Table 2 Characteristics and extracted data of the included studies

Bilder et al./2013 [23]

Utah

ADDM Network cohort (Population-based)

ASD, n = 128

8

Intellectual disability, maternal and paternal age and education, parity, pre-pregnancy BMI, gestational age, gender, birth year

Continuous (based on a 5-lb increase)

DSM-IV-TR

Each 5 pounds of weight gained was significantly associated with ASD risk [1.10 (1.03–1.17)].

Control, n = 10,920/Singleton births

A 2 SD increase in GWG was associated with risk of ASD (25.00 lb).

Pre-pregnancy BMI was not associated with ASD.

Bilder et al./2013 [23]

Utah

Genetics Study cohort (research-based)

ASD, n = 288

8

Intellectual disability, maternal and paternal age and education, parity, pre-pregnancy BMI, gestational age, gender, birth year

Continuous (based on a 5-lb increase)

ADI-R and ADOS-G

Each 5 pounds of weight gained was significantly associated with ASD risk [1.17 (1.01–1.35)].

Unaffected siblings, n = 493/Singleton births

A 2 SD increase in GWG was associated with risk of ASD (24.74 lb).

Pre-pregnancy BMI was not associated with ASD.

Burstyn et al./2010 [24]

Canada

Cohort

218,890/Singleton births

4–10

Maternal age, weight, pre-pregnancy and gestational diabetes, bleeding, smoking, parity and socio-economic status, child’s birth year, gestational age, gender

Poor GWG: <  0.5 kg/week (26–36 weeks)

ICD-9

Poor weight gain (26–36 weeks, <  0.5 kg/week) was not associated with increased risk of ASD [0.95 (0.57–1.59)].

Low maternal pre-pregnancy weight (<  45 kg) was associated with increased risk of ASD [RR: 2.15 (1.20–3.85)].

Maternal weight > 91 kg was not associated with increased risk of ASD [1.18 (0.96–1.44)].

Dodds et al./2011 [25]

Canada

Retrospective, longitudinal Cohort

Total births:

1–17

Year of birth, genetic susceptibility (sibling with ASD, maternal psychiatric disorders, maternal neurological illness), major CNS anomaly, breastfeeding at discharge, infant sex, type of labour, maternal conditions (pulmonary disease, heart disease, renal disease and anemia), pre-pregnancy weight, gestational age.

Excess GWG: ≥18 kg

ICD-9 or ICD-10 codes

GWG ≥ 18 kg in total sample [RR: 1.19 (1.02–1.39)] and in mothers of children with low genetic susceptibility [1.21 (1.03–1.43)] was associated with increased risk of autism.

n = 129,733

Autism: n = 924/(adjusted for multiple births)

Poor GWG: < 7 kg

Inadequate GWG (<  7 kg) was not associated with the risk of autism.

Pre-pregnancy weight ≥ 90 kg in total sample [1.58 (1.26–1.98)] and in mothers of children with low genetic susceptibility [1.69 (1.34–2.14)] was associated with increased risk of autism.

Gardner et al./2015 [26]

Sweden

Prospective cohort

Total children:

≥ 4

Maternal BMI, gestational age, infant sex, birth year, parity, maternal age, paternal age, maternal country of birth, socioeconomic status, parental psychiatric history, genetic susceptibility, intellectual disability.

Based on IOM guidelines

ICD-9, ICD-10, and DSM-IV codes

Both insufficient [1.17 (1.04–1.31)] and excessive GWG [1.12 (1.01–1.25)] were independently associated with ASD.

n = 333,057

ASD: 6420/Singleton births

Every 2.3 kg (5 lb) increase in GWG was associated with increased risk of autism [1.03 (1.00–1.06)].

Matched sibling analyses showed elevated risk of ASD with excessive GWG [1.48 (0.93–2.38)].

Maternal overweight/obesity was associated with increased risk of offspring ASD [25 ≤ BMI < 30: 1.31, (1.21–1.41); BMI ≥ 30: 1.94 (1.72–2.17)].

Shen et al./2018 [28]

China

Case-control

Autism: n = 705

2–9

Child’s gender and age, parental age and family income

Based on Chinese GWG guidelines

DSM-IV-TR criteria

Excessive GWG was associated with autism risk in the entire sample [(1.327, 1.021–1.725)].

Control: n = 2236/Singleton births

Excessive GWG increased the risk of autism in overweight/obese mothers [2.468 (1.102–5.526)].

Inadequate GWG was not associated with the risk of autism.

Maternal pre-pregnancy BMI might not be independently associated with the risk of autism.

Windham et al./2019 [29]

USA

Case-control

ASD: n = 540

2–5

Pre-pregnancy BMI, maternal age, education, race/ethnicity, parity, smoking, intellectual disability, child sex

Based on IOM guidelines

SCQ score ≥ 11, ADI-R, ADOS

Maternal total GWG was higher in the ASD group than other groups.

SEED

Developmental delays: n = 720

Control: n = 776/Singleton births

Continuous model: 1.06 (1.02–1.10)

Exceeds IOM/ACOG recommendation: 1.29 (1.00–1.66)

Exceeding clinically recommended GWG was significantly higher (51%) in ASD group than other groups.

Associations of ASD with higher GWG were stronger in quintiles 4 and 5.

Quintile 4 (35– ≤ 44 pounds): 1.52 (1.05–2.22)

Quintile 5 (≥44 pounds): 1.58 (1.08–2.31)

In the continuous model, each 5-pound increase of GWG was associated with6% increased odds of ASD.

Insufficient GWG was not associated with risk of ASD.

Maternal pre-pregnancy BMI was not associated with ASD.

Xiang et al./2015 [30]

USA

Retrospective longitudinal cohort

322,323 children

A median of 5.5 years after birth

Birth year, maternal age, parity, gestational age, education, maternal race/ethnicity, household income, history of comorbidity (≥1 diagnosis of heart, lung, kidney, or liver disease; cancer), child sex

Continuous (per 4 kg)

ICD-9 codes 299.x or equivalent KPSC codes

GWG (per 4 kg) [HR: 1.67 (1.10–2.53)] and maternal pre-pregnancy BMI were modestly and positively associated with ASD risk.

ASD: n = 3388/Singleton births