Nutritional status and birth outcomes among pregnant teenagers in Ghana

Background: Inadequate nutritional status during pregnancy can lead to adverse birth outcomes but this is more so for pregnant adolescents as they require nutrients to meet their needs for growth and that of the foetus. The study examined the relationship between nutritional status and birth outcomes among Ghanaian teenagers in rural and urban districts of the Ashanti Region, Ghana. Methods In this prospective cohort study, 416 pregnant teenagers recruited at hospitals/health centres during antenatal care were followed until delivery. Data on weight, height (for body mass index), mid-upper arm circumference (MUAC), and nutrients intake using a repeated 24-hour dietary recall were determined. Blood samples were also taken and analysed for haemoglobin (Hb), serum levels of ferritin, prealbumin, vitamin A, total antioxidant capacity (TAC), C-reactive protein (CRP), and zinc protoporphyrin (ZPP). Birth weight, gestational age at birth, type of delivery, and outcome of delivery were collected as each teenager gave birth. Data were analysed using Chi-square (Fisher’s exact) analysis, t-test, two-way ANOVA and multiple linear regression. Results: 15.9% had low birth weight, 12.5% had preterm birth, 11.5% had a postpartum haemorrhage, and 3.1% neonatal deaths. Majority of the participants had inadequate intakes of energy (81.4%), protein (77.2%), vitamins A (97.1%) and E (97.7%), calcium (99.8%), iron (97.6%), folate (93.0%), zinc (83.7%), riboavin (88.4%), thiamin (74.4%), and ber (74.4%) intakes. Anaemia and wasting prevalence were 57.1% and 27.8%. Between-subject effects determined using Generalized Linear Modelling indicated MUAC (p=0.058), and gestational age at delivery (p=0.023) signicantly affected birth weight. A multiple linear regression analysis indicated MUAC (β=0.283, p=0.002), BMI (β=0.221, p=0.015), gestational age (β=0.285, p<0.001), and ZPP (β=0.131, p=0.057) were signicantly associated with birth weight (F(23, 223), P=0.008, R 2 =1.941). Conclusions: Inadequate nutrients intake, anaemia, and LBW were found in these pregnant teenagers. Maternal MUAC, gestational age, and serum ZPP were associated with birth weight. Interventions for improving the nutritional status of teenagers before and during pregnancy are urgently needed to reduce the risk of adverse birth outcomes. used for data analysis. Data were thoroughly cleaned for missing data. Absolute and relative frequencies were determined for nutrients intake, anthropometric, haemoglobin level, and biochemical variables. Kolmogorov-Smirnov test of normality was performed to determine whether the continuous variables data met parametric assumptions. A chi-square (Fisher’s exact test) cross-tabulation was performed to compare frequencies of nutrients intake, anthropometric variables, haemoglobin status, biochemical variables, and birth weight, and gestation age. An independent t-test and a two-way analysis of variance (ANOVA) (Generalized Linear Model test) were used for parametric comparisons, while Mann Whitney ‘U’ test was performed for non-parametric comparisons of all continuous variables, and birth weight, and gestation age. The mean differences of nutrients intake, anthropometric and biochemical variables were compared by birth weight and gestation age. Univariate and multivariate tests of variables associations were performed using the generalized linear model test. Data were presented as mean ± SD for the continuous variable. Multiple linear regression test was performed to determine predictors of birth weight. All tests were 2-tailed, and p-values < 0.05 were termed signicant.

A systematic review by Marvin-Dowle et al. [11] reported that pregnant adolescents have low macro and micronutrient intakes when compared to the recommended dietary allowances (RDA). Many studies suggest pre-pregnancy and pregnancy anthropometrics status such as excess weight gain, low weight, low MUAC, and maternal anaemia are strongly associated with preterm birth, and low birth weight infants in pregnant adults [12][13][14][15][16][17][18]. Another study in Ghana also reported an increase in oxidative biomarkers such as c-reactive protein, and low total antioxidant capacity during pregnancy [19]. During pregnancy, zinc protoporphyrin (ZPP) levels can be elevated when there is iron de ciency during erythropoiesis [20], and is less likely to be affected by in ammation [21]. It is therefore imperative to ensure adequate nutrition for women before and during pregnancy, to prevent adverse birth outcomes.
Because of the added nutrients requirements for teenage pregnant girls need for their growth, the above-reported de ciencies would have a severer consequence for pregnant teenagers. The lifelong negative consequences of poor birth outcomes make the need to address poor nutrition among adults, but more importantly teenagers even more crucial [2,18].
However, among Ghanaian pregnant adolescents, much remains to be explored regarding the effects of nutrition on birth outcomes [22]. Ghana seeks to reach the set target for LBW reduction, giving considerable attention to pregnant adolescents is necessary [18]. This study sought to understand the determinants of birth outcomes among teenage pregnant girls in selected rural and urban districts of Ghana. To the best of our knowledge, this is the rst of such studies and very timely as we seek to bridge the understanding gap on the in uencers of birth outcomes in Ghana, to provide an impetus for strengthening interventions to improve birth outcomes among pregnant women in Ghana.

Study Design
This is a hospital-based longitudinal cohort study that involved 416 pregnant adolescents, aged 13 to 19 years old. Participants were recruited in communities in three urban districts (Kumasi Metropolis, Asante Akim Central and Ejisu Juaben) and ve rural districts (Bosomtwi, Asante Akim South and North and Ahafo Ano North and South) in Ashanti Region, Ghana, from May to August 2018. The ow diagram for the study is shown in Fig. 1.

Study area and population
The study took place in 29 communities in three urban districts (Kumasi Metropolis, Asante Akim Central and Ejisu Juaben) and ve rural districts (Bosomtwi, Asante Akim South and North and Ahafo Ano North and South) in Ashanti Region, Ghana. Ashanti Region has the largest population in Ghana, with an estimated population of 5,792,187, accounting for 19% of Ghana's total population in 2019 [23]. The adolescent population in 2017 stood at 19.8% of the total population, for those aged 15-19 years [24]. The report from the 2017 Ghana Maternal Health Survey showed a high prevalence of teenage pregnancy (12.2%) and adverse birth outcomes in the region [25]. All the selected districts have at least hospitals or Community-based Health Planning Services (CHPS) compounds/health centres depending on the geographical area of the district [26]. The study population were pregnant adolescents (aged 13 to 19 years old) with gestational age up to 32 gestational weeks, who were attending antenatal clinics at their district health facilities.

Recruitment
Using a reference low birth weight prevalence of 23% reported in a previous study by Ayensu et al. [27], and a marginal error of 5%, the sample size calculated was 420 but 416 pregnant adolescents were recruited within the study period. The sample size at birth outcome was reduced to 270, due to participants fallout from the study, while others traveled out of the region. Participants were recruited on a rst come rst served basis from hospitals and health centres during antenatal care visits. We obtained days in which antenatal clinics were held for pregnant adolescents in the hospitals/health centres involved in this study. On these dates, researchers visited the hospitals/health centres, and any pregnant adolescent within the required age group (13-19 years) was eligible for recruitment. Although special antenatal clinics were held for teenagers, attendance was low due to stigma.
Hence, in the rural districts, community information centres were asked to announce and invite all pregnant adolescents to the health centres. Those that came and consented to the study were recruited.

Baseline study
Upon recruitment and signing of informed consent, dietary assessment and anthropometrics were done and blood samples collected for haematology and biochemical analysis. Questionnaires were used to collect data on antenatal visits, antenatal interventions uptake, and other pregnancy-related practices.

Dietary assessment
Dietary intakes of participants were taken using a repeated 24-hour dietary recall on 2 weekdays and 1 weekend. Household food measures were used to quantify food consumed by participants and these were later converted into grams using a standardized excel spreadsheet which has the measuring weight of Ghanaian foods. These grams were further computed into a nutrient analysis Microsoft excel software designed by the University of Ghana, Department of Food Science and Nutrition [28] which included mostly consumed Ghanaian foods. The mean nutrients intakes of macro and micronutrients were calculated by the software, and nutrients intakes were compared with recommended dietary allowances (RDAs)/estimated energy requirement (EER)/adequate intake (AI) [29].

Assessment of anthropometry
Weight, height, mid-upper arm circumference (MUAC) of participants was taken. Weight was measured with a weighing scale (model: DT602, India), to the nearest 0.1 kg while, height measurement in centimeter was taken using a portable stadiometer (Secca 213, India), all with barefoot. The values of the weight and height were used to calculate the body mass index (BMI) of the participants, and categorized using the WHO 2004 classi cation for BMI [30]. MUAC measurement was determined as a proxy indicator of maternal weight status since it has good speci city in determining weight during pregnancy. An inelastic tape measure was used to take MUAC of participants by locating the midpoint between the acromion and olecranon bone on the left hand of participants. MUAC measurement less than 22.0 cm was termed severe wasting, 22.0 cm to less than 24.0 cm as mild/moderate wasting, and 24.0 cm and above as normal [31]. All anthropometric measurements were done in duplicates and averages were used.

Assessment of Haemoglobin (Hb) level
Two milliliters venous blood sample of participants was collected into anticoagulant EDTA tubes and used to determine haemoglobin levels. The haemoglobin level was determined using Sysmex Haematology system (USA). Using the WHO cut-off for Hb, haemoglobin values less than 11.0 g/dL was termed anaemia while, 11.0 g/dL, and above was termed no anaemia [32].

Assessment of Biochemical parameters
Serum ferritin is recommended by WHO in the assessment of iron levels [33]. Because serum ferritin is likely to be high in the presence of in ammation or infection, an in ammatory marker, CRP, was also assessed [34]. Five milliliters venous blood were taken from participants into serum gel separator tubes, and later centrifuged at 4000 rpm for 5 minutes. The supernatant serum was used for the analysis of serum levels of ferritin, prealbumin, vitamin A, c-reactive protein (CRP), total antioxidant capacity (TAC) and zinc protoporphyrin (ZPPP) using a sandwich enzyme-linked immunosorbent assay (ELISA) technique (from R&D system Inc, USA) at the Clinical Analysis Laboratory of the Department of Biochemistry and Biotechnology, Kwame Nkrumah University of Science and Technology. The optical density for serum ferritin, serum prealbumin, serum vitamin A and serum ZPP was measured at 450 nm wavelength within 15 minutes, with the help of multipurpose microplate ELISA reader (Mindray MR-96A). TAC was determined using the ferric reducing ability of plasma (FRAP) assay protocol described by Benzie and Strain [35]. TAC was estimated at wavelength 593 nm using a spectrophotometer (Mindray BA-88A, China). The standard curves of known concentrations of the respective recombinant value were used for the calculation of biochemical variables. All biochemical analyses were done in duplicates. Serum vitamin A de ciency was characterized when retinol levels were < 0.70 mmol/L [36].
Serum iron de ciency was termed as serum ferritin less than 15 µg/L [37]. Low serum prealbumin was de ned as serum levels less than 50-500 mg/L [38].

Follow up
After the baseline study, participants' telephone numbers and addresses were collected and entered into a book, and the expected date of delivery was entered for each participant. All participants were called the exact date to check if they had delivered, and if yes visited for the follow-up study.

Follow up study
A structured questionnaire was used to gather data on birth experiences for each participant. The questionnaire was divided into four sections; 1). birth outcome (e.g: mode of delivery, birth status, place of delivery), 2). maternal morbidity (e.g: postpartum haemorrhage, hypertension, pelvic in ammatory disease) and mortality (e.g: the period of death, cause of death), 3). infant mortality (e.g: the period of death, cause of death), and 4). infant anthropometrics (e.g: birth weight, birth length, head circumference, gestational age at delivery).
Birth outcomes were then recorded from birth records kept by the health centre/hospital for each participant.
Physical examination and clinical evaluation of the pregnant adolescent were done by quali ed midwives to determine the gestation ages of some participants in the rural districts while, for others, a scan report of the gestation age from their antenatal card was collected. Birth outcomes such as gestational age at birth, birth weight, birth delivery method, neonatal mortality, stillbirth, spontaneous and induced abortions, postpartum morbidity of mothers were obtained from the hospitals/health centres records and categorized into those that were normal or adverse.
Preterm birth was de ned as delivery less than 37 completed weeks of gestation [39]. Stillbirth was termed as the death of foetus that occurs before the complete ejection from its mother during conception, regardless of the duration of gravidity [39]. Postpartum haemorrhage was de ned as a blood loss of more than 500 ml to 1,000 ml within the rst day after birth [40]. Low birth weight was de ned as a birth weight of newborns less than 2500 g (2.5 Kg) [41].

Data analysis
The IBM Statistical Package for Social Sciences version 25 (SPSS IBM Inc Chicago, USA) was used for data analysis. Data were thoroughly cleaned for missing data. Absolute and relative frequencies were determined for nutrients intake, anthropometric, haemoglobin level, and biochemical variables. Kolmogorov-Smirnov test of normality was performed to determine whether the continuous variables data met parametric assumptions. A chi-square (Fisher's exact test) cross-tabulation was performed to compare frequencies of nutrients intake, anthropometric variables, haemoglobin status, biochemical variables, and birth weight, and gestation age. An independent t-test and a two-way analysis of variance (ANOVA) (Generalized Linear Model test) were used for parametric comparisons, while Mann Whitney 'U' test was performed for non-parametric comparisons of all continuous variables, and birth weight, and gestation age. The mean differences of nutrients intake, anthropometric and biochemical variables were compared by birth weight and gestation age. Univariate and multivariate tests of variables associations were performed using the generalized linear model test. Data were presented as mean ± SD for the continuous variable. Multiple linear regression test was performed to determine predictors of birth weight. All tests were 2-tailed, and p-values < 0.05 were termed signi cant. Ethical approval for the study was obtained from the Committee on Human Research Publication and Ethics (CHRPE) of the Kwame Nkrumah University of Science and Technology, KNUST, (Kumasi, Ghana) (Reference: CHPRE/ AP/ 236/18). Study protocols/aims were rst explained to all participants in their local language (Asante Twi). Written and signed informed consent was obtained from all participants by following CHRPE regulations before recruiting for the study. Also, parents/guardians signed a written consent form on behalf of participants less than 16 years old. Figure 2 shows birth outcomes associated with the pregnant adolescents. Adverse birth outcomes found among pregnant adolescents were low birth weight (15.9%), preterm birth (12.5%), stillbirth (2.2%), neonatal mortality (3.1%), postpartum haemorrhage (11.5%) and postpartum hypertension (2.2%). However, a greater percentage of the adolescent mothers had normal birth weight of child (84.1%), term birth (87.5%), and survived neonates (92.6%).

Results
Nutrients adequacies and inadequacies among pregnant adolescents and their relationship with birth weight and gestational age are displayed in Table 1. The majority of the pregnant adolescents reported inadequacies in energy (81.4%), protein (77.2%), vitamins A (97.1%) and E (97.7%), calcium (99.8%), iron (97.6%), folate (93.0%), zinc (83.7%), ribo avin (88.4%), thiamin (74.4%) and ber (74.4%) intakes when compared to recommended dietary allowance (RDA)/estimated energy requirement (EER)/adequate intake (AI). The proportions of adolescent mothers who gave birth to low and normal-weight children did not signi cantly vary by nutrients intake (p>0.05). Proportions of adolescent mothers who had preterm and term birth were not signi cantly related to nutrients intake except for ber intake, in which, more teenagers who experienced preterm births had inadequate intake than termed birth participants (86.8% versus 70.6%, p = 0.050).
The relationship between anthropometric and birth outcomes is presented in Table 2. More than a fourth (27.8%) of the pregnant adolescents were either overweight/obese while, 1.9%, were underweight. Generally, 28.8% of the participants were wasted during pregnancy; of which, 5.5% were severely wasted. Higher proportions of adolescent mothers who had low birth weight (7.0%) were severely wasted than those who had normal birth weight (5.0%, p = 0.032). The proportions of adolescent mothers who had low and normal birth weight did not vary by BMI status(p>0.05). Also, proportions of adolescent mothers who had preterm and term birth did not vary by BMI status and MUAC status (p>0.05). Table 3 indicates how haemoglobin, biochemical indices affect birth weight and gestation age. Anaemia prevalence (57.1%) was high among pregnant adolescents. The majority of the participants had low serum prealbumin (97.2%), serum vitamin A (86.4%) but normal serum ferritin (96.4%). The proportions for haemoglobin (Hb), serum ferritin, prealbumin, and vitamin A did not signi cantly vary by birth weight and gestation age (p>0.05). Also, the means of serum CRP, TAC, and ZPP did not vary by birth weight status and gestation (p>0.05).
The results of between-subject effects, conducted in univariate and multivariate Generalized Linear Model is presented in Table   4. Among all the multiple variables tests of associations on birth weight, only MUAC (p=0.058) and gestational age (p=0.023) had a signi cant effect on the child's birth weight.

Discussion
The present study aimed at determining nutritional factors in uencing birth outcomes among pregnant adolescents in rural and urban communities in Ghana. The major ndings are that the prevalence of low birth weight, preterm birth, stillbirth, neonatal mortality, haemorrhage, and postpartum hypertension were 15.9%, 12.5%, 2.2%, 3.1%, 11.5%, and 2.2%. Among the nutritional factors assessed, only MUAC status during pregnancy was associated with birth weight, while, ber intake was associated with the gestational age of the mothers. Combined analyses of study variables showed that both the MUAC and gestational age of participants was effectual in determining birth weight. Gestational age and MUAC status during pregnancy were the signi cant predictors of birth weight.
The prevalence of low birth weight and preterm birth in our study is higher than that reported in other regions of Ghana [42,43], the current 2017 Ghana Maternal Health Survey [44], in Brazil [8,22], and Ethiopia [45]. Low birth weight is a signi cant underlying cause of neonatal and infant mortality in low-and-middle-income countries such as Ghana [44], which implies that participants with low birth weight are at increased risk of neonatal and infant mortality. A multiple regression analysis showed that maternal anthropometry and gestation age were co-effective in determining birth weight. The result indicated that MUAC (β = 0.283, p = 0.002), BMI (β = 0.221, p = 0.015), gestation age (β = 0.285, p < 0.001), and ZPP (β = 0.131, p = 0.057) were signi cantly associated with birth weight. The association from the standardized beta coe cient was positive which implies that a point increase in MUAC, BMI, gestation age, and ZPP values would likely result in an increase in the birth weight of babies, and vice versa. The results suggest that pregnancy anthropometrics and ZPP were the nutritional factors that co-effectively determined birth weight. Also, gestational age attained can predict birth weight in adolescent pregnancy. Studies have predicted that maternal gestational age and weight status are associated with birth weight [46][47][48][49].
The study found that the adolescent girls had inadequate intake of macronutrients; energy, protein, ber, and micronutrients; thiamin, ribo avin, folate, vitamins A and E, iron, calcium during pregnancy and their mean intakes for the participants were far lower than RDA/EER/AI. Poor dietary intakes owing to their poor income status at that stage of life might have contributed to nutrients de ciencies in these populations. Another school of thought is that pregnancy comes along eating disorders including anorexia, food aversion, hyperemesis, and other conditions that might limit food intake [50,51]. A study by Baker et al. [52] also found micronutrient de ciencies among pregnant adolescents in the United Kingdom. In our case, most of these teenagers were having their rst pregnancy, and are less likely to adapt to these pregnancy symptoms, which can reduce their appetite for food, and consequently lead to nutrients de ciencies. However, there was no direct effect of nutrients intake on the infant's birth weight, which suggests dietary intake might be indirectly linked with birth weight, and in uence by maternal weight. The same association was found between adolescent mothers with preterm and term birth, except for inadequate ber intake, in which, preterm birth participants were more likely to have inadequate ber intake. On the contrary, some studies had reported a relationship between maternal dietary intake and weight gain [53,54], and weight gain and birth weight [49,55,56], low iron and folate intake and small-for-gestational-age babies [52] which suggest that fetal development may rely on nutrients stores during pregnancy. Notwithstanding to exception that, when maternal nutrients stores are depleted under certain circumstances such as severe hunger, there might also be an association between maternal diet and birth outcomes such as birth weight and gestation age [57,58].
Although, maternal weight gain, nutritional status, and anaemia has a strong association with birth outcomes such as birth weight and gestational age [46-48, 54, 59]. Only a few prior studies have examined these risk associations among pregnant adolescents in a developing country like Ghana. In the current study, we found that 27.8% of the participants were either overweight or obese during pregnancy while, few of them were underweight. Based on the MUAC assessment, 28.8% of the participants were wasted. We found a signi cant relationship between birth weight and MUAC status during pregnancy, which was consistent with a study by Assefa et al. [12]. Low birth weight adolescent mothers were more likely to be severely wasted during pregnancy. Additionally, we found a high prevalence of maternal anaemia, low serum levels of prealbumin (97.2%), and vitamin A (86.4%). High maternal anaemia prevalence among teenagers has been reported in the study area [60], another region in Ghana [61], and Malaysia [62]. This may be due to poor nutrition at adolescent age, since there is a competition of nutrients for their tissue growth, before that of the developing foetus [2,60]. Anaemia prevalence was not signi cantly associated with birth outcome, but anaemic mothers had higher low birth weight. In contrast to our ndings, studies done in Ghana, China and India have reported that iron de ciency anemia was signi cantly associated with low birth weight and preterm birth among pregnant adults [43,63,64]. Maternal low serum levels of prealbumin, ferritin, and vitamin A were found non-signi cantly higher among low birth weight participants than normal birth weight participants. In this study, adolescent mothers who had preterm birth were also non-signi cantly more likely to have anaemia, low serum levels of prealbumin, and vitamin A during pregnancy than those who had term birth. The study found that zinc protoporphyrin (ZPP) levels were associated with an infant's birth weight (β = 0.131, p = 0.057). ZPP as a biomarker is used to indicate an adequate supply of body iron to red blood cells (RBC) in the bone marrow [20]. ZPP is usually elevated when there is iron de ciency during erythropoiesis, which can be measured and used as a screening marker in pregnant women [20]. A study by Harthoorn-Lasthuizen et al. [65] reported that pregnant women who were anaemic throughout pregnancy had an elevated ZPP, indicating iron de ciency anaemia during RBC formation.
Another study Schiman et al. [66] reported that an elevated ZPP level in the rst trimester can predict 56% for third-trimester anaemia. In contrast to Schiman et al. ndings, a study by Tchai et al. [67] reported that using only ZPP as a marker shows lower e ciency in predicting iron de ciency anaemia in pregnancy when compared to serum ferritin and ZPP/heme ratio. The current study did not report the prevalence of elevated ZPP in the participants, however, the majority of the study participants reported low haemoglobin levels (anaemia). It is unclear whether the anaemia prevalence in the participants affected the ZPP levels and/or, can result in elevated ZPP levels, which resulted in a signi cant association with birth weight. Therefore, using reports from previous studies of ZPP and iron de ciency anaemia in pregnancy, the current study suggests that monitoring of ZPP throughout pregnancy in the clinical setting can a good marker in deciding which maternal women will need iron supplementation. Maternal biochemical parameters such as total antioxidant capacity, c reactive protein did not show any signi cant association with birth outcomes, because some of these parameters are related in advanced maternal age births [19]. For adolescents, it is less likely to observe oxidative stress and other in ammatory conditions that would in uence birth outcomes.
The implication of these results for clinical practice is that pre-pregnancy and pregnancy nutritional status of high-risk teenagers may need monitoring and intervention. Additionally, early intervention to modify the weight status of pregnant adolescents by improving the quantity and quality of maternal diet may in uence the infant's birth weight and consequently lead to a reduction in other adverse birth outcomes such as preterm birth, stillbirth, neonatal, and infant mortality. In developing countries like Ghana, intervention programs and policies on women's health, especially pregnancy, have largely focused on the adults, with less attention to the vulnerable teenagers. This is because the ethical conduct of the society frowns at teenage pregnancy. Let not forget that, some of these nutritional factors such as maternal poor diet and low pregnancy anthropometry can alter fetal growth and development which would consequently harm their adult life [68]. We advocate that health intervention programs and nutrition education on improving nutritional status, and early recognition of pregnancy complications should be made inclusive, to incorporate the adolescents, to prevent poor nutritional status leading to adverse pregnancy outcomes among these populations.
Although the study reports interesting results among these cohorts in Ghana, some limitations may affect the interpretation of the results. To begin with, the small sample size for the birth outcome data might not have in uenced some statistical associations and analyses. Secondly, we could not follow up on participants throughout pregnancy to collect nutritional factors, therefore the trends in changes in nutritional parameters that could affect birth outcomes were not observed. Lastly, this was a cross-sectional study, hence, future longitudinal cohorts/interventional studies would properly enlighten our understanding of nutritional factors underpinning adverse birth outcomes among teenagers.

Conclusions
Nutrients de ciencies, wasting, anaemia, low serum prealbumin, and vitamin A were found high among pregnant adolescents during pregnancy. Also, we observed poor adverse birth outcomes such as low birth weight, preterm birth, neonatal death, stillbirth, postpartum haemorrhage and hypertension among the participants. Maternal weight and gestational age were associated with birth weight. Health intervention programs and nutrition education on improving the nutritional status, and early recognition of pregnancy complications should be made inclusive, to incorporate the adolescents, to prevent poor nutritional status leading to adverse pregnancy outcomes among these populations.

Declarations
Ethics approval and consent to participate Ethical approval for the study was obtained from the Committee on Human Research Publication and Ethics (CHRPE) of the Kwame Nkrumah University of Science and Technology, KNUST, (Kumasi, Ghana) (Reference: CHPRE/ AP/ 236/18). Written and signed informed consent was obtained from all participants by following CHRPE regulations before recruiting for the study. Also, parents/guardians signed a written consent form on behalf of participants less than 16 years old.

Not applicable
Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare no competing interests regarding the publication of this study.

Funding
The study received funding support from Nestle Foundation. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Authors' contribution LAG, RAA, CA, HEL, WA, and AKE conceptualized the design and wrote the protocols for the study as well as participated in the supervision of data collection. LAG, RAA, CA, and OAB performed the analysis and interpretation of the results. LAG and OAB wrote the rst and nal draft manuscript. RAA and CA participated in the analysis and manuscript reviews. All authors read and approved the nal manuscript.   Figure 1 Study ow show date and duration, study design and data collected This is a list of supplementary les associated with this preprint. Click to download.