. 2022 Feb 8;8(2):e08915. doi: 10.1016/j.heliyon.2022.e08915

Table 3.

A summary of included studies on pregnancy outcomes associated with EMF exposure.

Authors and Year	Country	Design and Sample size	Results	Study Limitations	Rating
Authors and Year	Country	Design and Sample size	Results	Study Limitations	Strength/Quality
Tsarna et al. (2019)	4 studies were included from Denmark, South Korea, Spain, and the Netherlands.	A meta-analysis, longitudinal design N = 55507 pregnant women GA^a = Not identified	The risk of giving birth with small GA^a > in the intermediate exposure group than non-exposure group (hazard ratio = 1.04, 95% CI^c: 1.01, 1.07). Mobile phone use was associated with short pregnancy duration (P < 0.001) and PTB^d (P = 0.003).	Self-report measures	III/B
Lu et al. (2017)	Japan	A cross-sectional correlational design N = 461 mother-child pairs GA^a = <27 Wk^b.	Heavy use of mobile devices during pregnancy is associated with high premature birth PMB^d, NICU^e, and small birth chest circumference (P < 0.05).	Small sample size	III/A
Saadia (2018)	Pakistan	Cross sectional correlational N = 69 women (22 obese & 47 non obese) GA^a = 37–38 Wk^b.	The fetal heart rate variability value was 1.28 among women who were non-mobile phone users and 1.53 among women who were mobile phone users (P = 0.017, BMI<30).	Interview biases	III/B
Li et al. (2017)	California	A prospective cohort study. N = 913 pregnant women GA^a = <10 Wk^b.	Women who were exposed to higher EMF^f levels (>2.5 mG) had 2.72 times the risk of miscarriage (hazard ratio = 2.72, 95% CI^c: 1.42–5.19) than those with lower EMF^f exposure (<2.5 mG).	Interview biases	III/A
Karuserci et al. (2019)	Turkey	A cross-sectional correlational N = 400 women who have a child aged between 0-5 (boys 199 & 201girls) GA^a = 27 to 41 Wk^b.	HC^g < among infants whose mothers watched TV for more than 6 h (watching TV: 35.8 ± 2.8 cm (28–43 cm); not watching TV; 37.0 ± 2.3 cm (33–41 cm); p = 0.040. HC^g < among mothers based on a nearby base station (no base station near the home: 36.4 ± 2.6 cm (28–43 cm); base station near the home: 35.5 ± 2.7 cm (30–41 cm); p = 0.041).	Use self-report measures	III/B
Birks et al. (2017)	Denmark, Spain, Netherlands, Norway, and Korea	A meta-analysis, prospective and retrospective cohort designs N = 83,884 mother-child pairs GA^a = Not determined	The risk of having children with hyperactivity/inattention problems increased among mothers who used cell phones during pregnancy (the odds ratio for mothers of medium and high users was 1.11, 95% CI^c: 1.01, 1.22 and odds ratio 1.28, 95% CI^c: 1.12, 1.48 respectively).	Recall biases	III/A
Takei et al. (2018)	Japan	A model calculation N = one pregnant mother GA^a = 13 to 26 Wk^b.	The maternal body absorbs 0.15, 0.11, and 0.35 of magnetic waves per kg at GA^a 13, 18, and 26 wks^b respectively. -Fetal temperatures > by 0.025, 0.030, and 0.017 (K) in GA^a 13, 18, and 26 wks^b respectively.	Use one case model	V/B
Zarei et al. (2019)	Iran	A cross-sectional design N = 75 mothers of healthy children (control group) and 110 mothers of children aged 3–7 years diagnosed with speech problems GA^a = Not determined	Associations were found between the use of cordless phones and offspring speech problems for both before pregnancy and during pregnancy maternal exposure (P = 0.005 and P = 0.014, respectively).	Small sample size	III/C
Yuskel et al. (2016)	Turkey	An experimental design N = 32 rats and their 40-newborn offspring GA^a = NA	Progesterone, prolactin, and estrogen levels in the 900, 1800, and 2450 MHz/1hr groups were lower compared to the control group (p < 0.001, p < 0.05, and p < 0.05 respectively). Estrogen levels were lower in the 2450 MHz group compared to those in the 900 and 1800 MHz groups (p < 0.05).	Use animal subjects	I/A

^a GA = gestational age, ^b wks = weeks, ^c CI = Confidence Interval, ^d PMB = premature birth, ^e NICU = Neonatal Intensive Care Unit, ^f EMF = electromagnetic field, ^g HC = head circumference.