This study enrolled 836 expectant women of whom 209 were below 20 years of age while 627 were aged between 20–30 years in a ratio of 1:3. This ratio was higher than three studies under comparison where the authors opted for a 1:2 [5, 24] and 1:1[23] adolescent to adult ratio respectively. The high number of controls in comparative studies help clearly determine whether the effect of risk factors seen among the cases is isolated or universal irrespective of the risk status, because this study design is used to determine causality. This was also witnessed in a nested case-control study in a cohort of 4,591 expectant women from Northern England where the authors [25] adopted a higher case to control ratio of 1:6 compared to the current study. This study hypothesized that being aged below 20 years is a risk factor for undesirable pregnancy outcome. Furthermore, it is easier to identify women within safe age of pregnancy (> 20 years) compared to adolescent mothers in many public healthcare facilities in Kenya.
This study reports that about two thirds (65.4%, n = 547) of the expectant mothers enrolled were married while 288 (34.4%) were single and 0.2% were widowed. A high proportion of married women was also reported in a study conducted at the Cape Coast Metropolis in Ghana[26] where the proportion of those married (51.2%) was higher than single women (46.3%), while the rest were either divorced or opted not to respond to the marital status question. Similarly, in Northern England, 55.2% of those enrolled were married while 23.1% were single. The authors (Marvin-Dowle et al., 2018) further noted that 21.5% were not married but living with a partner. This high proportion of unmarried women compared to the current study could be attributed to two factors. First, methodologically, the authors focused on adolescent pregnancies reported in national surveys while the current study was a hospital-based study. Very low proportions of marriages are reported among adolescents as adolescent marriages are both forbidden by law and most of adolescent pregnancies are either undesired or unplanned for.
When the level of education of the participants was assessed, majority (42.5%) of this study’s participants attained a secondary level of education followed by those with a tertiary (39.5%) level of education and primary (17.9%) level of education. Only one participant was reported as not having any formal education. This finding is similar to that conducted in major towns in Kenya where nearly half (44.5%) of the study’s participants had attained a secondary level of education[27]. The authors[27] merged those with primary level of education or less and got a higher proportion (34.9%) compared to that of the current study. Furthermore, their proportion of tertiary level of education stood at 20.6% [27]. This high proportion of secondary and tertiary level of education reported in both studies could be attributed to the fact that Kenya has high literacy levels courtesy of state sanctioned universal primary education and transition to secondary level of education. However, the proportionate distribution of level of education reported in this study varies from that reported across studies conducted in Africa. In Ethiopia, the authors [5] reported a higher (8.28%) proportion of those with no formal education compared to the current study. This study collected level of education data because previous studies have correlated highest level of education and reproductive health awareness. Higher level of education leads to lower rates of adolescent pregnancy due to its positive effect on contraceptive acceptance and use. Education also increases autonomy and decision-making power of an individual. Furthermore, it increases economic independence leading to postponement of marriage and reduction of fertility. The lower the level of education, the earlier the age of consent to first sexual encounter which increases the risk of adolescent pregnancy.
About half (48.6%) of this study’s participants were unemployed followed by students (23.8%), those who were self-employed (21.1%) and the formally employed (6.5%) being the least represented. This high prevalence of students and unemployed could be attributed to the fact that this study enrolled one-third of its participants who were adolescent mothers. Furthermore, majority of the adult mothers were young and yet to be engaged in meaningful employment as most of them did not attain tertiary level of education. The proportionate findings of this study match that reported in a local study conducted in Kenya’s major towns and cities (Nairobi, Mombasa, Kisumu, Machakos, and Kakamega) where 44.3% were unemployed, 32.6% were students and 44.3% were employed [27].
The mean number of antenatal visits during the pregnancy of this study’s participants (irrespective of their age grouping) was 3.960 ± 1.721 visits. This finding is close to that reported in Ethiopia[5] where the average visits for adolescent pregnancies 3.26 ± 1.24 (cases) and 3.54 ± 1.5 for adult pregnancies (controls). When the antenatal visits were stratified by frequency, 55 (6.6%) of all this study’s participants did not attend any antenatal clinic, 212 (25.3%) made 1–3 visits while majority of them (n = 569; 68.1%) had at least 4 visits. In South Africa’s KwaZulu-Natal, the authors[28] determined that most of the study participants had two antenatal visits during their pregnancy under study. The difference in the average number of antenatal visits could be attributed to the fact that maternity services are free in Kenya thus an increased access by the pregnant mothers.
When we compared the rates of contraceptives use, the least uptake was adolescent expectant mothers (11.5%) compared to their adult counterparts (48.3%). This study’s average is higher than that reported in East and Southern African Countries at 23.08%[29] and a local study conducted in Kenya[27] at 18.07%. Higher overall contraception rate than that found in this study was reported in a Wogedi town in Northeast Ethiopia[30] at 46.3%.
The main maternal outcomes of interest in this study were anaemia at the time of labour, hypertension in pregnancy, preterm labour, caesarean section and perineal trauma. Anaemia at the time of labour was more prevalent in adolescent pregnancies at 16.3% compared to adult pregnancies at 9.4%. This relationship was statistically significant (p = 0.008), with adolescent mothers being more at risk of anaemia at the time compared to adult expectant mothers (RR = 1.556; 95% CI: 1.156, 2.096). The proportion of expectant mothers with anaemia at the time of labour, who were enrolled in this study was close to that reported in Raymond Mhlaba sub-District of South Africa [23] where 13% of adolescent mothers were diagnosed with anaemia at labour compared to 10% of adult mothers. Higher prevalence of anaemia at labour for adolescent mothers of 46% compared to 32% among adult mothers was reported in Lahore-Pakistan [31] in a study conducted among 300 participants. The authors [31] used a haemoglobin (Hb) cut-off of < 11 g/dL noted a statistically significant relationship (p = 0.010). The high prevalence of anaemia in adolescents than in older women might be because of their nutritional deficiencies related to social and environmental factors, growth spurt and onset of menstruation which increase iron requirements [23].
Hypertension in pregnancy was seen in 14.8% of the adolescent mothers compared to 8.5% of the adult expectant women enrolled in this study, a relationship that was statistically significant (p = 0.012; RR = 1.543; 95% CI: 1.133, 2.100). This finding was higher than that reported in South Africa at 2.1% among adolescent mothers and 2.0% among adult mothers. In a large prospective study conducted in Northern England[25] that enrolled 640 adolescent and 3951 adult mothers, the overall prevalence hypertension in pregnancy was 3.6%. The proportionate difference could be attributed to spatial and sample size differences in both studies. Prevalence rates tend to be lower in large sample sizes compared to studies with fewer participants. Secondly, because of the difference in health promotion and primary care services offered in Northern England -through the National Health Service- compared to Kenya, mitigating interventions could be instituted earlier hence the lower prevalence. In Nigeria, there were nearly equal proportions of hypertension in pregnancy for both adolescent and adult mothers at 4.1% and 3.8% respectively, a relationship that was not statistically significant (p = 0.934). This current study findings contrast those reported in Egypt where the authors[19] reported lower proportion (6.9%) of hypertension in adolescents compared to the adult mothers (8.3%), a relationship that was found to be statistically significant (p = 0.001). These differences could be attributed to both study methodology and settings. Whereas this study adopted a 1:3 ratio, the study from Egypt used a 1:5 ratio. Larger case-control ratios directly affect the proportions of variable of interest. Additionally, the socioeconomic differences in Egypt and Kenya, where Egypt is a developed economy compared to Kenya could also attribute to a more sedentary lifestyle further increasing the risk of hypertension among adolescents [10, 19]. A similar finding was also reported in South Africa [32], the rate of hypertension was noted to be higher in adolescents (39.8%) while it was at 26.4% in adult mothers, a difference that was noted to be statistically significant (p = 0.003). However, in Turkey [33], lower proportion (13.7%) of hypertension among adolescent mothers compared to their adult compatriots at 14.3% was reported, a relationship that was not statistically significant (p = 0.622).
Preterm labour was witnessed more in adolescents (30.1%) compared to adult (21.4%) mothers enrolled in this study, a relationship that was found to be statistically significant (p = 0.011; RR = 1.395; 95% CI: 1.092, 1.795). This finding compares to that reported in Lahore-Pakistan at 11.3% among adolescent mothers and 4.6% of adult mothers delivering below 37 weeks’ gestation, a relationship that was statistically significant [31]. Additionally, caesarean Section was noted more among adolescent mothers (24.4%) compared to adult mothers (23.3%); however, the difference was not statistically significant (p = 0.778; RR = 1.050; 95% CI: 0.800, 1.379). A similar trend was noted in Raymond Mhlaba sub-District of South Africa at 27.4% among adolescent mothers and 13.7% among adult mothers [23]. Contrasting findings of 16.6% among adolescent mothers who underwent caesarean section was reported in Lahore-Pakistan [31] compared to 21.33% among adult mothers with no statistically significant association noted (p = 0.290). In Turkey [33] where 66.8% of the adolescent mothers underwent a caesarean section a proportion that was lower than the 87.4% of adult mothers who underwent the same procedure. The variance could be attributed to methodological differences in the studies under review compared to the current study. In Pakistan [31], the adolescent to adult ratio was 1:1 compared to the 1:3 adopted in this study. In Turkey, the authors opted for a larger sample size (n = 1861) compared to the 836 enrolled in the current study. Furthermore, in Turkey [33] adopted a retrospective design compared to the prospective nature of the current study. Because of the large sample size and study design differences, the findings of these two studies on caesarean section rates for adolescent and adult mothers were likely to contrast.
This study reports that more than one-third (37.8%) of adolescent mothers had perineal trauma (perineal tears and episiotomy) compared to adult mothers at 27.0%, a relationship that was found to be statistically significant (p = 0.004; RR = 1.144; 95% CI: 1.040, 1.259). A similar trend was reported in a study conducted in the Northwest region of Ethiopia [22] where 30.9% of adolescent mothers had perineal tears compared to 10.9% of adult mothers. The findings in the current study are higher than that reported from a retrospective study conducted in Tigray region of Ethiopia where perineal tears were witnessed in 7.4% of adolescent mothers compared to 5.2% of adult mothers. This proportionate difference could be attributed to difference in study designs. Retrospective studies are often fraught with incomplete data, a phenomenon that could explain the marked proportionate difference.
This study reports that sub-optimal antenatal care attendance is not a significant predictor of adverse maternal outcomes such as anaemia in labour, hypertension and preterm labour. These findings match that reported in Kilifi County [34] where the authors stated that although there were higher proportions (31%) of preterm labour among adolescent mothers, the possible biological explanation could be prematurity as both the physiological and anatomical systems were still developing. Although the authors also attributed this preterm labour to inadequate and lack of proper antenatal care offered to adolescent mothers to address their medical and nutrition needs, the relationship was not statistically significant.
In this study, more adverse maternal complications were noted among adolescent pregnant mothers compared to their adult counterparts. Perineal trauma was the outcome with the highest incidence of 29.7% while hypertension in pregnancy had the least incidence of 10.0% These findings compared to those of a study conducted in Nigeria [15] where more adverse maternal complications were also noted among adolescent pregnant mothers compared to their adult counterparts however the most common adverse outcome was anaemia at labour and it was attributed to inadequate nutrition during the antenatal period which could also be linked to adverse fetal outcomes such as low birth weight.
The three main fetal outcomes of interest in this study were low birth weight, low 5-minute Apgar score and stillbirth. One quarter (25.4%) of adolescent mothers gave birth to newborns with a low birthweight compared to 19.1% of their older compatriots, a difference that was not statistically significant (p = 0.061). A higher proportion of low birthweight among adolescent mothers was also reported in a number of studies under comparison. In Turkey [33], 44.5% of adolescent mothers had low birthweight newborns compared to 21.8% of adult mothers, a difference that was found to be statistically significant (p < 0.001). Similarly, in a study conducted in Tigray-Ethiopia [14] where 17.5% of adolescent mothers gave birth to low birthweight newborns compared to 6.8% of adult mothers. These findings contrast that of a retrospective comparative study done by Hoque et al in a tertiary hospital in South Africa [35] where the rate of low birth weight was almost similar between the adolescents and adult mothers with no statistical difference, p = 0.174. The difference could be attributed to the study designs and sample size differences in both studies. The current study was a prospective study whereas the study in South Africa was retrospective thus possibly prone to missing information compared to the current prospective study.
The study further assessed stillbirths as a fetal outcome among the study’s participants who were stratified by age. Younger (adolescents) mothers (1.9%) were less likely to have still births compared to adult mothers at 3.7%. These finding contrasts that reported in a retrospective study conducted in Nigeria where the rate of still births was higher among adolescent mothers at 16.2% compared to older women above 20 years, which was 12.4% [15]. The difference could be attributed to methodological difference in the study in Nigeria compared to current study. In Nigeria, the study was a retrospective review and over a period of 6 years while this study, was prospective and over 6 months. Additionally, the findings of this study contrasts that reported in retrospective study conducted in Ethiopia where the rate of stillbirths was more in adolescents at 6.1% than (Abebe et al., 2018). This is similar to the findings of a study done in Cameroon [36] where there was also no statistical significance, p = 0.080. A phenomenon that could be explained by the fact that our study was not designed to detect risk factors such as smoking or alcohol abuse among the adolescent population.
Lastly, when the study controlled for the effect adverse maternal outcomes on fetal complications in the regression model and noted that preterm labour significantly (p = 0.005) increased the risk of low 5-minute Apgar score by five folds (RR = 5.714; 95% CI: 1.569, 20.816). Similar to this study’s findings, a study conducted in Kilifi County[34] also noted a high prevalence of low birthweight newborns among adolescent mothers. The authors[34] argued that this phenomenon could be since a high proportion of preterm labour increases the risk of low birth weight as these newborns had not attained their full intrauterine growth weight. A situation more common among adolescent mothers who were malnourished and had a low number of antenatal visits. Additionally, in a prospective case-control study conducted in Northeast India among 495 expectant mothers [37], adolescent mothers were more at risk (RR = 1.655; 95% CI: 1.039, 2.636) of preterm labour after controlling for adverse maternal outcomes compared to their adult counterparts, a finding that is similar to the current study. The authors[37] argued that the although the mechanism of preterm labour among adolescent mothers is still unclear; one of the explanations could be immaturity of the uterine and cervical blood supply in young mothers which leads to increase in prostaglandin production leading to preterm labour. Secondly, lack of proper menstrual preconditioning among young adolescent interferes with the process of decidualization and trophoblastic invasion that increases the risk of defective deep placentation. This could increase the likelihood of preterm deliveries in adolescent primigravidae [37]. In Pakistan, the authors[31] noted that maternal youth is a risk factor for preterm labour and adverse fetal outcomes. This creates the need to screen for biological risks among expectant adolescent mothers such as cervical shortness, infections and inadequate nutrition. Education about these risks should be given and frequent antenatal visits advised.