Keywords
low birth weight - cardiovascular disease risk factors - anthropometry - low socioeconomic
status - adolescence
Palavras-chave
baixo peso ao nascimento - risco cardiovascular - antropometria - baixo nível socioeconômico
- adolescência
Introduction
Over the past two decades, an extensive and growing literature has investigated the
associations between low birth weight (LBW) and chronic diseases. Several clinical
studies demonstrate that LBW is an important risk factor for atherosclerosis, type
2 diabetes, hypertension, metabolic syndrome and endothelial dysfunction.[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8] Many of those studies were conducted in countries with high average socioeconomic
status and the majority demonstrated a strong association between LBW and increased
risk of cardiovascular disease.
Nevertheless, many other studies have only found a positive or a non-association between
LBW and cardiovascular disease risk factors in young men and women or in children
and adolescents. Moreover, there is convincing evidence that postnatal weight gain
may have a greater influence on blood pressure than birth weight.[4]
[9]
[10]
[11]
There are not many studies of cardiovascular risk factors for LBW pediatric populations
in the Brazilian northeast.[12]
[13] This region has the lowest average socioeconomic status and the highest under-five
mortality rates when compared with the other regions of the country.[14] In recent years, the Brazilian government has invested heavily in programs in the
northeast to ensure better child nutrition. However, these increases in the population's
income have not been reflected in health indicators, and the prevalence of overweightness
and obesity (important predictors for several cardiovascular disease risk factors)
is rising.[15]
The hypothesis of this study is that LBW adolescents in low average socioeconomic
status regions have the poorest health status. Thus, the research was conducted to
investigate whether LBW is associated with changes during adolescence in anthropometric
and biochemical risk factors, in a low average socioeconomic status population in
the Brazilian northeast.
Methods
We conducted a retrospective cohort study with male and female participants with ages
varying from 10 to 20 years, born from 1992 to 2002, with LBW (less than 2,500 g at
birth) or normal birth weight (NBW) (≥ 2,500 g) and who were born and lived in Fortaleza,
Brazil. This study was approved by The Research Ethics Committee of the Assis Chateaubriand
Maternity Teaching Hospital and conducted between February and August 2013. A similar
sample size was applied in previous studies which assessed the endothelial and metabolic
disorders and fetal growth restriction in adolescents.[16]
[17]
Hospital birth records were used to check perinatal conditions and select potential
subjects. Eligibility criteria included being healthy at the moment of the evaluation
and having no cardiovascular and/or endocrine conditions or relevant cardiovascular
disease risk factors. Participants were contacted through personal visits, phone calls
and letters.
A total of 101 LBW adolescents were located; however, the mothers of 15 of these adolescents
refused to sign the consent for the adolescent's participation in the study. As for
the NBW group, 102 adolescents were located, with the mothers of 16 of them refusing
to sign the consent.
The evaluation consisted of an interview to investigate medical history and familial
occurrence of cardiovascular disease risk factors, and to conduct laboratorial exams
and anthropometric and clinical measurements.
Weight and total body fat percentage, with the participants wearing light clothes
and no shoes, were measured to the nearest 0.1 kg / 0.1% using a calibrated model
W835 body analyzer (Wanchai-Ho, China). A wall-mounted stadiometer Seca (model 206,
Hamburg, Germany) to the nearest 0.1 cm was used to measure standing height. With
these measurements the body mass index (BMI) was calculated (in kilograms per square
meters) and evaluated according to age and gender, based on World Health Organization
(WHO) references.[18]
An anthropometric tape measure Wiso (model T872) was used to measure the circumference
of the waist, abdomen and hips. The measurements were taken at the end of a gentle
expiration, with the reference point halfway between the lower rib and the top of
the iliac crest; the umbilical scar and the largest point of the outer hip, respectively.
Other two measurements were waist/hip ratio and biceps skinfold were also measured.
To measure blood pressure, a Microlife calibrated semi-automatic sphygmomanometer
(model BP 3BTO-H, Dunedin, USA) was used, after a 30-minute rest, and giving a 1-minute
interval between the two measurements. If the difference between the first and second
measurements was ≥ 20 mm Hg, a new one would be taken, and the average of the two
closest measurements was used as the final result. The Fourth Report on the Diagnosis,
Evaluation, and Treatment of High Blood Pressure in Children and Adolescents was used
as a reference for blood pressure according to age and gender.[19]
For laboratorial evaluation, fasting blood sugar test samples were collected, as well
as samples to measure total cholesterol, high-density lipoprotein cholesterol and
triglycerides. Low density lipoprotein cholesterol was calculated when triglycerides < 400
mg/dL.[20] The I Guideline for Prevention of Atherosclerosis in Childhood and Adolescence was
adopted as reference value.[10] Each participant received their laboratory results and a cardiovascular report.
Statistical Analysis
Mean and standard deviations were calculated to all normality variables using Student's
t test, and proportions were compared by Chi-square (χ2) or Fisher exact tests. Kruskall–Wallis and Mann–Whitney U tests were used for non-normal
distribution variables and described as quartiles: first quartile (Q1), median (M)
and third quartiles (Q3). Some variables were converted to Z-score by using the formula:
(x − mean)/standard deviation. Analysis were performed using Stata program (College
Station, TX, USA) version 12.0 for Mac and statistically significant if p < 0.05. An alpha error of 5% and a beta error of 20% were adopted to calculate the
sample size according to a relative risk of 2.3 to hypertension in preterm and low
birth weight subjects.[21]
Results
Evaluations were obtained from 86 LBW and 86 NBW participants. The sample consisted
of 56 females (65.1%) in the LBW group and 52 females (60.5%) in the NBW group (p = 0.5), with an average age of 12.8 years for LBW (SD = 2.2) and 13.6 years for NBW
(SD = 2.9) (p = 0.2). Of the total of 58 adolescents born small for gestational age, 49 (57.0%)
were also LBW (p < 0.05).
The maternal characteristics were similar, except for the mode of delivery (p < 0.05) ([Table 1]). No significant differences were identified regarding age, gender distribution,
or current anthropometric data (weight, height, abdominal circumference, BMI, classification
according to BMI, body fat mass and biceps skinfold thickness) ([Table 2]). The clinical parameters were similar between the groups. Systolic blood pressure
and diastolic blood pressure levels were also similar in the LBW group (systolic blood
pressure = 2.6 mm Hg; diastolic blood pressure = 1.5 mm Hg). Subjects in the LBW group
tended to have lower height (p = 0.1). No differences were found in the family history of cardiovascular diseases
(p = 0.1), family incomes (p = 0.7) or maternal schooling (p = 0.8). We also expressed anthropometric variables as mean Z-scores. No significant
differences were found. All the values were expressed as mean and standard deviations
([Table 2]).
Table 1
Maternal characteristics during childbirth according offspring's birth weight
|
LBW
Q1 (M) Q3
|
NBW
Q1 (M) Q3
|
p
|
|
Age (years)
|
18.0 (24.5) 30.0
|
20.0 (23.0) 28.0
|
0.5*
|
|
Pregnancy complications (n,%)
|
|
|
|
|
|
|
Preeclampsia
|
32
|
37.2
|
23
|
26.7
|
< 0.001†
|
|
Diabetes
|
2
|
2.3
|
0
|
0.0
|
|
|
Other diseases
|
43
|
49.9
|
13
|
15.2
|
|
|
Mode of delivery (n,%)
|
|
|
|
|
|
|
Vaginal
|
45
|
52.3
|
32
|
37.2
|
< 0.05§
|
|
Cesarean
|
41
|
47.7
|
54
|
62.8
|
|
|
Number of pregnancies
|
0.8 ± 1.2
|
1.4 ± 2.7
|
0.2*
|
|
Number of births
|
0.6 ± 0.9
|
0.9 ± 1.8
|
0.5*
|
|
Number of abortions
|
0.2 ± 0.5
|
0.5 ± 1.3
|
0.3*
|
|
Complications in other pregnancies (n,%)
|
22
|
25.6
|
20
|
23.3
|
0.4§
|
|
Smoking (n,%)
|
2
|
2.4
|
8
|
9.6
|
0.1§
|
Abbreviations: LBW, low birth weight; M, median; NBW, normal birth weight; Q1, first
quartile; Q3, third quartile.
Note: Values are median and interquartile range for continuous variables; number of
participants and percentages are presented for categorical variables however.
*Mann–Whitney U test, †Fisher's exact, §Chi-square (χ2) test.
Table 2
Perinatal, anthropometric, and clinical current data according to birth weight groups
|
LBW
Q1 (M) Q3
|
NBW
Q1 (M) Q3
|
p
|
|
Gender
|
|
|
|
|
Boys (n,%)
|
30
|
34.9
|
34
|
39.5
|
0.5†
|
|
Girls (n,%)
|
56
|
65.1
|
52
|
60.5
|
|
|
SGA (n,%)
|
49
|
57.0
|
9
|
10.5
|
< 0.001†
|
|
Current data
|
|
|
|
|
Age (years)
|
11.0 (12.0) 14.0
|
11.0 (13.0) 16.0
|
0.2*
|
|
Weight (kg)
|
40.2 (47.7) 56.7
|
40.6 (49) 58.5
|
0.8*
|
|
Z-score weight
|
-0.01 ± 1.0
|
0.01 ± 1.0
|
0.8§
|
|
Height (m)
|
1.5 (1.5) 1.6
|
1.5 (1.6) 1.6
|
0.1*
|
|
Z-score height
|
-0.1 ± 1.0
|
0.1 ± 1.0
|
0.1§
|
|
Abdominal circumference (m)
|
0.6 (0.7) 0.8
|
0.6 (0.7) 0.8
|
0.6*
|
|
Z-score abdominal circumference
|
0.02 ± 1.0
|
-0.02 ± 1.0
|
0.6§
|
|
Hip circumference (m)
|
0.6 (0.7) 0.8
|
0.6 (0.7) 0.7
|
0.8§
|
|
Z-score hip circumference
|
0.03 ± 1.0
|
-0.03 ± 1.0
|
0.5§
|
|
Waist/hip ratio
|
0.8 (0.8) 0.8
|
0.7 (0.8) 0.8
|
0.2*
|
|
Z-score waist/hip ratio
|
0.1 ± 0.1
|
-0.1 ± 0.1
|
0.1§
|
|
BMI (mean)
|
17.6 (20.1) 23.9
|
17.5 (19.9) 23.1
|
0.8*
|
|
Z-score BMI
|
0.46 ± 1.0
|
-0.46 ± 1.0
|
0.5§
|
|
Classification according to BMI
|
|
|
|
|
Overweightness/obesity (n,%)
|
23
|
26.8
|
23
|
26.8
|
0.5†
|
|
Normal (n,%)
|
54
|
62.8
|
55
|
64.0
|
|
|
Body fat mass (%)
|
20.9 (26.5) 35.2
|
20.7 (25.2) 33.3
|
0.4*
|
|
Z-score body fat mass
|
0.1 ± 0.1
|
-0.1 ± 0.1
|
0.4§
|
|
Biceps skinfold thickness (m)
|
0.005 (0.007) 0.011
|
0.005 (0.007) 0.009
|
0.6*
|
|
Z-score biceps skinfold thickness
|
0.05 ± 1.0
|
-0.05 ± 1.0
|
0.6§
|
|
SBP (mmHg)
|
97 (108) 110
|
90 (100) 108
|
0.1*
|
|
Z-score SBP
|
0.1 ± 0.1
|
-0.1 ± 0.1
|
0.1§
|
|
PAS ≥ 95th percentile
|
9
|
10.5
|
9
|
10.5
|
0.9†
|
|
DBP (mmHg)
|
60 (65) 70
|
58 (63.5) 67
|
0.1*
|
|
Z-score DBP
|
0.1 ± 1.1
|
-1.1 ± 0.8
|
0.1§
|
|
PAD ≥ 95th percentile
|
1
|
1.2
|
0
|
0
|
0.1†
|
|
Family history of CVDʃ
|
|
|
|
|
Yes (n,%)
|
83
|
96.5
|
78
|
90.7
|
0.1†
|
|
No (n,%)
|
3
|
3.5
|
8
|
9.3
|
|
|
Menarche age
|
11 (11) 12
|
11 (12) 13
|
0.2*
|
|
Familiar incomes (MW)
|
1 (1.5) 2
|
1 (1) 2
|
0.8*
|
|
Maternal school education (years)
|
5 (8.5) 11
|
5 (8) 11
|
0.8*
|
Abbreviations: BMI, body mass index (in kilograms per square meters); CVD, cardiovascular
disease; DBP, diastolic blood pressure; LBW, low birth weight; M, median; MW, minimum
wage; NBW, normal birth weight; Q1, first quartile; Q3, third quartile; SBP, systolic
blood pressure; SGA, small for gestation age.
Note: Values are median and interquartile range for continuous variables; number of
participants and percentages are presented for categorical variables however.
*Mann–Whitney U test, †Chi-square test (χ2) test, §Student's t-test.
The data from the biochemical evaluation showed very similar results between LBW and
NBW participants. Although low-density lipoprotein cholesterol, very low-density lipoprotein
cholesterol and triglycerides levels were higher in the LBW group, there was no significant
difference. Similarly, even though total cholesterol, high-density lipoprotein cholesterol
and glucose levels were higher in the NBW group, no significant differences were observed
([Table 3]).
Table 3
Blood lipids and insulin profile according to birth weight groups
|
LBW
|
NBW
|
p*
|
|
Glucose (mg/dL)
|
79.7 (9.6)
|
82.3 (9.5)
|
0.2
|
|
TC (mg/dL)
|
143.0 (24.4)
|
143.4 (26.2)
|
0.9
|
|
HDL-C (mg/dL)
|
43.4 (10.2)
|
44.3 (10.1)
|
0.5
|
|
LDL-C (mg/dL)
|
84.6 (21.1)
|
84.4 (23.6)
|
0.7
|
|
VLDL (mg/dL)
|
14.6 (6.5)
|
13.9 (7.1)
|
0.4
|
|
TG (mg/dL)
|
73.0 (32.6)
|
70.9 (30.0)
|
0.5
|
Abbreviations: HDL-C, high-density lipoprotein cholesterol; LBW, low birth weight;
LDL-C, low-density lipoprotein cholesterol; NBW, normal birth weight; TC, total cholesterol;
TG, triglycerides; VLDL-C, very low-density lipoprotein cholesterol.
Note: Values are mean and standard variation.
*Mann–Whitney U test.
Because more than 50% of the eligible patients were lost, we performed some statistical
tests between participants and non-participants. LBW participants and non-participants
had similar performances regarding perinatal characteristics such as maternal age
(p = 0.4), number of maternal gestations and abortions (p = 0.5; p = 0.8 respectively), maternal preeclampsia and diabetes (p = 0.5; p = 0.2 respectively), age (the average age of participants was 13, and the average
age of non-participants was 12; p < 0.05), gender (p = 0.7), prematurity (p = 0.1), small size for gestational age (higher frequency of adolescents born small
for gestational age in the participant's group, p < 0.05).
Discussion
The findings presented suggest that low birth weight was not associated with poor
health outcomes among adolescents in the Brazilian northeast. The differences were
non-significant for all risk factors except for blood pressure, where a borderline
significant difference was observed. The adolescents in the LBW group had higher blood
pressures (2.6 mm Hg for systolic blood pressure and 1.5 mm Hg for diastolic blood
pressure), and they tended to have lower height than adolescents in the NBW group.
These three variables showed the highest difference in the anthropometric data and
tended to be higher in the LBW group, but were not significant. Although we selected
subjects born on the same day, from 1992 to 2002, the groups had a different quartile
of age. This occurred because we couldn't localize the adolescents born exactly on
the same day. However, this didn't affect the results.
One of the strongest predictors of hypertension in adulthood is a high level of blood
pressure in childhood.[22] Another important risk factor is gender, which may influence the probability of
elevated blood pressure during adulthood.[23] Among women, the risk ratio ranged from 2.6 to 5.7, while among men it ranged from
2.3 to 4.3. However, some studies have found opposite results. A longitudinal prospective
study with 250 subjects with ages between 11 and 14 years found no correlation between
birth weight and blood pressure, weight or BMI. The authors suggested there are risk
factors more important than LBW or gestational age that are related to increased blood
pressure in childhood, such as high maternal BMI and a high birth weight (HBW).[24]
In Iceland, a study[25] investigated 857 children with ages between 9 and 10 years, 51.9% of which were
girls. They found higher blood pressure levels among the boys, but no correlation
was found between birth weight and absolute blood pressure values. Gestational age
also did not correlate with blood pressure. Studies with other populations had similar
results in the US,[26] China,[9] England,[4] Italy[27] and Brazil.[28]
Despite socioeconomic factors such as family incomes and maternal education being
comparable, the LBW group did not have an increased risk of cardiovascular disease.
Studies show that women who deliver LBW infants take meticulous care of them and pay
closer attention to medical advices.[29] This could be the reason for less cardiovascular disease risk factors associated
with catch-up growth,[3] a determinant factor for cardiovascular diseases in LBW newborns. On the other hand,
Fortaleza currently ranks as the ninth worst Brazilian capital on the Municipal Human
Development Index.[30] This index encompasses longevity, education and income. Thus, in this study NBW
newborns were also exposed to poor conditions, which may explain why their results
were similar to those of LBW newborns, even though the groups had a similar family
history of cardiovascular diseases.
Countries have some peculiarities regarding socioeconomic status. It has been estimated
that in regions in the world with low average socioeconomic status more than 200 million
children less than 5 years old are not fulfilling their development potential. This
fact can be the result of poverty, nutritional deficiencies and inadequate learning
opportunities.[31] In Brazil specifically, there have been fast changes in major social health determinants
and in the organization of health services over the past three decades. The main changes
during this period were economic growth, reduction in income inequality, urbanization,
improvements in the education of women, decreased fertility rates, a government cash
transfer program and improvements in the provision of water and sanitation.[32]
Furthermore, since the 1970's a nutritional transition is occurring in Brazil, involving
a decline in malnutrition in children and an increase in obesity and overweightness
in adults. It is an apparent paradox that the highest frequencies of normal anthropometric
measurements in Brazilian adults were found in the poorer northern and northeastern
regions of Brazil.[33] However, a recent study showed that cardiovascular disease increased in these regions.[34]
The results contribute toward testing the validity of the Barker hypothesis in countries
with low average socioeconomic status, especially in the poorest region of Brazil.
An important feature in this study is that socioeconomic status and the same day of
birth were accounted for and the groups were similar in average chronological age,
gender and time frame for the menarche. We used international parameters to facilitate
future comparisons with our data.
Some study limitations need to be addressed. Firstly, the neonatal data obtained from
the hospital records had only been collected prior to the present study for other
purposes and a long time ago. For that reason, some data was missing. Secondly, adolescents
with birth weight between 2,500 and 3,000 g (n = 25) were included in the NBW group; however, this group could include some subjects
who did not reach their potential for intrauterine growth, which lead to confusing
results. Although some authors categorized birth weight in a manner similar to this
research, they had different results.[35] Thirdly, complete information regarding pubertal status was not available, so bias
may have been introduced by giving all participants the same pubertal status. Fourthly,
we can't increase our sample. As we have described, there were some difficulties in
obtaining the hospital records that have maternal and neonatal data. Despite this,
we got more than one thousand records. Unfortunately, we were unable find most of
the patients. However, as we have shown, participants and non-participants had similar
characteristics, and non-participants were not the reason behind the few differences
between the groups that we found. Fifthly, we based our sample size calculation on
hypertension; however a few number of subjects had high levels of blood pressure.
This could be the reason why no other differences were found.
Summarizing, LBW did not increase cardiovascular disease risk factors in young adults
in the second decade of life. The absence of an association between LBW and poor health
outcomes among adolescents in a low average socioeconomic status population from a
capital in the Brazilian northeast corroborates previous findings in other countries
with low average socioeconomic status. We suggest future prospective studies with
more subjects to investigate this association.
