Birth weight and renal markers in children aged 5–10 years in Cameroon: a cross-sectional study

Background A relationship exists between birth weight (BW) and glomerular filtration rate (GFR) in postnatal kidney. Willing to fill a gap of knowledge in sub-Saharan Africa, we assessed the effect of BW on blood pressure (BP), proteinuria and GFR among Cameroonians children. Methods This was a cross-sectional hospital-based study from January to April 2018 at the Yaounde Gynaeco-Obstetric and Paediatric Hospital (YGOPH). We recruited low BW (LBW) [< 2500 g], normal BW (NBW) [2500-3999 g] and high BW (HBW) [> 4000 g] children, aged 5–10 years, born and followed-up at YGOPH. We collected socio-demographic, clinical (weight, height, BP), laboratory (proteinuria, creatinine), maternal and birth data. The estimated GFR was calculated using the Schwartz equation. Results We included 80 children (61.2% boys) with 21 (26.2%) LBW, 45 (56.2%) NBW and 14 (15.5%) HBW; the median (interquartile range) age was 7.3 (6.3–8.1) years and 17 (21.2%) were overweight/obese. Two (2.5%) children, all with a NBW (4.4%), had an elevated BP whereas 2 (2.5%) other children, all with a LBW (9.5%), had hypertension (p = 0.233). Seven (8.7%) children had proteinuria with 19, 2.2 and 14.3% having LBW, NBW and HBW, respectively (p = 0.051). Equivalent figures were 18 (22.5%), 14.3, 24.2 and 28.6% for decreased GFR, respectively (p = 0.818). There was a trend towards an inverse relationship between BW and BP, proteinuria and GFR (p > 0.05). Conclusion Proteinuria is more pronounced in childhood with a history of LBW and HBW while LBW children are more prone to develop hypertension. Regular follow-up is needed to implement early nephroprotective measures among children with abnormal BW.


Background
There is a correlation between birth weight (BW) and glomerular number, density, volume, size and filtration rate in postnatal kidney [1][2][3][4][5][6]. This is supported by Barker's and Brenner's hypothesis; the latter, which supports the intrauterine origin of later-life health, explains the relationship between BW and the number of nephrons as well as the risk of developing hypertension and chronic kidney disease (CKD). Consequently, long-term cardiovascular and renal health disorders are likely to occur [7][8][9].
It has been evidenced that low birth weight (LBW), prematurity and growth restriction are markers of an adverse intrauterine environment whereas high birth weight (HBW), exposure to maternal diabetes and rapid growth during early childhood emerge as developmental risk factors for chronic diseases [10]. Compared with normal birth weight (NBW), LBW is associated with reduced kidney volume, increased risk of perinatal morbi-mortality, underweight and short stature. It is also linked to elevated blood pressure, hypertension, proteinuria, natriuresis, podocytopenia and focal segmental glomerulosclerosis, and occurrence and rapid progression of CKD [4,5,[9][10][11][12][13][14][15][16]. Moreover, studies revealed that LBW is associated with male sex predilection of CKD and early onset of end stage renal disease in patients with autosomal dominant polycystic kidney disease, without racial predominance [12,[17][18][19]. In contrast, HBW increases the risk of hypertension in children and decreases the risk in adults; furthermore, it increases proteinuria in children with maternal diabetes as well as the risk of diabetes-associated end stage renal disease [10].
In sub-Saharan Africa, the relationship between BW and renal markers has been scarcely explored, especially in school-aged children. Studies assessing the relationship between BW and blood pressure revealed a consistently positive association in neonates, an inverse association among children and inconsistent results in adolescents [20][21][22][23][24][25]. LBW was also identified as a nontraditional risk factor for CKD in this setting, and contributed to the burden of disease [26].
In Cameroon, CKD is highly prevalent in adult populations, ranging from 10 to 14.2% and driven by known modifiable risk factors for chronic nephropathy [27,28]. In children, studies revealed no association between BW and blood pressure (BP) [29,30]. However, no study had yet assessed the spectrum of renal markers according to BW among children. In this context, we undertook the present study aiming to evaluate the relationship between BW and blood pressure, proteinuria and glomerular filtration rate (GFR) in children aged 5-10 years and living in Yaoundé, Cameroon.

Study design and setting
This was a cross-sectional hospital-based study carried out over a period of 4 months (January to April 2018) at the Yaounde Gynaeco-obstetric and Paediatric Hospital (YGOPH). The YGOPH is one of the tertiary health care facilities in Yaounde, the capital city of Cameroon. Inaugurated in 2002, the YGOPH has a capacity of 240 beds; it offers mainly gynaeco-obstetric and paediatric care to patients. The study was approved by the Institutional Review Board of the Faculty of Heath Sciences of the University of Buea and the Ethical Committee of the YGOPH.

Study participants
The study involved children between 5 and 10 years, born between 2007 and 2012 at YGOPH, who used to consult in this hospital for pediatric problems, and whose parents/guardians gave their consent. We excluded from the study children suffering from renal and urinary tract malformations, CKD, diabetes, HIV infection, hepatitis B and C infection, and sickle cell disease. Children born between 2007 and 2012 were identified from the maternity registers and recorded in a book. Based on the BW of the maternity register, children were divided into 3 groups: LBW (< 2500 g), NBW (2500-3999 g) and HBW (≥4000 g). We used the sample size calculation formula for cross-sectional studies and considered the clinical prevalence of hypertension in children of 3.5% to obtain a minimum sample size of 70 participants [31,32]; we selected 1 HBW child for 2 LBW and 4 NBW children. The recruitment of LBW and HBW participants was consecutive on their appearance in the register while NBW children were randomly selected. In case of refusal to participate, the next name on the register was selected and the procedure repeated.

Data collection
For each eligible participant, the child's parent/guardian was contacted through a phone call during which the study procedures were fully explained and an appointment fixed based on their availability. During the meeting with the parent/guardian and their child, the study was once again explained to them and an assent form signed. A self-designed and pre-tested questionnaire was used for data collection. Data collected included sociodemographic details (age, sex), clinical characteristics (weight, height, systolic and diastolic blood pressure), maternal history of pregnancy (age, type of pregnancy, maternal illness, smoking and alcohol consumption), birth characterisitics (weight, gestational term and age, and history of child reanimation and/or hospitalization) and laboratory parameters (proteinuria, and serum and urinary creatinine). We used an appropriate cuff size of 13.5 to 22 cm according to the American Academy of Pediatrics guidelines to measure BP [32]. After 5 min of rest with the participant in the sitting position, the back supported and feet uncrossed on the floor, we used an automated sphygmomanometer (OMRON HEM705CP, Omron Matsusaka Co, Matsusaka City, Mie-Ken, Japan) to measure BP on the right arm placed at the level of the heart, stretched out on the table with the palm facing up. The cuff was appropriately placed and then the machine was switched on. Three readings were taken consecutively and their mean calculated and recorded. Using BP table levels for sex, age and height percentiles, the BP percentile was recorded. When the BP was ≥90th percentile, it was repeated weekly up to two times; when it remained the same on the third measurement, it was measured using an aneroid sphygmomanometer twice consecutively; the mean of these measures were calculated and recorded as the final value.
Each participant provided 50 ml first morning midstream urine for an immediate semi-quantitative measurement of dipstick proteinuria using the CombiScreen 7SL PLUS 7 test strips (Analyticon Biotechnologies AG, D-35104 Lichentenfeis, Germany). Participants with at least traces on urine dipstick for proteinuria on the first sample were given an appointment 1 week later to repeat the urine dipstick. Those with a second urine sample still showing at least traces for proteinuria were seen 1 week later for another urine dipstick test. When the proteinuria persisted even just as traces after the second repeated urine dipstick, we proceeded to estimate the 24-h proteinuria from urine protein to creatinine ratio (PCR). We equally collected 3 ml of whole blood from an antecubital vein for serum creatinine and subsequent calculation of the GFR. Urine and blood samples were transported to the laboratory for processing. Serum and urinary creatinine were measured with a kinetic modification of the Jaffé reaction using a Human visual spectrophotometer (Human Gesellschaft, Biochemica und Diagnostica mbH, Wiesbaden, Germany) and Beckman creatinine analyzer (Beckman CX systems instruments, Anaheim, CA, USA) while urinary protein was measured using pyrogallol red-molybdate complex with Teco diagnostics tests (Teco Diagnostics, Anaheim, CA, USA).

Definitions and calculations
Delivery was categorized as preterm (< 37 weeks of gestation), normal (37 to 42 weeks) or post-term (> 42 weeks). Small for gestational age (SGA) was defined by BW < 10th centile for that gestational age (GA) whereas large for GA (LGA) was a BW > 90th centile for that GA and appropriate for GA (AGA) corresponded to BW between the 10th and 90th centile for that GA. We grouped children in percentiles using the World Health Organisation height and weight for age percentile reference charts release in 2007 [33]. According to weight, underweight (<5th percentile for age), normal weight (5-< 95th percentile for age) or overweight (≥95th percentile for age) were distinguished. For height, short stature (< 5th percentile for age), normal height (5-< 95th percentile for age) or tall stature (≥95th percentile for age) were considered. BMI was estimated as weight (kg)/square height (m 2 ). It was stratified into underweight (<5th percentile for age and sex), normal weight (5-85th percentile for age and sex), overweight (85-95th percentile for age and sex) and obesity (≥95th percentile for age and sex). BP was either normal [systolic blood pressure (SBP) and diastolic blood pressure (DBP)] <90th percentile for sex, age and height), elevated (SBP and/or DBP ≥90-< 95th percentile for sex, age and height) whereas hypertension was defined as SBP and/or DBP ≥ 95th percentile for sex, age and height after three occasions according to the American Academy of Pediatrics guidelines to define BP categories and stages [32]. Dipstick proteinuria was defined by a persistent proteinuria (at least traces) after three measurements. The 24 h proteinuria was estimated from PCR and proteinuria corresponded to a PCR ≥ 200 mg/g. The Schwartz equation was used for estimate glomerular filtration rate (eGFR) [34]; it was either increased (≥120 ml/min/1.73 m 2 ), normal (90-< 120 ml/min/ 1.73 m 2 ) or decreased (< 90 ml/min/1.73 m 2 ).

Statistical analysis
Data were entered and coded using EPI info version 7.0 and analysed using Statistical Package for Social Science (SPSS) version 23

Maternal and birth history of participants
As presented in Table 2, the median (IQR) maternal age was 28 (23.2-32.0) years with no significant difference in BW (p = 0.486). Amongst the 10 (12.5%) of multiple pregnancies, 9 (90%) led to LBW infants with a statistical significance (p < 0.001). We observed that 3 (3.7%) women had diabetes mellitus among whom 2 (66.7%) delivered HBW children with a statistical significance (p = 0.048). In Table 3, the median (IQR) BW was 3200   Table 4). None of the HBW children had an elevated BP or hypertension. Hypertension and elevated BP were not significantly associated with BW (p = 0.233).  Effect of prematurity, weight for gestational age and overweight/obesity on blood pressure, proteinuria and glomerular filtration rate As presented in Table 6, GA did not significantly affect any of these parameters. When comparing with NBW, LBW had significantly higher SBP (p = 0.029); however, LBW and HBW children had a trend toward an increase DBP, PCR and eGFR with no statistical significance. We observed that overweight/obese children had significantly high SBP and DBP, and a reduced eGFR (all p < 0.037).   Estimates are median and interquartile range. AGA Appropriate for gestational age, DBP Diastolic blood pressure, eGFR Estimated glomerular filtration rate, HBW High birth weight, LBW Low birth weight, LGA Large for gestational age, NBW Normal birth weight, SBP Systolic blood pressure, SCr Serum creatinine, SGA Small for gestational age

Correlations between birth weight and variables
There was an inverse relationship between BW and SBP, DBP, PCR and eGFR although with no significant difference (all p > 0.05). However, we observed a significant weak positive correlation between BW and weight (r = 0.231) as well as BMI (r = 0.269), and a negative correlation between BW and weight percentile (r = − 0.241) (all p < 0.039) ( Table 7).

Discussion
This study revealed a high prevalence of elevated blood pressure and hypertension, observed respectively in NBW and LBW children only. Nearly one out of ten children had proteinuria which was significantly associated with LBW and HBW. More than one out of five children presented with a decreased eGFR associated with an increased prevalence with regards to BW. We noticed an inverse relationship between BW and renal markers (blood pressure, proteinuria and GFR), with no significant association. The reported high prevalence of hypertension as well as an elevated blood pressure were previously observed in this setting [30]. We did not find any association between BW and BP overall as well as with SBP or DBP as reported earlier [29,30]. However, we found that all children with hypertension were of the LBW group, which could be explained by the Brenner's hypothesis. Indeed, it supports a relationship between BW and the number of nephrons as well as the risk of developing hypertension [8,20,21]. Regarding the presence of hypertension in only LBW children, there is need for regular BP follow-up in this group of children.
Further, we observed that LBW and HBW children had significant proteinuria compared to NBW children, as reported elsewhere [10,12]. This could be related to hyperfiltration mechanisms related either to reduced glomerular number in LBW infants, the association between HBW and maternal diabetes or glomerular lesions such as podocytopenia and focal segmental glomerulosclerosis [10].
There was no significant association between eGFR and BW as reported elsewhere in children with similar age using creatinine [19,35,36]. However, the eGFR using cystatin C showed an inverse relationship with BW [36]. This could be explained by the fact that cystatin C reflects better the eGFR in comparison to creatinine and suggests its preferential use in LBW children. Nevertheless, higher GFRs turned to be more frequent in the LBW children compared to NBW and HBW. This could be explained by the fact that LBW is associated with a reduced number of nephrons and a low glomerular density at birth; glomerular enlargement volume, hyperfiltration, increased GFR and proteinuria occur during childhood and adolescence due compensatory maladaptive changes as consequences of compromised nephrogenesis; furthermore with advanced age favored by overweight/obesity, there will be a reduction in the GFR and worsening of proteinuria due to focal segemental glomerulosclerosis lesions [1][2][3]37].
We observed an inverse relationship between BW and renal markers without any statistical significance as previously observed [21,29]. However, studies in SSA reported a significant inverse relationship between BW and BP [20,25]. Meanwhile, some other studies showed a positive correlation between BW and SBP at birth [24]. The difference of correlation between renal markers and BW observed in this study compared to others could be related to the difference in the age of participants and the method of renal markers measurement.

Strengths and limitations
The main limitations of this study were the estimation of GFR with creatinine instead of cystatin C which reflects better, the kidney function. Furthermore, the small sample size as well as the number of events might have an influence on possible lack of associations. Nevertheless, this study is the first in Central Africa, to the very best of our knowledge, to assess the relationship between BW and overall renal markers in childhood. We used up-to-date guidelines for blood pressure and proteinuria diagnosis in children [32,38]. These findings contribute to enrich data on renal markers and BW in the sub-Saharan Africa setting and suggest further research using cystatin C-based equations for eGFR.