Table 2 Poisson regression results using quartile categories of explanatory variables.

From: Spatial analysis of the geographical distribution of thyroid cancer cases from the first-round thyroid ultrasound examination in Fukushima Prefecture

Variable name

unit

Exp (coefficient of explanatory variable)

P-value of Wald test (P-value for trend)

Residual Deviance

AIC

estimate

95% CI

Proportion of estimated external radiation does ≥ 1 mSv

proportion among surveyed people

Q1

Reference

(0.757)

46.554

132.62

Q2

1.169

(0.584, 2.340)

0.659

  

Q3

1.059

(0.508, 2.211)

0.878

  

Q4

1.149

(0.602, 2.196)

0.673

  

Distance from the FNPP

1 km

Q1

Reference

(0.370)

45.924

131.99

Q2

0.958

(0.631, 1.454)

0.840

  

Q3

0.945

(0.537, 1.664)

0.845

  

Q4

0.622

(0.221, 1.747)

0.368

  

Altitude

100 m

Q1

Reference

(0.623)

43.624

129.69

Q2

1.439

(0.964, 2.149)

0.075

  

Q3

1.183

(0.612, 2.286)

0.618

  

Q4

1.299

(0.612, 2.756)

0.495

  

Population density

1000 persons per square kilometre

Q1

Reference

(0.560)

45.536

131.60

Q2

1.159

(0.257, 5.227)

0.848

  

Q3

1.653

(0.390, 7.011)

0.495

  

Q4

1.357

(0.333, 5.520)

0.670

  

Proportion of workers in agriculture, forestry, and fisheries industry

proportion among workers

Q1

Reference

(0.970)

44.080

130.14

Q2

0.889

(0.546, 1.448)

0.638

  

Q3

0.547

(0.239, 1.252)

0.153

  

Q4

1.198

(0.486, 2.953)

0.695

  

Unemployment

proportion among labour force

Q1

Reference

(0.371)

43.628

129.69

Q2

0.944

(0.328, 2.722)

0.916

  

Q3

1.351

(0.488, 3.745)

0.562

  

Q4

1.428

(0.510, 4.001)

0.498

  

Proportion of professional and technical workers

proportion among workers

Q1

Reference

(0.995)

45.785

131.85

Q2

1.239

(0.346, 4.441)

0.742

  

Q3

0.880

(0.256, 3.021)

0.839

  

Q4

1.125

(0.356, 3.559)

0.841

  
  1. Each estimation result of Poisson regression model using quartile categories (Q1, Q2, Q3, and Q4) of one explanatory variable (shown in the 1st column) consists of four rows to report the estimated coefficients of the four quartile categories. For each model, the lowest quartile (Q1) is set as the reference category. Estimates of intercept terms were omitted. n = 59 (municipalities) for all of the models.
  2. A small Wald test p-value indicates that the coefficient is considerably different from that of the reference category. AIC, Akaike information criterion; FNPP, Fukushima Daiichi Nuclear Power Plant; N/A, not applicable; CI, confidence interval.
Back to article page