Table 3.
Summary of statistical analysis, adjustments made and significant results reported across the 11 included studies
| First author and year | Statistical analysis | Significant results reported* | |
|---|---|---|---|
| Exposure | Outcome | ||
| Adane 201829 | Logistic regression | Interpregnancy weight gain (≥ 4 kg/m2) (Ref: Stable interpregnancy weight (± 1 kg/m2)) | Obesity (aOR 2.20, 95% CI [1.02, 4.75]) |
| Adjustments: Maternal characteristics prior to the second birth (age, area of residence, education, smoking and physical activity), interpregnancy interval and BMI prior to the first birth | |||
| Albers 201840 | Individual patient data meta-analysis on five studies. Multilevel model separating within-family and between-family effects | Each additional cigarette smoked a day in sibling pregnancies | BMI z-score (β = 0.007, 95% CI [0.006, 0.009]) |
| Adjustments: Maternal weight status, breastfeeding and maternal education attained by the start of the respective pregnancy | |||
| Aucott 201730 | Multinomial multilevel logistic modelling and a two level multivariate model. Analysis undertaken with and without sibling analysis | Maternal weight loss >10% (Ref: weight change ±3%) | Overweight (aOR 1.3, 95% CI [1.0, 1.6]) |
| Obesity (aOR 1.9, 95% CI [1.4, 2.7]) | |||
| Maternal weight gain ≥10% (Ref: weight change ±3%) | Mean BMI z-score (β = 0.13, 95% CI [0.05, 0.20]) | ||
| Maternal smoking: | |||
| With sibling analysis: | |||
| ● Starting to smoke between pregnancies | Mean BMI z-score (β = 0.19, 95% CI [0.01, 0.36]) | ||
| ● Smoking in both pregnancies | Mean BMI z-score (β = 0.10, 95% CI [0.01, 0.20]) | ||
| Without sibling analysis: | |||
| ● Starting to smoke between pregnancies | Mean BMI z-score (β = 0.22, 95% CI [0.07, 0.36]) | ||
| ● Smoking in both pregnancies | Mean BMI z-score (β = 0.32, 95% CI [0.24, 0.40]) | ||
| ● Quitting between pregnancies | Mean BMI z-score (β = 0.29, 95% CI [0.18, 0.40]) | ||
| Odds ratios (Ref: never smokers) | |||
| ● Starting to smoke between pregnancies | Overweight (aOR 1.5, (95% CI) 1.0, 2.2]) | ||
| Obesity (aOR 2.0, 95% CI [1.1, 3.6]) | |||
| ● Smoking in both pregnancies | Overweight (aOR 1.5, 95% CI [1.2, 2.8]) Obesity (aOR 1.8, 95% CI [1.3, 2.6]) | ||
| ● Quit between pregnancies | |||
| Overweight (aOR 1.5, 95% CI [1.0, 2.0]) | |||
| Obesity (aOR 1.6, 95% CI [1.0, 2.5]) | |||
| Adjustments: Child sex, maternal weight category, maternal smoking (for weight change exposure)/percentage weight change (for smoking change exposure), socio-economic status, parity, maternal age and birthweight z-score | |||
| Barclay 201831 | Analysis: Between-family analysis using linear regression and within-family analysis, using sibling fixed effects | Preceding interval length (Ref: 25–30 months) | |
| ● Between-family analysis: (months) | Overweight/obesity: | ||
| 31–36 | (β = 0.012, 95% CI [0.005, 0.019]) | ||
| 37–42 | (β = 0.013, 95% CI [0.005, 0.020]) | ||
| 43–48 | (β = 0.016, 95% CI [0.008, 0.024]) | ||
| 49–54 | (β = 0.018, 95% CI [0.009, 0.027]) | ||
| 55–60 | (β = 0.033, 95% CI [0.023, 0.043]) | ||
| 61–66 | (β = 0.036, 95% CI [0.024, 0.047]) | ||
| 67–72 | (β = 0.042, 95% CI [0.029, 0.055]) | ||
| 73–78 | (β = 0.044, 95% CI [0.030, 0.059]) | ||
| 79–84 | (β = 0.057, 95% CI [0.040, 0.073]) | ||
| 85–90 | (β = 0.057, 95% CI [0.038, 0.076]) | ||
| 91–96 | (β = 0.075, 95% CI [0.052, 0.097]) | ||
| 97+ | (β = 0.084, 95% CI [0.070, | ||
| 0.098]) | |||
| ● Within-family analysis: (months) | Overweight/obesity: | ||
| 31–36 | (β = 0.014, 95% CI [0.000, 0.028]) | ||
| 55–60 | (β = 0.021, 95% CI [0.002, 0.040]) | ||
| 67–72 | (β = 0.034. 95% CI [0.009, 0.059]) | ||
| 79–84 | (β = 0.062, 95% CI [0.031, 0.093]) | ||
| 97+ | (β = 0.039, 95% CI [0.009, 0.068]) | ||
| Adjustments: Birth order, maternal age, birth year, sibling group size, age at and year of conscription | |||
| Devakumar 201632 | Multivariable linear regression | Birth interval ≥ 18 months (ref: <18 months) | Fat free mass (kg) (β = 1.717, 95% CI [0.242, 3.193]) |
| Continuous birth interval (females) | Fat free mass (kg) (β = 0.014, 95% CI [0.000, 0.027]) | ||
| Visceral fat (cm) (β = 0.004, 95% CI [0.000, 0.007]) | |||
| Adjustments: Maternal age, education, BMI at the beginning of pregnancy, family income at birth and birth order | |||
| Huttly 199233 | ANOVA (p-value < 0.001) | Birth interval in months | Mean weight-for-height z-scores |
| <18 | 0.07 | ||
| 18–23 | -0.03 | ||
| 24–35 | 0.06 | ||
| 36–47 | 0.23 | ||
| 48–71 | 0.19 | ||
| >71 | 0.27 | ||
| Adjustments: Maternal income, race, education, cohabitation, age and parity | |||
| Iliadou 201034 | Analysis: Logistic regression | Smoking in both male pregnancies (Ref: non-smokers in pregnancy) | Overweight (aOR 1.71, 95% CI [1.39, 2.09]) |
| Adjustments: Maternal age, height, BMI, pregnancy weight gain, maternal and paternal socio-economic category and education, offspring birthweight, head circumference, gestational age, urban living and age at conscription | |||
| Li 201835 | Generalized linear regression models | Length of the interpregnancy interval | No significant results reported |
| Adjustments: Infant gender, maternal race, maternal marriage status, SES, mother’s educational level, smoking status, pregestational/gestational diabetes, pregestational/gestational hypertension (in the continuous model only), gestational age, prepregnancy BMI, weight gain in pregnancy, mode of labour onset, birthweight for gestational age, Apgar score at 5 minutes, birthweight (in the categorical model only), feeding pattern, rapid weight gain in the first year of life, and maternal age at birth | |||
| Smithers 201736 | Augmented inverse probability weighted analysis | Mode of birth | No significant results reported |
| Adjustments: Maternal age, type of antenatal care, number of antenatal visits, medical conditions in pregnancy (asthma, diabetes, hypertension), smoking in pregnancy, gestational age, birthweight for gestational age z-scores, maternal partnership status, maternal ethnicity, maternal occupation, neighbourhood-level indicators of socioeconomic disadvantage and remote residence | |||
| Willmer 201337 | Fixed-effects regression | Maternal weight change | No significant results reported |
| Adjustments: Sex of siblings, birth order, mother’s age and smoking in pregnancy | |||
| Yuan 201638 | Within-family analysis using conditional logistic regression | Vaginal birth after caesarean birth | Obesity (RR 0.69, 95% CI [0.53, 0.83]) |
| Maternal age at birth, race, prepregnancy BMI group, maternal height, gestational diabetes, pre-eclampsia, pregnancy induced hypertension, child sex, year of birth, gestational age at birth, birth order, birth weight group, prepregnancy smoking and region of residence at birth | |||
Abbreviations: aOR, adjusted odds ratio; BMI, body mass index (kg/m2); CI, confidence interval; RR, relative risk.
*
Significant results are those where the confidence interval for association does not include the null (1.0 for odds and risk ratios, 0.0 for linear coefficients), or where p values <0.05 if these are unavailable.