Skip to main content
. 2013 Apr 8;13:316. doi: 10.1186/1471-2458-13-316

Table 3.

Estimates from the population-averaged GEE model investigating associations with vitamin D level

Parameter Estimate Standard error P-value Overall p-value
Intercept
64.5883
2.4201
<.0001
<.0001
Age (years)
 
 
 
<.0001
  age_RCS_lin
−0.2738
0.0422
<.0001
 
  age_RCS_S1
0.0015
0.0001
<.0001
 
  age_RCS_S2
−0.0030
0.0004
<.0001
 
  age_RCS_S3
0.0008
0.0004
0.0823
 
Median household income ($CDN 10000)
 
 
 
<.0001
  income_RCS_lin
−0.4078
0.2350
0.0826
 
  income_RCS_S1
0.0543
0.0263
0.0392
 
  income_RCS_S2
−0.1270
0.0793
0.1092
 
  income_RCS_S3
0.0824
0.0855
0.3352
 
Percentage of immigrants
 
 
 
<.0001
  percentimmi__RCS_lin
−7.6190
9.4844
0.4218
 
  percentimmi__RCS_S1
−172.823
567.4589
0.7607
 
  percentimmi__RCS_S2
−1122.79
1547.330
0.4681
 
  percentimmi__RCS_S3
3152.879
1511.398
0.0370
 
Percentage of aboriginal people
 
 
 
<.0001
  percentabor__RCS_lin
−68.5534
15.3580
<.0001
 
  percentabor__RCS_S1
13868.36
5311.648
0.0090
 
Percentage of subjects with some post-secondary education
 
 
 
<.0001
  percentuniv_RCS_lin
13.6596
4.2690
0.0014
 
  percentuniv_RCS_S1
237.8126
69.0070
0.0006
 
  percentuniv_RCS_S2
−842.990
263.7487
0.0014
 
  percentuniv_RCS_S3
1090.294
509.4453
0.0323
 
Gender
 
 
 
<.0001
  Female
7.4674
0.1930
<.0001
 
  Male
0.0000
0.0000
.
 
Month of vitamin D testing
 
 
 
<.0001
  December
−1.3870
0.5024
0.0058
 
  November
0.1357
0.4583
0.7672
 
  October
−1.3138
0.4593
0.0042
 
  September
−0.1319
0.4590
0.7738
 
  August
4.7383
0.4252
<.0001
 
  July
3.7308
0.4142
<.0001
 
  June
0.5034
0.3959
0.2035
 
  May
3.5073
0.4330
<.0001
 
  April
0.6403
0.4129
0.1209
 
  March
−3.6969
0.4045
<.0001
 
  February
0.4483
0.4108
0.2751
 
  January 0.0000 0.0000 .  

Note: Age, median household income, percentage of immigrants, and percentage of subjects with some post-secondary education were coded using restricted cubic spline (RCS) functions with five knots located at the 5th, 25th, 50th, 75th, and 95th percentiles of their distribution, and percentage of aboriginal people was coded using the 5th, 50th and 95th percentiles of its distribution. The variables listed underneath each of the continuous variables are their corresponding spline variables (e.g., age_RCS_lin, age_RCS_S1, age_RCS_S2, and age_RCS_S3 are the 4 spline variables of age).