Temporal association of maternal weight gain with early-term and preterm birth and low birth weight babies

Abstract

Background:

Low birth weight and preterm or early-term babies may have a higher risk of poor health. One of the main factors is the weight gain of a pregnant woman during gestational weeks in the second and third trimesters. Changes in weight over a month in a pregnant woman might also have an impact on infant outcomes. This study aimed to investigate the association between maternal weight at different time points and low birth weight and preterm or early-term babies (premature babies).

Methods:

A total of 156 pregnant women were recruited. Maternal weight was collected at different gestational weeks. Maternal age, body mass index, delivery mode, delivery week, and infant weight were also recorded. Maternal data were restructured into a person-period format before mixed-effects multiple logistic regression was used. Various weight variables with either a fixed effect or time-varying effects were tested in the model.

Results:

Thirty (19.23%) women had delivered low birth weight or premature babies. Multiple logistic regression model demonstrated that mothers with higher increases in weight at 32 weeks of gestation than that in the predelivery stage had a lower probability of having a low birth weight or premature baby (odds ratio [OR] = 0.64; 95% CI, 0.49-0.85; p < 0.001). Women with a weight increase of more than 2 kg in a 4-week gestation period had a higher probability of having a low birth weight or premature baby than those with an increment of <1 kg (OR = 8.43; 95% CI, 2.90-24.54; p < 0.001).

Conclusion:

An increase in weight gain after 32 weeks was shown to reduce the risk of low birth weight and premature babies. Maternal weight monitoring was suggested to be conducted every 4 weeks to minimize the chance of having a low birth weight and premature baby.

1. INTRODUCTION

According to the Maternal and Child Health 2016, the percentages of low birth weight (<2.5 kg) and preterm birth (before 37 weeks) in Malaysia were 9.7% and 6.7%, respectively.¹ Conversely, the percentage of low birth weight in Malaysia varied among previously published studies, reported at 6.50%, 11.08%, 12.60%, 13.50%, 13.96%, and 20.80%.^2–7 Similar phenomena were also reported for the prevalence of preterm birth, noted at 5.38%,⁸ 9.06%,² and 25.40%.^2,3,8

Literature suggests that birth weight and gestational age are two crucial factors that could affect infant outcomes. Birth weight was negatively associated with the risk of preterm or early-term babies (37–38 weeks).⁹ Full-term babies (6.4%) had lower incidence of low birth weight than that in preterm babies (12.14%).² Undeniably, low birth weight and preterm babies could also have a higher risk of mortality, poor health conditions, and development problems, as well as the probability of child diseases.^4,10–12 Emerging concern regarding early-term birth globally is that these babies had been proved to have a high risk of poor cardiorespiratory fitness, high request of oxygen therapy, poor cognitive outcome, hypoglycemia, neonatal death, neonatal intensive care unit (NICU) admission, and respiratory morbidities.^10,11,13,14 Therefore, good strategies and policies to reduce the prevalence of low birth weight and preterm or early-term babies are required.

To date, weight at the prepregnancy stage and during gestational weeks has been reported to be negatively associated with having a low birth weight and preterm baby.^9,15,16 Women with a lower body mass index (BMI) at the prepregnancy stage, as well as low gestational weight gain, are more likely to have a higher probability of low birth weight and preterm baby.⁹ The Institute of Medicine Weight Gain Recommendation for Pregnancy recommends a total weight gain during gestational weeks and weight gain in the second and third trimesters according to different prepregnancy weight categories,^15,16 suggesting the importance of total weight gain in different trimesters to reduce the risk of low birth weight and preterm birth. Nonetheless, the changes in weight over a month in a pregnant woman might also impact pregnancy and indirectly increase the risk of having a low birth weight or preterm baby. Therefore, this study aimed to investigate the maternal weight of pregnant women associated with low birth weight and preterm or early-term babies (premature baby: ≥37 weeks).

2. METHODS

2.1. Data collection

Maternal data of 176 pregnant women were used in this study, where the participants were followed by the researchers from the first antenatal visit until delivery. Data were obtained from the Universiti Sains Malaysia Pregnancy Cohort Study, with previously published details.^17–19 Generally, the Universiti Sains Malaysia Pregnancy Cohort Study is led by an expert in the field of maternal nutrition, and the research team consists of obstetricians, pediatricians, biochemists, and statisticians. This cohort study aimed to follow up pregnant mothers from early pregnancy until the first year after delivery. Pregnant women were recruited using the purposive sampling method from the Obstetrics and Gynecology Clinic of Universiti Sains Malaysia Hospital [Universiti Sains Malaysia Hospital (HUSM)] and the Health Clinic of Kubang Kerian District in Kelantan, Malaysia. Inclusion criteria were Malay citizens with singleton pregnancy, aged 19–40 years, who were at <24 weeks of gestation, who planned to deliver in HUSM, and who lived within a 50-km distance from HUSM. Patients with preexisting chronic diseases, such as diabetes mellitus, hypertension, or heart disease, and those with a history of preterm delivery before 37 weeks of gestation were excluded. Ethical approval was obtained from the Human Research Ethics Committee of Universiti Sains Malaysia. Written consent was obtained from all participants who were included in the study.

Serial measurements of maternal weight of pregnant women were collected, including self-reported prepregnancy weight, as well as the measured weight at the first (average, 14 weeks of gestation), second (average, 18 weeks of gestation), and third visit (average, 32 weeks of gestation), and the last antenatal weight within 4 weeks before delivery. Prepregnancy BMI was also calculated. Maternal age, delivery mode, delivery week, and infant weight were recorded.

2.2. Statistical analysis

Descriptive analysis was conducted using IBM SPSS Statistics version 25 (SPSS Inc., Chicago, IL). Missing data and errors in data were checked. Twenty participants with no maternal weight data on either of the selected time points were excluded from the analysis. A total of 156 participants with complete maternal weight data were used for the modeling.

Data were restructured to a person-period format for each person,²⁰ with one record per time interval (4 weeks), as presented in Table 1. The number of records per participant depended on the gestational week, where each person had 9 to 11 time intervals. The binary outcome indicator was the delivered baby with birth weight <2.5 kg (≤2.49 kg) or premature baby (preterm and early term) with delivery ≤37 weeks (coded 1), depending on the outcome at the week of delivery (Table 1). In Table 1, women with identity document 1 (ID1) had delivered a baby with birth weight >2.5 kg or at >38 weeks of gestation; therefore, “0” was coded for all time intervals. Conversely, women of ID2 had delivered baby with birth weight <2.5 kg (2.49 kg and below) or premature baby (preterm and early term) with delivery at or <37 weeks at the ninth time interval; therefore, “1” was coded at the ninth time interval.

Table 1.

Person-period format data

Obs	ID	Time interval	Weeks	Week of gestational	Maternal weight, kg	Outcome
1	1	0	0	37	53.00	0
2	1	1	4	37	51.77	0
3	1	2	8	37	50.55	0
4	1	3	12	37	49.32	0
5	1	4	16	37	48.40	0
6	1	5	20	37	49.54	0
7	1	6	24	37	51.83	0
8	1	7	28	37	54.11	0
9	1	8	32	37	56.4	0
10	1	9	36	37	56.88	0
11	2	0	0	38	40.00	0
12	2	1	4	38	40.31	0
13	2	2	8	38	40.62	0
14	2	3	12	38	40.92	0
15	2	4	16	38	42.56	0
16	2	5	20	38	44.23	0
17	2	6	24	38	45.49	0
18	2	7	28	38	46.74	0
19	2	8	32	38	48.00	0
20	2	9	36	38	49.33	1

The maternal weight for each interval was estimated using the collected maternal weight at antenatal visits (Supplementary Appendix S1, http://links.lww.com/JCMA/A79).

A mixed-effect multilevel logistic regression model was set up to investigate the effect of maternal weight (monthly record) associated with the probability of having a low birth weight or premature birth.

A mixed-effect multilevel logistic regression model with two levels, the time interval nested within each woman, was calculated as follows:

where y_ij is the binary outcome with logit transformation, α_ij is the intercept of the equation, and i and j represent the ith time interval of the jth women, respectively. β₁ is the coefficient of variables X_ij at the first level (interval), β₂ is the coefficient of variables X_j at the second level (women), and μ_j is the random effect of the women’s level, with the variation σ².

Various weight variables with either fixed effects or time-varying effects were investigated in the model, including the following:

Fixed-effect variables (constant value throughout the study):

1. Weight and BMI at prepregnancy stage;

2. Weight at 12, 16, 20, 24, 28, 32, and 36 weeks and predelivery stage;

3. Changes in weight during 12, 16, 20, 24, 28, 32, and 36 weeks compared to the weight at the prepregnancy stage;

4. Changes in weight at prepregnancy stage and 12, 16, 20, 24, 28, 32, and 36 weeks compared to weight at the predelivery stage.

Time-varying effect variables (different values at each time interval):

1. Weight and BMI at each time interval (4 weeks of interval);

2. Changes in weight at each time interval compared to the prepregnancy stage;

3. Changes in weight at each time interval compared with the previous time interval

Other variables, including pregnant mother characteristics such as their age, marital status, educational level, occupational status, monthly household income, parity, smoking status, prepregnancy health issues, and delivery mode, were also tested in the model. Significant variables were then retested in the analysis.

The model was set up using MLwiN version 2.25 (Centre for Multilevel Modelling, University of Bristol, Bristol, UK), which was estimated using quasilikelihood method. Variables with p < 0.05 were retained in the final model. Possible confounders, including pregnant women characteristics, multicollinearity between variables, and interaction terms between variables, were checked. The final model was then estimated using the Markov chain Monte Carlo (MCMC) method with 500 000 interactions and a burn-in of 50 000. Chain mixing and stability were checked according to the MCMC trajectories and diagnostic window.

3. RESULTS

A total of 1567 time intervals created by 156 pregnant women were used in the analysis. The median gestational week was 39 (36–42) weeks. A total of 30 (19.23%) women delivered low birth weight or premature babies. Of the 30 women, 23 and 7 delivered low birth weight or premature babies before 38 weeks and between 38 and 42 weeks, respectively. The probability of having a low birth weight or premature baby and its cumulative probability are presented in Table 2 and Fig. 1, respectively.

Table 2.

Probability of having low birth weight or premature baby among 156 pregnant women

Time interval [wk)a	Total woman	Woman experienced outcomeb	Censoredc	Probability of having outcome
[36,37)	156	3	0	0.019
[37,38)	153	20	0	0.140
[38,39)	133	3	38	0.027
[39,40)	92	3	50	0.046
[40,41)	39	0	24	0.000
[41,42)	15	1	13	0.125
[42,43)	1	0	1	0.000

^a[) included within time.

^bOutcome: low birth weight or premature baby.

^cCensored: pregnant women leave the study without delivered low birth weight or premature baby.

Fig. 1.

Cumulative probability of having low birth weight or premature baby among 156 pregnant women.

Table 3 shows the distribution of pregnant women’s characteristics. The mean ± standard deviation of age for pregnant women was 29.93 ± 4.64. Majority were married (90.38%), worked (80.13%), never smoked (98.08%), and had no health issues (87.82%) before pregnancy.

Table 3.

Characteristics of pregnant women

Variable	Freq.	%
Age (mean ± standard deviation)	29.93 ± 4.64
Marital status
Married (staying with husband)	141	90.38
Others	15	9.62
Employment (mother)
Unemployed	31	19.87
Working	125	80.13
Employment (father, n = 155)
Unemployed	2	1.29
Working	153	98.71
Monthly household income [Ringgit Malaysia (RM)] (n = 155)
Low (<RM 2300)	61	39.35
Middle (RM 2300-RM 5599)	76	49.03
High (>RM 5600)	18	11.61
Education level (mother)
Primary/secondary school	52	33.33
Diploma	63	40.38
Degree and above	41	26.28
Education level (father, n = 155)
Primary/secondary school	63	40.65
Diploma	53	34.19
Degree and above	39	25.16
Smoking status (mother)
Never	153	98.08
Stop for more than 6 mo	3	1.92
Smoking status (father, n = 152)
Never	65	42.76
Stop for more than 6 mo	16	10.53
Smoker	71	46.71
Parity
0	42	26.92
1	46	29.49
≥2	68	43.59
Health status at prepregnancy
None	137	87.82
Gestational diabetes mellitus	5	3.21
Pregnancy-induced hypertension	3	1.92
Anemia	10	6.41
Others	1	0.64
Health status at prepregnancy (group)
No	137	87.82
Yes	19	12.18
Weight at prepregnancy, kg (mean ± standard deviation)	55.74 ± 11.35
BMI at prepregnancy, kg/m2
Thin (<18.50)	23	16.74
Normal (18.50-24.99)	85	54.49
Obese (≥25.00)	48	30.77
Labor method
Spontaneous vaginal delivery	131	83.97
Cesarean delivery	25	16.03

BMI = body mass index; RM = Ringgit Malaysia.

3.1. Univariable analysis

A list of fixed and time-varying effect of weight and BMI variables, as described in the methodology, was investigated using univariable logistic analysis (Table 4). There were four significant variables in this analysis, including one weight variable with a fixed effect (changes in weight at 32 weeks of gestation compared to the predelivery stage), two weight variables with a time-varying effect (weight at each interval and changes in weight at each interval compared to the previous interval), and the mode of delivery (p < 0.05).

Table 4.

Univariable logistic analysis of having low birth weight or premature baby among pregnant women

Variables	Freq.	Coefficient	SE	OR	95% CI		p
Fixed effect
Weight at prepregnancy, kg	1567	−0.006	0.02	0.99	0.96	1.03	0.72
BMI at prepregnancy, kg/m2
Thin (<18.50)	230	Ref.
Normal (18.50-24.99)	851	−0.22	0.52	0.81	0.29	2.25	0.68
Obese (≥25.00)	486	−0.06	0.55	0.95	0.32	2.80	0.92
Weight at 4 wk	1567	−0.005	0.02	1.00	0.96	1.03	0.77
Weight at 8 wk	1567	−0.005	0.02	1.00	0.96	1.03	0.75
Weight at 12 wk	1567	−0.004	0.02	1.00	0.97	1.03	0.80
Weight at 16 wk	1567	−0.003	0.02	1.00	0.97	1.03	0.85
Weight at 20 wk	1567	−0.003	0.02	1.00	0.97	1.03	0.85
Weight at 24 wk	1567	−0.002	0.02	1.00	0.97	1.03	0.90
Weight at 28 wk	1567	−0.001	0.02	1.00	0.97	1.03	0.95
Weight at 32 wk	1567	−0.001	0.02	1.00	0.97	1.03	0.95
Weight at 36 wk	1553	−0.02	0.02	0.99	0.94	1.03	0.50
Weigh at predelivery	1567	−0.006	0.02	0.99	0.96	1.03	0.71
Changes of weight at 12 wk vs prepregnancy	1567	0.04	0.07	1.04	0.91	1.18	0.59
Changes of weight at 16 wk vs prepregnancy	1567	0.04	0.06	1.04	0.93	1.17	0.52
Changes of weight at 20 wk vs prepregnancy	1567	0.03	0.05	1.03	0.93	1.14	0.62
Changes of weight at 24 wk vs prepregnancy	1567	0.04	0.05	1.04	0.94	1.14	0.44
Changes of weight at 28 wk vs prepregnancy	1567	0.03	0.04	1.03	0.95	1.13	0.43
Changes of weight at 32 wk vs prepregnancy	1567	0.03	0.04	1.03	0.95	1.11	0.46
Changes of weight at 36 wk vs prepregnancy	1553	−0.03	0.05	0.97	0.89	1.06	0.51
Changes of weight at prepregnancy vs predelivery	1567	−0.001	0.03	1.00	0.94	1.07	0.98
Changes of weight at 12 wk vs prepredelivery	1567	−0.02	0.04	0.98	0.90	1.07	0.71
Changes of weight at 16 wk vs prepredelivery	1567	−0.03	0.05	0.98	0.89	1.07	0.59
Changes of weight at 20 wk vs prepredelivery	1567	−0.03	0.05	0.97	0.88	1.08	0.61
Changes of weight at 24 wk vs prepredelivery	1567	−0.07	0.07	0.93	0.82	1.06	0.30
Changes of weight at 28 wk vs prepredelivery	1567	−0.14	0.09	0.87	0.73	1.04	0.14
Changes of weight at 32 wk vs prepredelivery	1567	−0.26	0.13	0.77	0.60	0.99	0.04*
Changes of weight at 36 wk vs prepredelivery	1553	−0.48	0.28	0.62	0.36	1.06	0.08
Time varying effect
Weight at each interval (4 wk)	1567	0.03	0.01	1.03	1.01	1.06	0.01*
Changes of weight at each interval vs previous interval (t, t − 1)	1411	0.61	0.15	1.85	1.39	2.46	<0.001*
Changes of weight at each interval vs previous interval (t, t − 1)
≤1.00 kg	634	Ref
1.01-2.00 kg	480	0.90	0.51	2.45	0.90	6.68	0.08
>2.00 kg	297	1.57	0.50	4.79	1.80	12.74	<0.001*
Fixed effect
Age of women	1567	0.01	0.04	1.01	0.74	1.09	0.94
Marital status
Married (staying with husband)	1415	Ref
Others	152	0.03	0.61	1.03	0.31	3.45	0.96
Employment (mother)
Unemployed	312	Ref
Working	1255	−0.01	0.46	0.99	0.40	2.45	0.99
Monthly household income (Ringgit Malaysia) (n = 155)
Low (<RM 2300)	625	Ref
Middle (RM 2300-RM 5599)	764	0.21	0.41	1.23	0.55	2.76	0.61
High (>RM 5600)	178	0.58	0.54	1.78	0.61	5.16	0.29
Education level (mother)
Primary/secondary school	520	Ref
Diploma	633	−0.12	0.41	0.89	0.40	1.96	0.77
Degree and above	414	−0.66	0.54	0.52	0.18	1.48	0.22
Education level (father, n = 155)
Primary/secondary school	642	Ref
Diploma	535	−0.08	0.42	0.92	0.40	2.11	0.85
Degree and above	390	−0.12	0.47	0.88	0.35	2.23	0.79
Smoking status (father, n = 152)
Never	691	Ref
Stop for more than 6 mo	160	−0.08	0.64	0.92	0.26	3.25	0.90
Smoker	716	−0.11	0.39	0.89	0.42	1.92	0.77
Parity
0	420	Ref
1	460	−0.46	0.50	0.63	0.24	1.68	0.36
≥2	687	−0.24	0.43	0.79	0.34	1.82	0.58
Health status at prepregnancy (group)
No	137	Ref
Yes	190	0.11	0.54	1.12	0.39	3.23	0.84
Labor method
Spontaneous vaginal delivery	1320	Ref
Cesarean delivery	247	1.01	0.39	2.74	1.27	5.93	0.01*

CI = credible interval; Freq. = frequency; OR = odd ratios.

*p value < 0.05.

3.2. Mixed-effect multiple logistic regression model

Four significant variables and two other variables with p < 0.25 (changes in weight at 28 and 36 weeks of gestation compared to the predelivery stage, respectively) were tested in the mixed-effect multiple logistic model. Due to the high correlation among the weight independent variables, the most clinically important and statistically significant variables (p < 0.05) were retained in the final model. In the final model, two variables, namely the changes in weight at 32 weeks of gestation compared to that in the predelivery stage (fixed-effect variable) and the changes in weight at each interval compared to that in the previous interval (time-varying variable), showed significant effect on the probability of having a low birth weight or premature baby (Table 5). Comparing the weight at 32 weeks of gestation and predelivery stage, women with higher increase in weight had lower probability of having a low birth weight or premature baby (odds ratio [OR] = 0.64, 95% [CI], 0.49-0.85; p < 0.001). When comparing the weight at each interval (t) and previous interval (t − 1), women with an increase in weight of >2 kg had a higher probability of having a low birth weight or premature baby than that in women with an increment of <1 kg (OR = 8.43; 95% CI, 2.90-24.54; p < 0.001).

Table 5.

Mixed effect logistic regression model of having low birth weight or premature baby among pregnant women

Variable	Freq.	Coefficient	SE	OR	95% CI		p
Intercept		−3.79	0.51
Changes of weight at 32 wk vs prepredelivery	1567	−0.44	0.14	0.64	0.49-0.85		<0.001*
Changes of weight at each interval vs previous interval (t, t − 1)
≤1.00 kg	634	Ref
1.01-2.00 kg	480	1.02	0.53	2.78	0.98	7.92	0.06
>2.00 kg	297	2.13	0.55	8.43	2.90	24.54	<0.001*

BMI = body mass index; CI = credible interval; Freq. = frequency; OR = odd ratios.

Possible confounders (pregnant women characteristics, weight, and BMI at prepregnancy stage) were tested in the final model. No significant effect was found on the final model.

*p value < 0.05.

4. DISCUSSION

In this study, 19.23% (n = 30) of women experienced either low birth weight or premature birth. This is comparable to the prevalence of low birth weight (9.7%) and preterm birth (6.7%), as reported by the Maternal and Child Health 2016 (total prevalence was 16.4%).¹ Low birth weight and premature babies (preterm and early term, ≤37 weeks) were chosen as the outcomes of this study, as many studies have demonstrated that a low birth weight baby or premature baby was associated with poor health outcomes in babies, including a higher mortality risk,^4,12 as well as health and development problems.²¹ Early-term babies have been associated with a higher probability of oxygen therapy, poor cardiorespiratory fitness, hypoglycemia, neonatal death, NICU admission, poorer cognitive outcome, and respiratory morbidities.^3,10–12 Therefore, many strategies have been devised from the government, non-governmental organizations, and other related parties to mitigate this situation, such as providing folic acid supplementation to pregnant mothers and more frequent antenatal checkups for high-risk mothers.

In this study, maternal weight was proven to be associated with low birth weight and premature birth, especially weight changes over the gestational weeks. The Institute of Medicine Weight Gain Recommendation for Pregnancy^15,16 has suggested gestational weight gain in the second and third trimesters according to prepregnancy weight categories. For instance, a 1-lb (~0.45-kg) gain per week is suggested for pregnant women with normal or underweight gain. Many studies have investigated the fixed-effect variable towards pregnancy outcome, such as total gestational gain throughout pregnancy.^9,15,16 However, this study also tested various time-varying effect variables (Table 4) as weight and gestational gain of pregnant women are not constant during pregnancy. Investigating the weight at each stage (or month) and the changes in gestational gain over time could provide more detailed information on the effects of weight on pregnancy outcomes. For instance, many studies have suggested that total gestational gain is positively associated with infant birth weight^9,15,16; however, the results of this study suggested that weight changes over 4 weeks were a risk factor for low birth weight or premature birth. Weight gain >2 kg (4.4 lb) over 4 weeks could lead to a higher probability of low birth weight or premature birth as compared to those who have gained <1 kg (2.2 lb) in weight. A huge increase in body weight over 4 weeks could indirectly indicate that a pregnant mother might have encountered health problems with the possibility of influencing the baby’s condition. A sudden increase might also indicate that the pregnant mother could have suffered from gestational diabetes or high blood pressure. In some traditional cultures, especially among the older generation, a pregnant mother who has gained more weight is believed to lead to a better baby outcome. However, this myth has to be corrected properly, in which the maintenance of a good weight gain throughout the gestational weeks is crucial, as this study suggested that body weight increment over 4 weeks has to be controlled to minimize the probability of giving low birth weight or premature babies.

The results also suggested that weight gain after 32 weeks was negatively associated with a higher probability of low birth weight or premature birth. A higher weight gain after 32 weeks was less likely to cause a low birth weight and premature birth, while weight gain within the first, second, and third trimesters was not significantly associated with the probability of having low birth weight or premature babies. This result has to be interpreted carefully as the mother’s weight in this study was not recorded every month but was estimated based on the method stated in Table 2. Therefore, it is difficult to conclude that the increase in body weight in different trimesters has a significant effect on the probability of low birth weight or premature birth. Future studies are required to record the weight of pregnant mothers every month to understand the effect of weight gain for each trimester on the probability of low birth weight or premature birth.

One common maternal factor that has been reported by several studies investigating the probability of low birth weight or premature birth was the prepregnancy BMI of the mothers.^7,9 The authors found that the prepregnancy BMI was negatively associated with the probability of low birth weight or premature birth. A mother with a higher BMI (>20) has a lower risk of low birth weight than a mother with a lower BMI (<20).^7,9 However, we could not detect any significant association between the BMI at the prepregnancy stage, probably due to the small sample size in this study, as only 30 women in this study delivered low birth weight or premature babies. They were five thin mothers, 15 mothers with normal BMI, and 10 obese mothers. Based on the previous literature, prepregnancy BMI is undoubtedly important, where more mothers should be recruited in future studies, particularly from different BMI groups at baseline.

One limitation of this study is that the estimation of weight at each interval was based on particular time points due to logistic challenges and variations in the clinic visit date. It was assumed that a constant change (increase/decrease) occurred during the 4-week intervals, which might not accurately represent the actual weight in each month. Future studies are required to record weight status per month. In addition, survival analysis such as multilevel discrete time event model could not be utilized in this study, as preterm or early-term babies are related to the duration of gestational week. However, a multilevel discrete time event model can be used to investigate the probability of having low birth weight, which is a health issue of concern lately. In fact, an increased number of low birth weight cases may allow utilization of survival analysis in a larger study population in the future.

In conclusion, maternal weight was monitored every 4 weeks to minimize the chances of having low birth weight and premature babies. An increase of >2 kg over 4 weeks in a pregnant woman should warrant consultation from doctors. An increase in weight gain after 32 weeks was shown to reduce the risk of low birth weight and premature babies.

APPENDIX A. SUPPLEMENTARY DATA

Supplementary data related to this article can be found at http://doi.org/10.1097/JCMA.0000000000000264.