Skip to main content
. 2019 Aug 9;16(8):e1002868. doi: 10.1371/journal.pmed.1002868

Table 1. Characteristics of the sons and parents according to quintiles of sons’ BMI.

 Subject  Variable Quintile of sons’ BMI Mean difference or odds ratio (95% CI) N
1st 2nd 3rd 4th 5th
Sons
BMIa,b (kg/m2) 18.5 20.2 21.3 22.7 26.2 2.90 (2.90–2.90) 1,036,817
Heighta,b (cm) 179.6 179.4 179.3 179.1 179.0 –0.19 (−0.20 to −0.18) 1,036,817
Smokersb,c (%) 64% 61% 59% 57% 56% 0.91 (0.89–0.93) 29,508
BMIa,b (kg/m2) 20.3 20.9 21.2 21.6 22.2 0.62 (0.60–0.64) 73,982
Heighta,b (cm) 178.4 178.4 178.2 178.2 178.2 –0.09 (−0.13 to −0.05) 74,001
Fathers Date of birtha 1936.4 1935.3 1934.9 1934.9 1935.8 0.003 (−0.02 to 0.03) 1,025,639
Smokersb,c (%) 62% 64% 62% 64% 67% 1.09 (1.06–1.12) 17,254
Age at sons’ birth (years)a 30.5 30.5 30.5 30.5 30.5 –0.003 (−0.02 to 0.01) 1,025,639
Educated >10 yearsc (%) 57% 56% 55% 53% 48% 0.88 (0.88–0.88) 973,704
In nonmanual workc (%) 51% 51% 50% 49% 42% 0.86 (0.86–0.87) 858,447
Mothers
Date of birth 1939.4 1938.4 1938.0 1938.0 1939.0 0.05 (0.03–0.07) 1,035,861
Age at sons’ birth (years)a 27.4 27.4 27.4 27.4 27.3 –0.05 (−0.06 to −0.04) 1,035,861
Educated >10 yearsc (%) 56% 55% 55% 53% 50% 0.92 (0.91–0.92) 1,016,365
In nonmanual workc (%) 50% 50% 49% 47% 42% 0.88 (0.88−0.88) 753,046

aContinuous variables are summarised as means in each quintile and linear regression produced mean differences per SD (2.90 kg/m2) of BMI preadjusted for each son’s age at examination, conscription office, and secular trends (date of birth).

bMeasured at preconscription medical examination. Smoking was only recorded at examinations in 1969 through 1970.

cBinary variables are summarised as percentages in each quintile and logistic regression produced odds ratios per SD (2.90 kg/m2) of BMI preadjusted for each son’s age at examination, conscription office, and secular trends (date of birth).

Abbreviations: BMI, body mass index; CI, confidence interval; SD, standard deviation