Complete regression results for Table 2. Dependent variable: serum 25-hydroxyvitamin D (ng/ml)
Regression model | Model 1† | Model 2 | Model 3 | Model 4† | Model 5† |
---|---|---|---|---|---|
Description | UV index | UV index – quartiles | UV index – median | UV index groups | Interaction with obesity |
Race/ethnicity (omitted: Non-Hispanic White) | |||||
Non-Hispanic Black | −7·47*** | −7·49*** | −7·87*** | −7·96*** | −8·01*** |
Hispanic | −3·41*** | −3·38*** | −3·39*** | −3·37*** | −3·32*** |
UV index | |||||
UV index 1 month before exam (continuous) | |||||
UV index 1 month before exam | 0·51*** | 0·66*** | |||
Non-Hispanic Black×UV index | −0·09 | ||||
BMI 25·0–30·0 kg/m2×UV index | −0·21 | ||||
BMI ≥ 30·0 kg/m2×UV index | −0·28* | ||||
Quartiles of UV index 1 month before exam (omitted: Quartile 1) | |||||
Quartile 2 | 0·41 | ||||
Quartile 3 | 1·45*** | ||||
Quartile 4 | 1·27*** | ||||
Non-Hispanic Black×quartile 2 | −0·82 | ||||
Non-Hispanic Black×quartile 3 | −0·62 | ||||
Non-Hispanic Black×quartile 4 | −0·62 | ||||
Half of UV index 1 month before exam (omitted: UV index <median) | |||||
UV index ≥median | 1·15*** | ||||
Non-Hispanic Black×UV index ≥median | −0·24 | ||||
Groups of UV index 1 month before exam (omitted: 6–7)† | |||||
UV index<3 | −3·15*** | ||||
3 ≤ UV index<4 | −2·50*** | ||||
4 ≤ UV index<5 | −2·23*** | ||||
5 ≤ UV index<6 | −0·88** | ||||
7 ≤ UV index<8 | −0·02 | ||||
UV index ≥8 | 0·39 | ||||
Demographics | |||||
Age | −0·05*** | −0·05*** | −0·05*** | −0·05*** | −0·05*** |
Female | −0·26 | −0·27 | −0·27 | −0·25 | −0·30 |
Married | 0·81*** | 0·83*** | 0·83*** | 0·81*** | 0·87*** |
Annual household income (omitted: <$US 25 000) | |||||
$US 25 000–34 999 | 0·16 | 0·15 | 0·14 | 0·17 | 0·20 |
≥$US 35 000 | 0·44* | 0·42* | 0·43* | 0·44* | 0·49* |
Skin type | |||||
Hair colour (omitted: Black) | |||||
Red | 1·78* | 1·80* | 1·80* | 1·72* | 1·84* |
Blonde | 2·25*** | 2·23*** | 2·25*** | 2·24*** | 2·30*** |
Light brown | 2·10*** | 2·11*** | 2·12*** | 2·09*** | 2·12*** |
Medium brown | 1·69*** | 1·66*** | 1·68*** | 1·67*** | 1·70*** |
Dark brown | 1·11*** | 1·08*** | 1·11*** | 1·06*** | 1·17*** |
Other | 2·65 | 2·79* | 2·82* | 2·68 | 2·70 |
Missing | 2·07*** | 2·09*** | 2·11*** | 2·03*** | 2·08*** |
Skin reaction to 1/2 h of sun exposure without sunscreen or protective clothing (omitted: none) | |||||
Severe sunburn with blisters | −0·18 | −0·18 | −0·19 | −0·18 | −0·19 |
Severe sunburn for a few days with peeling | −0·21 | −0·19 | −0·18 | −0·19 | −0·24 |
Mildly burned with some tanning | 0·91*** | 0·92*** | 0·92*** | 0·94*** | 0·94*** |
Turn darker without sunburn | 0·51* | 0·54* | 0·54* | 0·51* | 0·51* |
Other | −1·33 | −1·32 | −1·33 | −1·27 | −1·21 |
Missing | −0·59 | −0·63 | −0·62 | −0·60 | −0·49 |
Sun behaviours | |||||
How often stay in the shade on a sunny day, when outside for >1 h (omitted: sometimes) | |||||
Always | −1·95*** | −1·93*** | −1·94*** | −2·01*** | −2·17*** |
Most of the time | −1·69*** | −1·68*** | −1·68*** | −1·72*** | −1·77*** |
Rarely | 2·03*** | 2·01*** | 2·00*** | 2·05*** | 2·09*** |
Never | 1·46** | 1·42** | 1·42** | 1·48** | 1·53** |
Don’t go in the sun | 5·74 | 5·52 | 5·69 | 5·53 | 5·80 |
Missing | 6·92*** | 6·85*** | 6·97*** | 6·81*** | 6·53*** |
Number of sunburns in past year (omitted: zero) | |||||
1 | 1·16*** | 1·17*** | 1·17*** | 1·14*** | 1·26*** |
2 | 0·87*** | 0·87*** | 0·87*** | 0·86*** | 0·93*** |
≥3 | 0·61*** | 0·61*** | 0·62*** | 0·61*** | 0·66*** |
Missing | 5·13 | 5·08 | 5·10 | 4·91 | 5·50 |
How often wear a hat that covers face, ears and neck (omitted: never) | |||||
Always | 0·50 | 0·50 | 0·51 | 0·52 | 0·32 |
Most of the time | −0·01 | 0·03 | 0·04 | 0·00 | −0·03 |
Sometimes | 0·23 | 0·25 | 0·26 | 0·27 | 0·24 |
Rarely | −0·61 | −0·58 | −0·58 | −0·57 | −0·61 |
How often wear a long-sleeved shirt (omitted: never) | |||||
Always | −1·85* | −1·88* | −1·88* | −1·79* | −1·60* |
Most of the time | −2·09*** | −2·14*** | −2·18*** | −2·05*** | −1·95*** |
Sometimes | −0·49 | −0·53 | −0·53 | −0·43 | −0·48 |
Rarely | −0·93* | −0·93* | −0·94* | −0·91* | −0·82* |
How often wear sunscreen (omitted: never) | |||||
Always | 2·17 | 2·16 | 2·13 | 2·25 | 2·22 |
Most of the time | 1·88 | 1·87 | 1·85 | 1·95 | 1·87 |
Sometimes | 1·36 | 1·38 | 1·34 | 1·41 | 1·34 |
Rarely | 0·13 | 0·12 | 0·09 | 0·22 | 0·11 |
Sunscreen SPF used most often (omitted: <15) | |||||
15–29 | −0·45 | −0·38 | −0·36 | −0·49 | −0·53 |
≥30 | −0·72 | −0·67 | −0·66 | −0·75 | −0·80 |
Missing | 0·18 | 0·24 | 0·23 | 0·21 | 0·11 |
Missing data for hat, long-sleeved shirt, sunscreen | |||||
Missing | −10·54** | −10·39** | −10·54** | −10·14** | −10·49* |
Other personal behaviours or characteristics | |||||
Vitamin D supplementation during past 30 d (avg. IU/d) | 0·28*** | 0·28*** | 0·28*** | 0·28*** | 0·27*** |
Smoker | |||||
Every day | −1·26*** | −1·23*** | −1·23*** | −1·28** | −1·15*** |
Some days | 0·25 | 0·28 | 0·27 | 0·22 | 0·23 |
Missing | −0·33 | −0·30 | −0·31 | −0·33 | −0·34 |
BMI | |||||
BMI (kg/m2) | −0·28*** | −0·28*** | −0·28*** | −0·28*** | |
25·0–30·0 kg/m2 | −0·27 | ||||
≥30·0 kg/m2 | −2·51*** | ||||
Month/year of exam | |||||
Month (omitted: January) | |||||
February | −0·39 | −0·32 | −0·39 | −0·42 | −0·34 |
March | −1·11* | −0·97 | −0·90 | −1·25* | −1·16* |
April | 0·03 | 0·02 | 0·07 | −0·39 | 0·15 |
May | −0·58 | 0·68 | 0·68 | −1·25* | −0·50 |
June | 0·44 | 2·00*** | 2·04*** | −0·52 | 0·52 |
July | 1·98** | 3·16*** | 3·20*** | 1·33* | 2·12*** |
August | 2·19*** | 3·20*** | 3·24*** | 1·52** | 2·25*** |
September | 0·55 | 2·11*** | 2·22*** | −0·27 | 0·64 |
October | 0·44 | 1·96*** | 1·92*** | −0·36 | 0·55 |
November | 0·96 | 1·11* | 1·16* | 0·43 | 1·08* |
December | 1·23* | 1·30* | 1·38** | 0·94 | 1·29* |
Year (omitted: 2000) | |||||
2001–02 | −1·63*** | −1·63*** | −1·67*** | −1·43*** | −1·64*** |
2003–04 | −1·82 | −1·96 | −1·94 | −1·83 | −1·80 |
2005–06 | −0·86 | −1·04 | −1·05 | −0·79 | −0·87 |
Intercept | 30·60*** | 31·98*** | 32·16*** | 34·96*** | 23·49*** |
R 2 | 0·29 | 0·29 | 0·28 | 0·29 | 0·28 |
F | 84·89 | 80·04 | 84·73 | 80·88 | 81·12 |
n | 14 319 | 14 319 | 14 319 | 14 319 | 14 319 |
SPF, sun protection factor.
Robust standard errors were used.
*P<0·05, **P<0·01, ***P<0·001.
†Model presented in main article.