0%) 1(100%) -* CPD: cephalo-pelvic disproportion; ERM: early rupture of membrane a No explanation was found -Cannot be calculated because of zero frequency in a cell.
As described in Adverse birth outcomes

Discussion
It is often argued that the adverse reproductive outcome in teenage pregnancy is due to the social, economic and behavioral factors rather than the biological effect of young age. In this study statistical analysis was done by controlling economic status, family support, marital status, circumcision, and educational level of all mothers and pregnancy outcomes were computed.
The study showed that there was signi cant difference on educational level were teenagers (97%) and adults (100%) were high school (p=0.015). This nding was similar with studies in Baghdad and Nigeria were they found statistically lower educational level among teenagers with (p=0.0001) and (p=0.001) respectively (7), (8) . In this study unmarried teenagers were higher (6.25%) compared with adults (2.5%) though insigni cant (p=0.055).This was a similar result with study done in India were marital status was found to be insigni cant among the two groups (p=1) (9) . In contrast to studies done in Nigeria and Cameroon were they found that teenage mothers were signi cantly unmarried (p=0.000) and ( p 0.000) respectively (8), (2) .The main reason for less unmarried mothers attending hospital to deliver in our nding could be due to common practice for unmarried mothers to go either for termination of pregnancy or home delivery because of strong social taboo preventing them to attend large public hospitals .In our study we found out no statistically signi cant difference on antenatal visit among mothers , all mothers have attended antenatal care at list four times. This is higher than a study done in Cameroon were 71.92% of teenagers and 72.38% of adults have gone to antenatal care least 4 times (2) .Our nding was in contrast to a study done in Baghdad where they found that 53.5% of teenagers vs.73.9% of adults had attended antenatal care visits (p=0.0001) (7) .The reason for this high attendance of mothers for antenatal care in our study could be because our study was conducted in urban setting were all the mothers could have access to health care facility and the regular health education given to community through different channels. This study did not come across signi cant difference on circumcision status, economic status and family support.
In this study teenage mothers were having signi cantly higher anemia (24.3%) as compared to the adult mothers (11.1%) (OR: 2.568; 95% CI: 1.505-4.382). This nding was comparable to a study conducted in Baghdad teaching hospital were anemia was found signi cantly higher among teenagers (42.5%) compared to (21.3%) adults (OR: 6.13; 95%CI: 3.48-10.9) and in Nigeria were anemia was found among teenagers (22.9%) compared to adults (6.1%) (p=0.000) (7), (8) .Our study was different to studies conducted in Banaras Hindu University were anemia was insigni cantly different among teenagers (62.9%) compared to adults (63.9%) (p>0.05) and in Shadan Institute Of Medical Sciences India were (5%) teenagers and (6.7%) adults were anemic (p= 0.697 (10), (9) . The reason for this higher anemia among teenage mothers compared to adults in our study could be due to the effect of the girl becoming pregnant before her own growth and competing with developing fetus for nutrients (10) .
Post partum hemorrhage was found to be signi cantly higher among teenagers (24.3%) mean blood loss of 361.458(±SD116.4328) compared to adults (11.1%) (OR: 0.135; 95% CI: 0.072-0.253).Our nding was consistent with a study done Niger Delta University, were Post partum hemorrhage was found signi cantly higher 34.9% among teenagers and 12.8% among adults ( P = 0.000) (8) . In contrast to our study, studies done in Nepal, Baghdad, India and Cameron found no statistically signi cant difference on Postpartum hemorrhage among the two groups (11), (7), (9), (2) . The result of this study was also different from the study carried out in University Hospital of Caen, France were younger mothers aged less than 19 years had signi cantly decreased risks of post-partum hemorrhage compared to control mothers ( p<0.05 ) (12) .
In study perennial laceration was found to be signi cantly lower among teenagers (2.1%) compared to adults (11.1%) (OR: 6.288; 95% CI: 1.879-21.049). This was a similar nding to study conducted in a Swedish Medical Birth Register were women with advancing age revealed signi cantly increased risk of perennial lacerations (OR:0.65; 95%CI: 0.64 -0.66) (13) . In contrast to our study a study done in Cameroon found statistically insigni cant difference between the teenage and adult mothers on perennial laceration (OR: 0.82; 95%CI: 0.61-1.3) (2) . The reason for this lower incidence of perennial laceration in teenage mothers could be because teenagers deliver low birth weight babies compared to adults.  (2), (7), (10) .This statistically insigni cant difference in our nding could be ,because there is no special risk to adolescent mothers of hypertension associated with their young age. However, hypertension is the most common complication of pregnancy amongst women having their rst child and is therefore a common complication for many primigravida mothers (14) .
In study preterm delivery was found signi cantly higher among teenagers (60.40%) compared to adults (16.0%) (OR: 6.138; 95% CI: 3.749-10.104). This nding is similar to studies conducted in Nepal, , Cameroon ,United State, Nigeria, and Banaras Hindu University of India in which preterm delivery was found to be signi cantly higher among teenagers compared to adult mothers (OR: 1.59; 95% CI:1.14-2.20), ( OR:1.94; 95% CI: 1.34 -2.79),(RR:1.20; 95% CI:1.19, 1.20), (P =0.000) and (p <0.01) respectively (11),(2),(15),(8), (10) . It is also consistent to a study conducted in Benin, Gabon, Mozambique and Tanzania were preterm delivery was found to be highly associated with young maternal age (OR: 4.28; 95% CI: 2.05 -8.93) compared to the adult group (16) . Our nding is inconsistent with a study conducted in institute of medical science of India were preterm delivery was found to be statistically insigni cant in teenage mothers (P=0.762) (9) . This signi cantly higher result of preterm delivery we nd could be, because of Immaturity of the uterine or cervical blood supply in teenage pregnancy which in turn could increase the risk of subclinical infection and prostaglandin production, and lead to increased risk of pre-term delivery (9) .
According to our study low birth weight babies were found signi cantly higher among teenage mothers (49.3%) compared to ( (11),(15),(2), (8), (10) . Our result was inconsistent with study done in Shadan institute of medical sciences India were low birth weight was found to be statistically insigni cant among teenage mothers (p=0.153) (9) . Findings of our study were also different from a study conducted in Benin, Gabon, Mozambique and Tanzania were low birth weight was not statistically signi cant (OR: 1.29; 95% CI: 0.82 -2.01) (16) . Several reasons have been discussed in the scienti c literature, for the high risk of delivering low birth weight infant by adolescent mothers. Among others include anatomic immaturity and continued maternal growth which may represent biologic growth barriers for the fetus, this could explain the higher number of low birth weight infants among teenagers in our nding (16) .  ( 7), (8) . In contrast to our study, a study carried out in the Yaoundé General Hospital of Cameroon, found out statistically insigni cant difference between teenagers and adult mothers regarding type of delivery (OR: 1.07; 95%CI: 0.733-1, 55) (2) .Our nding was also different from study conducted in Nepal were abnormal delivery was signi cantly lower among teenagers compared to adults (OR: 0.69; 95% CI: 0.51-0.91) (11) . In our nding CPD as indication of abnormal delivery was signi cantly higher among teenagers (80%) compared to adults (20%) (p=0.000).This increased indication of CPD could be because of the fully immature pelvis of the young mother compared to adults.

Conclusion
Controlling marital status, economic status, circumcision, educational level and family support the results of this study suggested that adolescent pregnancy was associated with an increased risk of anemia, preterm birth, low birth weight, abnormal delivery and post-partum hemorrhage. However, preeclampsia and Apgar score were not signi cantly different between adolescents and adult mothers. Anemia, postpartum hemorrhage, preterm birth and low birth weight were found 2, 0.13, 6 and 0,094 times more among teenagers compared to adults. Abnormal delivery was 6 times more among teenagers. However perennial laceration was 6 times more among adults compared to teenagers.
We conclude that teenage pregnancy before the age of 20 years is associated with increased incidence of adverse maternal and fetal outcome, and we are in favor of the accepted opinion that adverse birth outcome associated with teenage pregnancy is attributable to biological immaturity rather than low socioeconomic status, and inadequate prenatal care.

Limitation Of The Study
Only mothers who were having telephone number and residents of Asmara were included in the study. Also economic status was categorized based on monthly income and monthly expense for house payment only. Ethical approval for this study was granted by Asmara College of Health Sciences research ethical approval committee. In addition, after brief explanation of the purpose of the study, written consent was obtained from the study participants and those who volunteered and participated in the study

Consent for publication
This manuscript has not been published elsewhere and is not under consideration by another journal. All authors have approved the nal manuscript and agreed for its publication.