Table 4

Statistics for r=1000 bootstrapped binary logistic regression (GSM exposure coefficients: increase in risk per increase in log GSM (μW/m2)

SymptomB*BootstrapNormal
BiasSE95% percentile intervalsSE95% CI
LowerUpperLowerUpper
Fatigue0.3290.0120.0970.1550.5390.1020.1280.529
Irritability0.4110.0160.1100.2410.6700.1040.2070.615
Headache0.3580.0220.1390.1490.6880.1130.1370.578
Nausea0.3190.0130.1240.0990.5900.1180.0880.550
Appetite0.4560.0260.1340.2640.7840.1280.2050.707
Sleep0.3960.0220.1240.1930.6900.1090.1810.610
Depression0.3460.0120.1020.1740.5830.1000.1510.541
Attention0.4290.0200.1180.2540.7110.1060.2220.636
Memory0.2370.0090.0980.0570.4480.0910.0580.415
Skin0.2170.0080.1100.0110.4510.1100.0010.433
Visual0.2030.0040.0930.0370.3980.0900.0260.379
Hearing−0.05−0.0020.089−0.2190.1430.093−0.2280.135
Vertigo0.3060.0100.1010.1270.5300.1020.1070.505
Walking−0.05−0.0060.098−0.2650.1200.098−0.2460.138
Vascular0.2740.0100.1090.0840.5200.1140.0510.497
  • Asymptotic SEs and 95% CIs are also shown for comparison.

  • *β coefficient (log OR).

  • GSM, Global System for Mobile Communication.