{
  "abstract": "Introduction We aimed to examine the impact of gestational weight gain (GWG) on perinatal outcomes based on body mass index (BMI) category and gestational diabetes mellitus (GDM) status.Research design and methods This retrospective study included 53 183 Japanese women with GDM and 734 028 with normal glucose tolerance (NGT). We assessed interaction effects of BMI categories (underweight <18.5, normal 18.5–24.9, overweight 25.0–29.9, obese ≥30.0 kg/m²) and GDM status on GWG and perinatal outcomes, including birth weight, hypertensive disorders of pregnancy, preterm delivery, mode of delivery, and a composite outcome, using generalized linear models with a logit link function.Results BMI category and GDM status had significant interaction effects on the association between GWG and all outcomes (all p<0.05, using the likelihood ratio test). Notably, the impact of increased GWG on the risk of having large-for-gestational-age newborns was greater in women with GDM than in those with NGT, a trend that was most pronounced in the obese group. For GWG, from 5–6 kg to 15–16 kg, the predicted probability of large-for-gestational-age newborns increased by 10.0% (from 4.9% to 14.9%) for NGT compared with 15.6% (from 6.3% to 21.8%) for GDM in the normal BMI group, and by 14.4% (from 17.6% to 32.0%) and 23.1% (from 23.5% to 46.6%), respectively, in the obese group.Conclusions The impact of GWG on perinatal outcomes differed according to the BMI category and GDM status. These findings support the need to take into account both factors when developing GWG recommendations to minimize the risk of adverse perinatal outcomes.",
  "authors": [
    {
      "affiliations": [
        "Department of Metabolic Medicine, Osaka University School of Medicine Graduate School of Medicine, Suita, Japan",
        "Department of Obstetric Medicine, Osaka Women's and Children’s Hospital, Izumi, Japan"
      ],
      "name": "Kei Fujikawa Shingu"
    },
    {
      "affiliations": [
        "Department of Laboratory Medicine, Osaka University School of Medicine Graduate School of Medicine, Suita, Japan"
      ],
      "name": "Mitsuyoshi Takahara"
    },
    {
      "affiliations": [
        "Department of Obstetric Medicine, Osaka Women's and Children’s Hospital, Izumi, Japan"
      ],
      "name": "Masako Waguri"
    },
    {
      "affiliations": [
        "Department of Obstetrics and Gynecology, Juntendo University Faculty of Medicine, Bunkyo, Japan"
      ],
      "name": "Jun Takeda"
    },
    {
      "affiliations": [
        "Department of Metabolic Medicine, Osaka University School of Medicine Graduate School of Medicine, Suita, Japan"
      ],
      "name": "Naoto Katakami"
    },
    {
      "affiliations": [
        "Department of Metabolic Medicine, Osaka University School of Medicine Graduate School of Medicine, Suita, Japan"
      ],
      "name": "Iichiro Shimomura"
    },
    {
      "affiliations": [
        "Department of Obstetrics and Gynecology, Juntendo University Faculty of Medicine, Bunkyo, Japan"
      ],
      "name": "Atsuo Itakura"
    },
    {
      "affiliations": [
        "Department of Obstetrics and Gynecology, Ehime University School of Medicine, Toon, Japan"
      ],
      "name": "Takashi Sugiyama"
    },
    {
      "affiliations": [
        "Department of Social Medicine, National Center for Child Health and Development, Setagaya, Japan"
      ],
      "name": "Naho Morisaki"
    }
  ],
  "full_text": "WHAT IS ALREADY KNOWN ON THIS TOPIC Achieving appropriate gestational weight gain (GWG) is important to reducing the risk of perinatal complications. Although GWG targets are conventionally based on pre-pregnancy body mass index (BMI) categories, it is unclear whether BMI modifies the link between GWG and perinatal risk. Gestational diabetes mellitus (GDM) may also modify this association, but its interaction with GWG is not well established.WHAT THIS STUDY ADDS This study demonstrates that pre-pregnancy BMI categories and GDM status significantly modified the associations between GWG and multiple perinatal complications, including birth weight, hypertensive disorders of pregnancy, mode of delivery, preterm delivery, and their composite outcome.HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY These findings support the need to develop GWG recommendations tailored to pre-pregnancy BMI categories and GDM status to minimize the risk of perinatal complications.Introduction Gestational weight gain (GWG) significantly impacts perinatal complications. Insufficient GWG increases the risk of low birthweight (LBW) infants and preterm delivery, whereas excessive GWG increases the risks of macrosomia, hypertensive disorders of pregnancy (HDP), and cesarean delivery. 1–4 Clinical guidelines recommend greater GWG for women with lower pre-pregnancy body mass index (BMI) and lesser GWG for those with higher pre-pregnancy BMI.5–7 This recommendation is based on the assumption that the impact of GWG on perinatal outcomes varies across pre-pregnancy BMI categories, indicating an interaction effect between pre-pregnancy BMI and GWG. Several studies have examined the interaction between pre-pregnancy BMI and GWG8–15; however, their findings have been inconsistent. This inconsistency may be owing to limited sample sizes, highlighting the need for an analysis based on a larger dataset.Another potential modifier of the association between GWG and perinatal risk is abnormal glucose metabolism during pregnancy, including gestational diabetes mellitus (GDM) and pregestational diabetes mellitus. Among these, GDM is particularly important given its high prevalence. However, few studies have investigated the interaction between GDM and GWG; moreover, the range of perinatal outcomes examined has been limited.16–18We aimed to examine the interaction effects between GWG and BMI categories, as well as between GWG and GDM status. Specifically, we aimed to assess the influence of GWG on perinatal outcomes across BMI categories and between women with GDM and those with normal glucose tolerance (NGT).Research design and methods Study design This retrospective study used data from the nationwide perinatal registry of the Japan Society of Obstetrics and Gynaecology (JSOG). Primarily, this registry includes data from tertiary care centers, regional perinatal centers, and general hospitals across Japan. At these facilities, obstetricians systematically record maternal characteristics, pre-existing maternal conditions, pregnancy complications, delivery details, and neonatal outcomes for all live births and stillbirths at ≥22 weeks of gestation, using a standardized format. In recent years, the registry has included approximately 25% of all births in Japan, making it the largest obstetric database in the country. We evaluated the perinatal risks and GWG in women with GDM and those with NGT who were registered in the JSOG database between 2015 and 2020.Participants The inclusion criteria were singleton pregnancies with GDM or NGT. We excluded cases with missing data on maternal age at delivery, height, pre-pregnancy weight, weight at delivery, number of fetuses, gestational age at delivery, birth weight, sex of the newborn, number of previous deliveries, mode of delivery, or obstetric complications. Additionally, we excluded women who gave birth before 18 or after 45 years of age, had chronic hypertension (diagnosed before pregnancy or before 20 weeks of gestation), pre-existing diabetes mellitus, or overt diabetes in pregnancy. Other exclusions included cases of stillbirth, deliveries before 28 weeks or after 42 weeks of gestation, use of psychotropic medications that could increase the blood sugar levels, and a history of cesarean delivery. Furthermore, deliveries by methods other than cesarean section, instrumental delivery, or spontaneous vaginal delivery, as well as cases with an estimated GWG of <−3 kg or ≥20 kg at 40 weeks of gestation, were excluded. The database used for this study did not include data on race or ethnicity. However, virtually all women giving birth in Japan are of Japanese ethnicity; therefore, the study population can reasonably be considered predominantly Japanese.Outcomes The outcomes assessed were macrosomia, LBW, large for gestational age (LGA), small for gestational age (SGA), pre-eclampsia, HDP diagnosed at ≥20 weeks of gestation, preterm delivery before 34 weeks of gestation, emergent cesarean delivery, instrumental delivery, and a composite outcome. The composite outcome was defined as the occurrence of any of the following: macrosomia, LBW, pre-eclampsia, preterm delivery before 34 weeks of gestation, emergent cesarean delivery, or instrumental delivery.Definitions GDM was determined based on a diagnosis recorded in the database. In Japan, GDM is diagnosed in women who do not have pre-existing diabetes mellitus or overt diabetes identified during pregnancy. 19 Definitions of pre-existing diabetes mellitus and overt diabetes in pregnancy, as well as the detailed screening algorithm and diagnostic criteria, are provided in online supplemental materials.SP110.1136/bmjdrc-2025-005709.supp1Supplementary dataMaternal pre-pregnancy BMI was calculated as the pre-pregnancy weight in kilograms divided by the square of the height in meters. BMI categories were as follows: underweight (<18.5 kg/m²), normal (18.5–24.9 kg/m²), overweight (25.0–29.9 kg/m²), and obese (≥30.0 kg/m²).20 We calculated the total GWG by subtracting pre-pregnancy weight from weight at delivery. To account for differences in gestational length at delivery, GWG was standardized to an estimated value at 40 weeks of gestation. This approach assumes that individual GWG follows an approximately linear trajectory during late pregnancy (28–41 weeks) with respect to gestational age, characterized by a subject-specific rate of weight gain, an assumption validated in the original study.21 Specifically, GWG Y (kg) at gestational age X (weeks) was modeled as a linear function of gestational length with an individual-specific slope A (kg/week), anchored at two constants (x′, y′). These constants, derived from a previously published model, were assumed to be common across individuals.21 In the analysis, the estimated GWG at 40 weeks of gestation was categorized into 1 kg intervals, such as −3 kg to <−2 kg, −2 kg to <−1 kg, and so on, up to 19 kg to <20 kg.Smoking status was categorized as follows: active smoking during pregnancy, active smoking before pregnancy, passive smoking (no active smoking before or during pregnancy), no history of smoking (no active or passive smoking before or during pregnancy), and unknown.Macrosomia and LBW were defined as birth weights >4000 g and <2500 g, respectively, whereas normal birth weight was defined as 2500–4000 g. LGA and SGA were defined as sex- and parity-specific birth weights for gestational age >90th percentile and <10th percentile, respectively, based on Japanese fetal growth curves, whereas appropriate for gestational age (AGA) was defined as birth weight between the 10th and 90th percentiles.22 HDP and pre-eclampsia were defined according to Japanese clinical criteria.23 Only cases diagnosed at ≥20 weeks of gestation were included. Detailed diagnostic criteria are provided in online supplemental materials. Preterm delivery before 34 weeks of gestation was defined as delivery before 34 complete weeks of gestation.Statistical analysis We developed the multinomial logistic regression models of GWG on birth weight categories (macrosomia, LBW, or normal birth weight), fetal growth categories (LGA, SGA, or AGA), and mode of delivery (emergent cesarean delivery, instrumental delivery, or either spontaneous vaginal delivery or elective cesarean delivery). However, binomial logistic regression models of GWG on HDP, pre-eclampsia, preterm delivery before 34 weeks of gestation, and a composite outcome were developed. To clarify the differences in the effect of GWG on perinatal outcomes across BMI categories and between women with GDM and those with NGT, we evaluated interaction effects using models 1–3. Model 1 included seven covariates without any interaction terms: GWG (estimated GWG at 40 weeks of gestation, modeled as a spline variable), BMI categories (underweight, normal, overweight, and obese), GDM status (GDM or NGT), maternal age, maternal height, maternal parity (primipara or multipara), and smoking status (active smoking during pregnancy, active smoking before pregnancy, passive smoking, no smoking history, and unknown). Model 2 included the interaction term between BMI category and GWG into model 1, whereas model 3 included the interaction terms among GDM status, BMI category, and GWG. First, we compared models 1 and 2, using the likelihood ratio test (LRT), to evaluate whether the interaction effect between BMI category and GWG was significant for perinatal outcomes. Second, models 2 and 3 were compared using the LRT to assess whether the interaction effect among GDM status, BMI category, and GWG was significant.The predicted probabilities of each outcome for GDM and NGT were calculated across GWG values within each BMI category using model 3. Pre-pregnancy BMI, age, and height were set to the mean values of all cases (GDM and NGT) within each BMI category, and parity was set to its expected value based on the same population. Smoking status was assumed to be “no history of smoking.” The mean predicted probabilities and their 95% CIs were calculated using the bootstrap method with 2000 resampling iterations.Additionally, we conducted several sensitivity analyses. First, instead of representing body type with BMI categories, we treated BMI as a continuous variable and evaluated its interaction with GWG. Second, we categorized estimated GWG at 40 weeks of gestation into three groups—below, within, or above the current Japanese GWG guidelines—and assessed its interaction with BMI category. Finally, we replaced estimated GWG at 40 weeks of gestation with total GWG (both modeled as spline variables) and evaluated its interaction with BMI category.We also examined the association between GWG and the risk of multiple overlapping perinatal complications (ie, cumulative risk), using the current Japanese GWG guidelines.6 We assessed this cumulative risk in women with GDM and those with NGT across all BMI categories, based on the current Japanese GWG guidelines.6 In model 3, we estimated the predicted probabilities for six outcomes—macrosomia, LBW, pre-eclampsia, preterm delivery before 34 weeks, emergent cesarean delivery, and instrumental delivery—separately for GDM and NGT groups and stratified them based on BMI category and GWG level. During the estimation, mean or expected values were substituted for the covariates within each BMI category and GDM status group, except for smoking status, which was considered to be “no history of smoking.” The mean cumulative risks and their corresponding 95% CIs were obtained using the bootstrap method with 2000 resamples.To account for the clinical importance of each outcome, we calculated the weighted cumulative risk according to the current Japanese GWG guidelines.6 For women with NGT, we applied weights based on a 2020 nationwide survey by the JSOG and the Japan Association of Obstetricians and Gynaecologists (JAOG).6 For women with GDM, we conducted a similar nationwide survey that included 161 of 712 Japanese Society of Diabetes and Pregnancy members in 2023. Respondents rated the importance of 10 perinatal outcomes related to weight management on a 5-point scale. Six of these outcomes—macrosomia, LBW, pre-eclampsia, preterm delivery before 34 weeks, emergent cesarean delivery, and instrumental delivery—were analyzed in accordance with the guidelines. The weight for each selected outcome was calculated by dividing its mean score by the total of all mean scores. For example, the weight for macrosomia reflected the combined rating of macrosomia and LGA, and the weight for LBW reflected the rating of fetal stunting and LBW. Each predicted probability (from model 3) was multiplied by its corresponding weight to calculate the weighted cumulative risk. The final weighted cumulative risks and their 95% CIs were estimated using 2000 bootstrap resamples. All statistical analyses were performed using the R software, V.4.4.2 (R Foundation for Statistical Computing, Vienna, Austria). Continuous variables are expressed as mean±SD and categorical variables as percentages (%).Patient and public involvement Patients and the public were not involved in the design, conduct, reporting, or dissemination plans of this research.Results Study sample, baseline characteristics, and perinatal outcomes Of the 1,399,066 deliveries in the database overall, we excluded 611,855 cases, mostly comprising those with missing data or a history of cesarean delivery, as well as others (see online supplemental figure 1 for the breakdown). The characteristics of women with GDM and NGT across BMI categories are presented in table 1. Among women with GDM (n=53,183), 10.6%, 62.5%, 18.0%, and 8.9% were underweight, normal, overweight, and obese, respectively. Mean GWG decreased with increasing BMI, ranging from 8.9±3.6 kg in the underweight category to 8.3±4.0 kg in the normal, 6.3±4.7 kg in the overweight, and 4.5±4.6 kg in the obese category. Among women with NGT (n=734,028), 17.5%, 72.5%, 7.8%, and 2.2% were underweight, normal, overweight, and obese, respectively. A similar trend in GWG was observed in the NGT group, with mean GWG highest in the underweight category (10.1±3.4 kg), followed by the normal (9.9±3.6 kg), overweight (8.0±4.5 kg), and obese (5.8±4.6 kg) categories. The number and incidence of individual perinatal outcomes ranged from 6104 (0.8%) for macrosomia to 102 863 (13.1%) for LBW (table 1). The detailed perinatal outcomes of women with GDM and NGT across BMI categories are presented in table 1.Table 1Maternal characteristics and perinatal outcomes of women with GDM stratified based on BMI categoriesAll casesCases with underweightCases with normal weightCases with overweightCases with obesity(GDM n=53,183; NGT n=734,028)(GDM n=5650; NGT n=128,336)(GDM n=33,240; NGT n=532,525)(GDM n=9579; NGT n=57,182)(GDM n=4714; NGT n=15,985)Age at delivery (years)GDM33.9±5.032.9±5.134.0±5.034.3±5.033.5±5.1NGT32.1±5.331.3±5.334.0±5.032.7±5.432.2±5.3BMI before pregnancy (kg/m2)GDM23.1±4.417.6±0.821.4±1.827.1±1.433.0±2.5NGT21.1±3.217.6±0.820.9±1.626.9±1.432.7±2.3Body weight before pregnancy (kg)GDM57.7±11.644.3±3.653.6±5.767.8±5.982.2±7.7NGT52.9±8.544.5±3.652.6±5.367.1±5.981.4±7.4Height (cm)GDM158.2±5.5158.6±5.5158.2±5.5157.9±5.6157.7±5.5NGT158.4±5.5158.9±5.5158.4±5.4157.9±5.6157.7±5.5Multipara (%)GDM45.041.544.547.747.2NGT42.641.042.446.745.5Smoking history Never smoking (%)GDM71.374.173.467.161.3NGT71.972.272.568.462.8 Active smoking during pregnancy (%)GDM1.31.11.01.82.8NGT1.01.10.91.62.5 Passive smoking during pregnancy (%)GDM7.66.57.58.58.4NGT7.27.37.17.77.6 Active smoking before pregnancy (%)GDM10.89.39.413.217.3NGT8.58.38.110.814.3 Unknown (%)GDM9.09.08.79.410.2NGT11.411.211.411.612.8Estimated GWG at 40 weeks of gestation (kg)GDM7.6±4.38.9±3.68.3±4.06.3±4.74.5±4.6NGT9.7±3.810.1±3.49.9±3.68.0±4.55.8±4.6Gestational week of deliveryGDM38.5±1.738.3±1.738.5±1.738.5±1.738.5±1.9NGT38.6±1.838.4±1.938.6±1.838.6±2.038.6±2.0Macrosomia (%)GDM1.50.31.12.53.9NGT0.70.30.71.62.4LBW (%)GDM11.818.411.510.19.4NGT13.217.912.311.710.8LGA (%)GDM13.86.412.219.023.6NGT9.35.29.415.318.7SGA (%)GDM6.911.16.85.55.2NGT8.511.97.96.96.3Pre-eclampsia (%)GDM1.91.01.42.84.8NGT1.81.31.73.34.8HDP (%)GDM6.63.44.810.016.5NGT5.33.64.910.416.3Preterm delivery before 34 weeks of gestation (%)GDM1.92.11.82.02.3NGT2.32.62.22.82.7Emergent cesarean delivery (%)GDM13.510.212.116.621.3NGT12.19.911.817.121.3Instrumental delivery (%)GDM9.710.310.38.47.4NGT9.29.89.37.96.5Composite outcome (%)GDM32.934.031.534.038.9NGT30.932.830.034.037.5Underweight, pre-pregnancy BMI<18.5 kg/m²; normal, pre-pregnancy BMI 18.5–24.9 kg/m²; overweight, pre-pregnancy BMI 25.0–29.9 kg/m²; obese, pre-pregnancy BMI≥30.0 kg/m²; composite outcome, occurrence of any of the following—macrosomia, LBW, pre-eclampsia, preterm delivery before 34 weeks of gestation, emergent cesarean delivery, or instrumental delivery.Number of instances of missing data on pre-eclampsia among women with GDM: 1115 cases (2.1%) in the total sample, 66 cases (1.2%) in the underweight group, 492 cases (1.5%) in the normal group, 311 cases (3.2%) in the overweight group, and 246 cases (5.2%) in the obese group.Number of instances of missing data on pre-eclampsia among women with NGT: 10,613 cases (1.4%) in the total sample, 1157 cases (0.9%) in the underweight group, 6949 cases (1.3%) in the normal group, 1764 cases (3.1%) in the overweight group, and 743 cases (4.6%) in the obese group.BMI, body mass index; GDM, gestational diabetes mellitus; GWG, gestational weight gain; HDP, hypertensive disorders of pregnancy; LBW, low birth weight; LGA, large for gestational age; NGT, normal glucose tolerance; SGA, small for gestational age.Interaction effects The interaction effect between BMI category and GWG was significant for all outcomes (all p<0.05, table 2). The interaction effect among GDM status, BMI category, and GWG was also statistically significant (all p<0.05, table 2).Table 2Interaction effects between BMI categories and GWG on perinatal outcomesInteraction variablesOutcomedfχ2P valueBMI categories and GWGBirth weight categories24845.766<0.001Fetal growth categories24682.081<0.001HDP12148.543<0.001Pre-eclampsia1289.474<0.001Mode of delivery24305.792<0.001Preterm delivery before 34 weeks of gestation12199.421<0.001Composite outcome121315.617<0.001GDM status, BMI category, and GWGBirth weight categories3864.9960.004Fetal growth categories38147.117<0.001HDP1944.950<0.001Pre-eclampsia1946.491<0.001Mode of delivery38119.418<0.001Preterm delivery before 34 weeks of gestation1950.846<0.001Composite outcome19105.364<0.001Birth weight categories are macrosomia, low birth weight (LBW), and normal birth weight; on the other hand, fetal growth categories are large for gestational age, small for gestational age, and appropriate for gestational age, based on Japanese fetal growth curves. Mode of delivery is categorized into emergent cesarean delivery, instrumental delivery, and either spontaneous vaginal delivery or elective cesarean delivery. The composite outcome includes the occurrence of any of the following perinatal outcomes: macrosomia, LBW, pre-eclampsia, preterm delivery before 34 weeks of gestation, emergent cesarean delivery, or instrumental delivery.BMI, body mass index; GDM, gestational diabetes mellitus; GWG, gestational weight gain; HDP, hypertensive disorders of pregnancy.Predicted probabilities of each outcome The predicted probabilities estimated using model 3, which includes interaction terms among GDM status, BMI category, and GWG, are presented in figure 1. The impact of GWG on the composite outcome was higher in women with GDM than in those with NGT, especially in the obese category. These differences in predicted probabilities between women with GDM and those with NGT were particularly pronounced for outcomes related to fetal overgrowth. For example, among women in the obese category, the predicted probability of LGA at a GWG of 5–6 kg was 17.59% (95% CI 16.63% to 18.63%) for NGT and 23.49% (21.57% to 25.51%) for GDM, increasing to 31.99% (29.22% to 34.78%) and 46.56% (40.40% to 52.75%), respectively, at a GWG of 15–16 kg. Therefore, the absolute increase in LGA probability from 5–6 kg to 15–16 kg of GWG was 14.40 percentage points for NGT and 23.07 percentage points for GDM. Similar trends were observed across other BMI categories: in the underweight group, the increase was 8.15 percentage points for NGT versus 12.56 percentage points for GDM; in the normal group, 9.99 versus 15.56; and in the overweight group, 12.13 versus 18.23, respectively. A similar pattern was observed for macrosomia, particularly in the obese category, although differences in the underweight group were minimal. From 5–6 kg to 15–16 kg of GWG, the predicted probability of macrosomia increased by 0.56 percentage points for NGT and 0.35 percentage points for GDM in the underweight group; by 0.99 and 1.99 in the normal group; by 2.01 and 2.86 in the overweight group; and most notably, by 3.56 and 6.59 in the obese group, respectively.Figure 1Predicted probabilities of macrosomia (A), LBW (B), LGA (C), SGA (D), pre-eclampsia (E), HDP (F), preterm delivery before 34 weeks of gestation (G), emergent cesarean delivery (H), instrumental delivery (I), and composite outcome (J) in women with GDM and those with NGT stratified according to BMI categories. Red dots and solid lines indicate the mean predicted probabilities of each perinatal outcome in women with GDM, whereas white diamonds and dashed lines indicate those in women with NGT. Error bars indicate 95% CIs of the predicted probabilities. Underweight, pre-pregnancy BMI<18.5 kg/m²; normal, 18.5–24.9 kg/m²; overweight, 25.0–29.9 kg/m²; obese, ≥30.0 kg/m²; Composite outcome, occurrence of any of the following: macrosomia, LBW, pre-eclampsia, preterm delivery before 34 weeks of gestation, emergent cesarean delivery, or instrumental delivery. BMI, body mass index; HDP, hypertensive disorders of pregnancy; GDM, gestational diabetes mellitus; GWG, gestational weight gain; LBW, low birth weight; LGA, large for gestational age; NGT, normal glucose tolerance; SGA, small for gestational age.Conversely, outcomes related to fetal growth restriction (SGA and LBW) showed an inverse association with GWG (figure 1), and the differences in predicted probabilities between women with GDM and NGT were generally less pronounced. The predicted probabilities of SGA decreased as GWG increased in the GDM and NGT groups, although the magnitude of change varied by BMI category. For example, in the normal BMI category, the predicted probability of SGA decreased from 11.29% (95% CI 11.08% to 11.50%) to 4.66% (4.55% to 4.78%) in NGT and 8.52% (8.01% to 9.07%) to 3.60% (3.10% to 4.10%) in GDM when GWG increased from 5–6 kg to 15–16 kg, corresponding to absolute reductions of 6.63 and 4.92 percentage points, respectively. In the obese category, the corresponding reductions were smaller: 5.47% (4.98% to 6.01%) to 3.57% (2.74% to 4.53%) in NGT (−1.90 percentage points) and 3.54% (2.83% to 4.29%) to 2.95% (1.38% to 4.97%) in GDM (−0.59 percentage points). Overall, LBW probabilities also tended to decrease with increasing GWG. In the normal BMI category, the predicted probability of LBW decreased from 17.42% (95% CI 17.16% to 17.67%) to 8.68% (8.50% to 8.85%) in NGT and 13.87% (13.17% to 14.55%) to 8.06% (7.30% to 8.84%) in GDM, as GWG increased from 5–6 kg to 15–16 kg, corresponding to absolute reductions of 8.74 and 5.81 percentage points, respectively. In the obese category, changes in LBW probability were smaller and less consistent (NGT: 9.46% (8.78% to 10.18%) to 7.28% (5.92% to 8.78%); GDM: 7.49% (6.40% to 8.64%) to 8.43% (5.06% to 12.29%)).For pre-eclampsia, HDP, preterm delivery, emergency cesarean section, and instrumental delivery, interaction terms between GWG, BMI category, and GDM status were statistically significant (table 2), indicating heterogeneity in associations across BMI categories and by GDM status. However, the absolute differences in predicted probabilities between women with GDM and NGT were generally modest in magnitude across the GWG range, in each BMI category (figure 1).Cumulative risk The plots of unweighted and weighted cumulative risk and 95% CIs in women with GDM and those with NGT across all BMI categories are shown in online supplemental figures 2 and 3. The weight for each selected outcome is shown in online supplemental table 1.Sensitivity analyses As shown in table 3, the results of the LRT remained consistent with those of the main analysis despite modifications to the explanatory variables. For all perinatal outcomes, the interaction effects between body type and GWG were statistically significant.Table 3Sensitivity analyses evaluating the interaction effect between GWG and maternal body type on perinatal outcomesOutcomedfχ2P valueSensitivity analyses 1 Birth weight categories8556.8<0.001 Fetal growth categories8675.0<0.001 HDP4167.7<0.001 Pre-eclampsia494.3<0.001 Mode of delivery8251.7<0.001 Preterm delivery before 34 weeks of gestation4146.1<0.001 Composite outcome4770.8<0.001Sensitivity analyses 2 Birth weight categories6356.20.008 Fetal growth categories6682.1<0.001 HDP3129.6<0.001 Pre-eclampsia374.5<0.001 Mode of delivery640.4<0.001 Preterm delivery before 34 weeks of gestation334.6<0.001 Composite outcome340.2<0.001Sensitivity analyses 3 Birth weight categories24877.60.004 Fetal growth categories24511.8<0.001 HDP12140.7<0.001 Pre-eclampsia1265.5<0.001 Mode of delivery24717.7<0.001 Preterm delivery before 34 weeks of gestation12306.6<0.001 Composite outcome122246.2<0.001Sensitivity analyses 1: interaction effects between BMI (continuous variable) and estimated GWG at 40 weeks of gestation (modeled as spline terms).Sensitivity analyses 2: interaction effects between BMI categories and GWG categories based on whether the estimated GWG at 40 weeks of gestation fell below, within, or above the recommended range in the current Japanese GWG guidelines.Sensitivity analyses 3: interaction effects between BMI categories and total GWG (modeled as spline terms).Birth weight categories: macrosomia, low birth weight (LBW), and normal birth weight.Fetal growth categories: large for gestational age (LGA), small for gestational age (SGA), and appropriate for gestational age (AGA), based on Japanese fetal growth curves.Mode of delivery: emergent cesarean delivery, instrumental delivery, and either spontaneous vaginal delivery or elective cesarean delivery.Composite outcome: occurrence of any of the following—macrosomia, LBW, pre-eclampsia, preterm delivery before 34 weeks of gestation, emergent cesarean delivery or instrumental delivery.BMI, body mass index; GWG, gestational weight gain; HDP, hypertensive disorders of pregnancy.Discussion In this study, we observed statistically significant interaction effects of BMI categories and GDM status on the association between GWG and perinatal outcomes, indicating heterogeneity in associations between women with GDM and those with NGT. Notably, increases in GWG were associated with larger absolute increases in fetal overgrowth risk among women with GDM than among those with NGT, particularly in the obese category. In contrast, the differences in predicted probabilities between GDM and NGT were smaller for outcomes related to fetal growth restriction and generally modest for other outcomes. Although statistically significant, the absolute differences in predicted probabilities were modest for several outcomes other than those related to fetal growth, and the clinical implications of these interaction effects should therefore be cautiously interpreted.Typically, GWG recommendations are stratified according to pre-pregnancy BMI category.5–7 Previous studies have examined interaction effects between pre-pregnancy body type and GWG in the development of perinatal outcomes, including birth weight, HDP, cesarean delivery, and preterm delivery.8–15 However, most studies failed to detect statistically significant interactions, suggesting that pre-pregnancy body type did not modify the impact of GWG. Accordingly, these findings do not support the stratification of GWG recommendations based on pre-pregnancy body type. Conversely, the present study identified significant interactions between BMI categories and GWG across all examined outcomes, including birth weight classifications, fetal growth classifications, pre-eclampsia, HDP, preterm delivery before 34 weeks of gestation, and mode of delivery. The discrepancy between previous findings and our study may be attributed to differences in sample size; previous studies often involved smaller cohorts compared with our large-scale dataset. Another explanation for the discrepancy may be differences in the treatment of pre-pregnancy body type and GWG during the analysis. Many previous studies considered pre-pregnancy body type as a continuous variable (BMI) and GWG as a categorical variable (classified as below, within, or above the frequently used Institute of Medicine recommendation). To address this, we conducted sensitivity analyses to reproduce their method of attributing those variables. Consequently, interaction effects between pre-pregnancy body type and GWG remained statistically significant across all perinatal outcomes (all p<0.01, table 3), suggesting that the variation in the treatment of the variable is unlikely to significantly affect the results. The lack of significant interactions in previous studies may rather be attributed to the relatively smaller sample sizes, ranging from a few hundred to a maximum of approximately 60,000 participants. In contrast, our study included approximately 780,000 participants, providing sufficient power to detect these interactions.To our knowledge, only three studies investigated the interaction effects between GDM status and GWG on perinatal outcomes, such as inappropriate birth weight and cesarean delivery, with findings consistent with those of the present study.16–18 Cheng et al17 identified significant interactions between GDM status and GWG in the risks of LGA and cesarean delivery, whereas Hong et al18 reported similar interactions in the development of macrosomia. Similarly, Mitanchez et al16 found significant interactions between GDM status and third-trimester GWG in relation to birth weight. Although Mitanchez et al16 also examined the interaction between BMI categories and GWG, they did not find statistical significance, possibly owing to a small sample size (approximately 400 participants). In contrast, our study, with its sufficiently large sample size, identified significant interaction effects between GDM status and GWG in the development of perinatal outcomes.To our knowledge, this study is among the first to comprehensively examine the interaction effects of body type and GWG, as well as GDM status and GWG, on a wide range of perinatal outcomes associated with weight gain, using a sufficiently large sample size. Our findings suggest that BMI category and GDM status are important when interpreting the relationship between GWG and perinatal outcomes. Current GWG guidelines define recommended ranges according to BMI category but do not specifically address women with GDM. These results may contribute to ongoing discussions on whether GWG recommendations could be further individualized according to GDM status. We explored GWG ranges associated with the lowest weighted cumulative risks of adverse outcomes within each BMI category. This weighted approach was used to account for the co-occurrence of multiple complications and differences in their clinical importance, rather than treating all outcomes equally as in simple composite measures. This approach is consistent with the framework adopted in current Japanese GWG guidelines. The corresponding GWG ranges (approximately 12–13 kg for underweight, 9–10 kg for normal BMI, 5–6 kg for overweight, and 1–2 kg for obesity) should be interpreted as exploratory and hypothesis-generating rather than prescriptive targets. Further studies are needed to confirm the clinical applicability of these observations.This study has several limitations. First, we were unable to evaluate trimester-specific GWG or GWG after GDM diagnosis. Some perinatal outcomes may be more strongly influenced by mid-to-late pregnancy GWG than early pregnancy GWG.24–27 Furthermore, GWG was standardized to an estimated value at 40 weeks, assuming a linear trajectory in late pregnancy. However, weight gain patterns may change after GDM diagnosis owing to lifestyle or treatment interventions, potentially leading to exposure misclassification. Second, detailed information on GDM severity, glycemic control, and treatment was unavailable. More severe hyperglycemia is associated with an increased risk of a broad range of perinatal outcomes examined in this study. Insulin therapy may contribute to excessive weight gain,28 which in turn could increase the risk of outcomes such as fetal overgrowth, HDP, and cesarean delivery. However, it may also mitigate hyperglycemia-related perinatal complications by improving glycemic control. Differences in disease severity and treatment intensity may influence both GWG and the risk of adverse perinatal outcomes, and residual confounding cannot be excluded. Third, we could not assess the long-term risks for mothers and offspring. Excessive GWG has been linked to postpartum weight retention and childhood obesity,3 4 and long-term effects warrant further investigation. Fourth, for hypertensive outcomes, the observed associations with GWG may reflect reverse causation, as associations between fluid retention and these conditions may result in increased maternal weight gain. Therefore, these findings should be interpreted as associations rather than causal effects. Fifth, because data from small obstetric clinics were not included, the study population may have been skewed toward higher-risk pregnancies. Moreover, in this multicenter study, variations in clinical practice across institutions, including differences in GWG counseling and GDM management, may have introduced residual heterogeneity. Sixth, most participants were Japanese women, which may limit the generalizability of our findings to other populations. Finally, owing to the small number of cases, we could not analyze pregnancies with pregestational diabetes mellitus. Pregestational diabetes mellitus may similarly modify the association between GWG and perinatal outcomes, warranting further investigations.Conclusion We identified significant interactions among BMI categories, GDM status, and GWG in the development of perinatal outcomes. These findings underscore the need for GWG recommendations that are tailored to BMI categories and GDM status.",
  "title": "Differences in association between gestational weight gain and perinatal outcomes according to pre-pregnancy body mass index categories and gestational diabetes mellitus status: findings from a large Japanese cohort",
  "uid": "de5989c9-1c5b-5873-a282-6e80a785d74a"
}