Search results, study selection and characteristics
The search strategy yielded 267 potentially relevant articles. Based on selection inclusion criteria, 18 articles were identified for further full-text assessment; finally, we included 15 articles, which included data of 68956 pregnant women. Table 1 presents a summary of studies, assessing adverse pregnancy outcomes among women with IMH.
Table 1
Characteristics of studies included in the Systematic review
First author (year); Country | Study design | Gestational age of IMH assessment | IMH definition | Sample size | Prevalence of IMH (%) | Feto-maternal outcome in women with vs. without IMH (%) | Neonatal outcome in women with vs. without IMH (%) | Significant associations between IMG and adverse outcomes |
Pop et al. (2004); Netherlands | prospective community-based cohort | First trimester and 24–32 weeks of gestation | TSH: 0.15–2.0 mIU/L) fT4 < 10th percentiles (12.4 pmol/L) | 1361 | 9.9% | Breech presentation: 9 vs. 2 | - | First trimester: Breech presentation: OR : 4.7 (95% CI:1.1–19) |
Casey et al. (2007); USA | prospective cohort | < 20 weeks of gestation | -TSH: 2.5th -97.5th (0.08 − 2.99 mU/L) -fT4 < 2.5th (0.86 ng/dL) | 17298 | 1.3% | gestational HTN: 8 vs.9; Severe preeclampsia: 4 vs 5; Diabetes: 4 vs 5; Placental abruption: 0.4 vs. 0.3; Preterm Delivery ≤ 36 w: 6 vs. 6; Preterm Delivery ≤ 34 w: 2 vs. 2.5; Preterm Delivery ≤ 32 w: 1 vs. 1; C/S: 25 vs. 25 | VLBW: 0.4 vs. 0.6; LBW: 3 vs. 6; macrosomia: 13 vs. 11; NICU: 1.3 vs. 2.2 ; 5-Min Apgar score ≤ 3: 0 vs. 0.7; umbilical artery blood pH < 7.0: 3 vs. 1.7 respiratory distress syndrome: 1.3 vs. 1.5; necrotizing enterocolitis: 0 vs. 0.1 intraventricular hemorrhage: 0.4 vs. 0.1; major malformations: 0.4 vs. 1.1; fetal death: 0 vs. 0.5; neonatal death: 0 vs. 0.2 | - |
Cleary-Goldman et al., (2008); USA | prospective cohort | First and second trimesters | -TSH: 2.5th-97.5th percentiles -fT4 < 2.5th (0.86 ng/dL) | 10990 | First trimester: 2.1% Second trimester 2.3% | First trimester: Miscarriage: 0 vs. 0.6; Gestational HTN: 0.4 vs. 5.5; Preeclampsia: 1.3 vs. 1; GDM: 6.2 vs. 3; Placenta previa: 0.4 vs. 0.4; Placental abruption: 1.8 vs. 0.9; Preterm labor: 8.4 vs. 6.1; Preterm PROM: 1.3 vs. 1.4; Preterm delivery: 9.3 vs. 7.2 Second trimester Miscarriage: 0 vs. 0.6; Gestational HTN: 9.5 vs. 5.4; Preeclampsia: 1.7 vs. 1; GDM: 7.4 vs. 3; Placenta previa: 0 vs. 0.4; Placental abruption: 0.4 vs. 0.9; Preterm labor: 7 vs. 6.2; Preterm PROM: 1.7 vs. 1.4; Preterm delivery: 9.1 vs. 7.3 | First trimester: LBW: 2.7 vs. 4.2; Macrosomia: 16.9 vs. 8.9; Perinatal mortality: 0.4 vs. 0.3 Second trimester LBW: 4.6 vs. 4.2; Macrosomia: 13.6 vs. 8.9; perinatal mortality: 0 vs. 0.3 | First trimester: Preterm labor: OR: 1.6 (95% CI: 1.00–2.62); Macrosomia: OR 1.97 (95% CI: 1.37–2.83). Second trimester GDM: OR: 1.7 (95% CI: 1.02–2.84). |
Hamme et al., (2009); Canada | prospective cohort | Second trimester | -TSH: 0.15–4.0 mU/L -f T4 ≤ 10th (8.5 pmol/L) | 879 | 10.1% | preterm delivery: 5.2 vs. 7.2 | SGA: 5.6 vs. 7.6 Apgar score < 7: 0 vs. 0 | - |
Mannisto et al,. (2010); Finland | prospective population-based cohort | First trimester | -TSH 5th – 95th percentiles -fT4 < 5th (11.96 pmol/L) | 5805 | 3.9% | Gestational HTM: 3.3 vs. 3; Preeclampsia: 1.4 vs. 1.9 GDM: 0 vs. 1; Placental abruption: 0.5 vs. 0.5; Maternal weight gain ˃20 kg: 8.7 vs. 9.8 | - | - |
Su et al., (2011); China | prospective population-based cohort | < 20 weeks of gestation | -TSH 5th – 95th percentiles fT4 < 5th (11.96 pmol/L) | 1017 | 2.9% | Spontaneous abortions: 0 vs. 1.3; Fetal deaths: 0 vs. 0.5; Fetal loss: 0 vs. 2.2; Medically induced labor: 0 vs. 0.5; Preterm births: 2.3 vs. 4.1 | Neural malformations: 0 vs. 0.4; Eye, ear, face malformations: 0 vs. 0.6; Circulation malformations: 4.7 vs. 1.3; Reproductive malformations: 0 vs. 0.1; Musculoskeletal malformations: 4.7 vs. 0.7; Other malformations: 0 vs. 1.2; Total malformations: 4.7 vs. 4.4; Fetal distress: 11.6 vs. 1.7; LBW: 2.3 vs. 1.8; Macrosomia: 7 vs. 9.9; SGA: 7 vs. 2.3; Neonatal death: 0 vs. 0.1; Poor vision development: 2.3 vs. 1; Hearing dysplasia: 0 vs. 0.5; Neurodevelopmental delay: 0 vs. 0.2 | fetal distress: OR:2.95 (95% CI:1.08–8.05) SGA: OR: 3.55 (95% CI:1.01–12.83) Musculoskeletal malformations: OR = 9.12 (95% CI:1.67– 49.70) |
Korevaar et al. (2013); Netherlands | prospective population-based cohort | Early pregnancy | -TSH: 2.5th -97.5th percentiles -fT4 < 2.5th (10.4 pmol/L) | 5971 | 2.6% | Preterm delivery < 37 w: 10.3 vs. 4.7; Preterm delivery < 34 w: NM; Spontaneous preterm delivery < 37 w: NM; Spontaneous preterm delivery < 34 w: NM; PROM < 37 w: 7.2 vs. 3.7; Spontaneous PROM < 37 w: NM | - | Preterm delivery < 37 w: OR = 2.54 (95% CI: 1.42– 4.54)λ Preterm delivery < 34 w: OR = 3.56 (95% CI: 1.50–8.43) λ Spontaneous preterm delivery < 37 w: OR = 3.44 (95% CI: 1.76–6.70) λ Spontaneous preterm delivery < 34 w: OR = 4.21 (95% CI: 1.34 − 13.3) λ PROM < 37 w: OR = 2.35 (95% CI: 1.18–4.69) λ Spontaneous PROM < 37 w: OR = 2.74 (95% CI: 1.30 − 5.75) λ |
Breathnach et al. (2013); Ireland | Cohort | < 20 weeks of gestation | -TSH 2.5th – 97.5th percentiles -fT4 < 2.5th | 904 | IMH: 1.9% | Placenta abruption: NM; Gestational HTN: NM; Preterm PROM: NM; GDM: NM; Preterm Birth: NM; IUGR: NM | Macrosomia: NM | Placenta abruption: P-value: 0.04 GDM: P-value: 0.001 |
Medici et al., (2014); Netherlands | prospective population-based cohort | Early pregnancy | -TSH 2.5th -97.5th percentiles -fT4 < 2.5th (10.4 pmol/L) | 5153 | NM | Hypertensive Disorders overall: 6.4 vs. 6.2; gestational HTN: 4.7 vs. 4; preeclampsia: 3.8 vs. 2.5 | - | - |
Ong et al., (2014); Australia | Cohort | First trimester | -TSH: 2.5th – 97.5th percentiles (0.02–2.15 mU/L) -fT4 < 10th (11.5 pmol/L) | 2411 | 10.1% | placenta previa: NM; placental abruption: NM preeclampsia: NM; pregnancy loss after 20 w: NM; preterm labor: NM; preterm birth: NM | SGA: NM Neonatal death: NM; birth defects: NM | - |
Leon et al., (2015); Spain | prospective population-based cohort | < 24 weeks of gestation | -TSH 5th – 95th -fT4 < 5th | 2170 | 2.3% | Preterm delivery: NM | Mean birth weight: NM; SGA:NM; LGA: NM | higher birthweight: β = 109 (95% CI: 31–187) |
Zhu et al., (2018); China | prospective population-based cohort | First and second trimesters | -TSH 2.5th – 97.5th percentiles -fT4 < 2.5th | 3178 | First trimester: 2.4% Second trimester 2.4% | - | First trimester: SGA:NM; LGA:NM Second trimester SGA:NM; LGA:NM | Second trimester LGA: OR = 2.088 (95% CI:1.193–3.654)** |
Rosario et al., (2018); Brazil | prospective cohort | First trimester | Three criteria: -TSH: 0.1–2.5 mIU/l and 1. fT4 < 5th (0.86 ng/dL) 2. fT4 < 10th (0.92 ng/dL) 3. Total T4 < 7.8 ng/dL | 596 | Criteria 1: 4.3% Criteria 2: 9% Criteria 3: 7% | Criteria 1: Gestational HTN: 7.7 vs. 8.6; GDM: 11.5 vs. 10.3; placental abruption: 0 vs. 0.3; Preterm delivery < 37 w: 3.8 vs. 2.7; Preterm delivery < 34 w: 0 vs. 0.9; Fetal loss: 0 vs. 2.9 Criteria 2: Gestational HTN: 9.2 vs. 8.6; GDM: 11.1 vs. 10.3; placental abruption: 0 vs. 0.3; Preterm delivery < 37 w: 3.7 vs. 2.7; Preterm delivery < 34 w: 1.8 vs. 0.9; Fetal loss: 3.7 vs. 2.9 Criteria 3:€ Gestational HTN: 9.5 vs. 9; GDM: 12 vs. 10.8; placental abruption: 0 vs. 0.3; Preterm delivery < 37 w: 4.7 vs. 2.8; Preterm delivery < 34 w: 2.4 vs. 0.9; Fetal loss: 4.7 vs. 2.8 | Criteria 1: Birth weight < 2500 g: 3.8 vs. 5.9; Birth weight < 1500 g: 0 vs. 1.4; NICU: 0 vs. 2 Ventilation > 24 h: 0 vs. 2; NEC: 0 vs. 0.2 ; IVH grade 3 or 4: 0 vs. 0; Malformations: 0 vs. 0.9 ; Neonatal death: 0 vs. 0.5 Criteria 2: Birth weight < 2500 g: 3.7 vs. 5.9; Birth weight < 1500 g: 1.8 vs. 1.4; NICU: 1.8 vs. 2 Ventilation > 24 h: 1.8 vs. 2; NEC: 0 vs. 0.2; IVH grade 3 or 4: 0 vs. 0; Malformations: 1.8 vs. 0.9 ; Neonatal death: 0 vs. 0.5 Criteria 3: Birth weight < 2500 g: 4.7 vs. 5.9; Birth weight < 1500 g: 2.3 vs. 1.6; NICU: 2.3 vs. 2; Ventilation > 24 h: 2.3 vs. 2; NEC: 0 vs. 0.2; IVH grade 3 or 4: 0 vs. 0; Malformations: 0 vs. 1; Neonatal death: 0 vs. 0.5 | - |
Gong et al. (2019); China | prospective cohort | First and second trimester | -TSH 2.5th–97.5th percentile -fT4 < 2.5th (13.35 pmol/L) | 3398 | First trimester: 7.3% Second trimester: 18.8% | First trimester: Miscarriage: 5.7 vs. 7.4; gestational HTN: 4.7 vs. 3; eclampsia: 0.9 vs. 0.6; GDM: 11.3 vs. 15.9; placental abruption: 0 vs. 0.2; PROM: 0 vs. 0.5; preterm delivery: 1.9 vs. 3.2; breech delivery: 6.6 vs. 4.2; Second trimester Miscarriage: 0.8 vs. 0.8; gestational HTN: 8.3 vs. 3.5; eclampsia: 0 vs. 1.2; GDM: 19 vs. 14.8; placental abruption: 0 vs. 0.4; PROM: 1.7 vs. 0.6; preterm delivery: 3.3 vs. 3.3; breech delivery: 4.1 vs. 3.3 | First trimester: LBW: 1.9 vs. 2; Macrosomia: 14.2 vs. 10.5 Second trimester LBW: 1.7 vs. 2.5; Macrosomia: 21.5 vs. 13.4 | Second trimester gestational HTN: P-value: 0.019 OR: 4.2 (95% CI: 1.61–10.97) µ Macrosomia: P-value: 0.024 |
Su et al. (2019);China | hospital-based Retrospective cohort | < 20 weeks of gestation | -TSH: 0.06–3.83 mIU/L -fT4 < 2.5th (1.01 ng/dL) | 8173 | 4.18% | GDM: 18.2% vs. 13.3%; gestational HTN: 5.9% vs. 2.8%; Preeclampsia: 1.2 vs. 0.9; preterm delivery: 4.7% vs. 4.2%; placenta previa: 1.2% vs. 0.4; placenta abruption: 0.9% vs. 0.8% | LBW: 3.2 vs. 2.7; Macrosomia: 9 vs. 5.8; SGA: 3.8 vs. 4.4; LGA:18.4 vs. 13.1 | gestational HTN: RR: 2.21 (1.28–3.82)* Macrosomia: RR: 1.64 (1.01–2.67)* |
IMH: Isolated maternal hypothyroxinemia. SCH: subclinical hypothyroidism; SGA: small for gestational age; LGA: large for gestational age; GDM: gestational diabetes mellitus; PROM: Premature rupture of membranes; LBW: low birth weight; IUGR: Intrauterine growth retardation; PIH: pregnancy induced hypertension; C/S: cesarean section; NEC: Necrotizing enterocolitis; IVH: Intraventricular hemorrhage; HTN: hypertension; NM: Not mentioned. Bold indicates statistical significance, P < 0.05. * Adjusted for BMI, health insurance, gravidity, parity, family history of chronic disease and newborn sex €: Compared to TT4 ≥ 7.8 # Adjusted for maternal age, prior pregnancy, body mass index, and study site. ð Adjusted for cohort, maternal age, country of origin, employed during pregnancy, maternal and paternal height, maternal body mass index, parity, weight gain during pregnancy, smoking during pregnancy, and season of delivery. ** Adjusted for maternal age, paternal age, pre-pregnant BMI, gestational age, metabolic dysfunctions, parity, birth type, GWG and fetal gender λ adjusted for gestational age at blood sampling, maternal age, smoking, SES, parity, ethnicity, maternal BMI, maternal height and child sex µ adjusted for smoking, passive smoking, alcohol, GW, AC, SBP, DBP, HR, TSH, maternal education, social-economic status, multiparous |
Details of the quality assessment of studies included are presented in supplementary table 2. This assessment showed that 11 studies were classified as being of high quality (6, 14, 15, 20–27) and four had moderate quality (11, 16, 28, 29). In addition, cohort studies had a low risk of bias for selection of exposed and non-exposed cohorts, assessment of exposure, presence of outcome of interest at start of study, outcome assessment, and adequacy of follow up of cohorts; however, approximately 30% had a problem risk of bias in the domain of control of prognostic variables, 13% in existence of outcome at start of study and 7% in outcome evaluation (Supplementary Fig. 1).
The articles were published in various geographical region: North America (15) and USA (14, 20)], South America [Brazil (16)], Europe [Netherland (21, 24, 29), Spain (22), Finland (23) and Ireland (28)] and Asia / Australia [China (6, 11, 26, 27) and Australia (25)]. All studies were prospective or retrospective cohorts and 47% (7/15) had a population-based design (21–23, 26, 27, 29, 30). In five studies, IMH was diagnosed in the first trimester (16, 21, 23–25), 5 in the first and second trimesters, before 20–24 weeks of gestations (11, 14, 22, 26, 28), 4 in the both first and second trimesters, separately (6, 20, 27, 29), one study in only in second trimester (15).
The prevalence of IMH among included studies in the first and second trimesters of pregnancy varied widely and ranged from 1.3% (14) to 18.8% (6), although, its prevalence in epidemiological data of population based studies included were less sparse, ranging between 2% -3% (21–23, 26, 27).
Diagnostic criteria used in studies included were quite variable and heterogeneous. In this respect, in terms of TSH, 8 studies used population-derived 2.5th − 97.5th (6, 14, 20, 21, 24, 25, 27, 28) percentiles as the TSH reference interval for diagnosis and 3 studies used the population‐derived of 5th − 95th percentiles (22, 23, 26). Two studies used the ATA 2017 fixed ranges of 0.05-4 mIU/L (11, 15) and two study used the ATA 2011 fixed ranges of 0.1–2.5 mIU/l during pregnancy (16, 29). Regarding fT4, the cut point of fT4 also varied between studies. Three studies applied the population‐derived ˃10th percentile (15, 25, 29), Three studies used the population‐derived ˃5th percentile (22, 23, 26) and also Eight studies used the population‐derived ˃2.5th percentile (6, 11, 14, 20, 21, 24, 27, 28), and one study used the three criteria of the population‐derived ˃10th and ˃5th percentiles as the fT4 cut point and also total T4 < 7.8 ng/dL for diagnosis of IMH (16).
Feto-maternal Outcomes
The association between IMH and feto-maternal outcomes, investigated by 14 studies (6, 11, 14–16, 20–26, 28, 29), had wide variations in amplitude of findings between studies included in this review. Regard this association, 11 studies examined the risk of preterm birth among women with IMH (6, 11, 14–16, 20–22, 25, 26, 28). The prevalence of preterm birth among women with IMH ranged between 2.3%-10.3%, although 90.9% of the studies included found no significant association between preterm birth and IMH, a prospective population-based cohort study with large sample size from Netherlands reported that maternal hypothyroxinemia in the first trimester of pregnancy was associated with a 2.5-fold increased risk of preterm birth (adjusted OR: 2.54, 95% CI: 1.42– 4.54), a 3.4-fold increased risk of spontaneous preterm birth (adjusted OR: 3.44, 95% CI: 1.76–6.70) and a 3.6-fold increased risk of early preterm birth before 34 week of gestations (adjusted OR: 3.56, 95% CI: 1.50–8.43) (all P ≤ .01) (21). However, one study reported that IMH in the first trimester was associated with preterm labor (OR: 1.62, 95% CI: 1.00–2.62) without increasing the risk of preterm birth (20) and one (21) of four studies (6, 20, 21, 28) evaluating the risk of preterm PROM, showed a positive association between IMH and preterm PROM (adjusted OR: 2.35, 95% CI: 1.18–4.69)
Of publications included, 6 evaluated the risk of GDM among women with hypothyroxinemia in first and second trimesters of pregnancy (6, 11, 16, 20, 23, 28) and reported that prevalence of GDM varied between 0-18.2% and 1-14.7% in women with and without IMH; of these studies, 4 found no association (6, 11, 16, 23), two reported that maternal hypothyroxinemia in the second trimester of pregnancy was significantly associated with a higher prevalence / risk of GDM compared to non-IMH counterparts (20, 28).
Nine studies investigated the association of maternal IMH and gestational HTN, preeclampsia and eclampsia (6, 11, 14, 16, 20, 23–25, 28). Neither preeclampsia nor eclampsia were associated with IMH diagnosed in first or second trimesters of pregnancy; in addition, all the above studies except for two (6, 11) found no significant association between maternal IMH and gestational HTN. Gong et al. (2019) however reported that IMH identified in the second trimester was associated with increased risk of only gestational HTN, particularly among women with BMI < 25 kg/m2, (adjusted OR: 4.2, 95% CI: 1.61–10.96) (6). Moreover, Su et al. (2019), showed that IMH was associated with a 2.2-fold increased risk of gestational HTN (adjusted OR: 2.2, 95% CI: 1.28–3.82) (11). Of 8 studies (6, 11, 14, 16, 20, 23, 25, 28) that assessed the association between maternal IMH and placenta abruption, all except one (28), showed no association between IMH and placenta abruption. Two studies assessed the risk of breech presentation in mothers with IMH (6, 29) and one (29) reported increased risk of breech presentation in women diagnosed with IMH in the first trimester of pregnancy (adjusted OR: 4.7, 95% CI: 1.1–19).
Moreover, there were no associations between maternal IMH and other adverse feto-maternal outcomes, including cesarean Sect. (14), miscarriage (6, 20, 26), placenta previa (11, 20, 25), maternal weight gain ˃20 kg (23), fetal deaths (26), fetal loss (16, 25, 26) or IUGR (28) among studies included.
Neonatal Outcomes
Among studies included, 6 examined the association between IMH and macrosomia (6, 11, 14, 20, 26, 28); 50% of these studies showed positive associations, indicating that the IMH diagnosed in the first (20), second (6) and < 20 weeks of gestation (11) was associated with around 1.5-fold increased risk of macrosomia. Furthermore, 2 other studies showed an increased risk of LGA and among IMH women in the second trimester (OR: 2.088, 95% CI: 1.193–3.654) (27) and significant higher birthweight (22) in the first half of pregnancy. Six studies assessed the risk of SGA among women diagnosed with IMH (11, 15, 22, 25–27), and Of just one (26) demonstrated that IMH was related to SGA (adjusted OR: 3.55, 95% CI:1.01–12.83). This study also showed that isolated hypothyroxinemia was associated with fetal distress (adjusted OR:2.95, 95% CI:1.08–8.05) and musculoskeletal malformations (adjusted OR:9.12, 95% CI:1.67– 49.70) (26).
However, IMH was not associated with other neonatal outcomes including NICU admission (14, 16), low Apgar score (14, 15), umbilical artery blood pH < 7 (14), RDS (14), necrotizing enterocolitis (14, 16), intraventricular haemorrhage (14, 16), major malformations (14, 16, 25, 26), perinatal mortality and neonatal death (14, 16, 20, 25, 26), or neurodevelopmental disturbances (26).