Skip to main content
. 2018 Nov 20;18:1273. doi: 10.1186/s12889-018-6168-0

Table 3.

Multiple linear regression analysis showing the association between coffee drinking and T-scores among study participants

Variables β P-value
Coffee drinking (Reference: No)
 Medium 0.049 0.3523
 High 0.158 0.0038*
P-trend 0.0046*
Age group (Reference: 30–49 years)
 50–69 − 0.447 <.0001*
  ≥ 70 − 0.619 <.0001*
Sex (Reference: Women)
 Men −0.200 0.0008*
Education (Reference:=College and above)
 Senior high school −0.140 0.0131*
 Junior high school −0.257 0.0001*
 Elementary and below −0.312 <.0001*
BMI (Reference: Normal)
 Underweight −0.314 0.0141*
 Overweight 0.181 0.0021*
 Obese 0.223 0.0049*
Smoking (Reference: Never)
 Quit −0.017 0.8427
 Yes −0.094 0.1546
Drinking (Reference: Never)
 Quit −0.038 0.7940
 Yes 0.012 0.8607
Exercise (Reference: No)
 Yes 0.096 0.0305*
Vegetarian diet (Reference: No)
 Yes −0.001 0.9729
Personal disease history
 Diabetes −0.185 0.0817
 Hypertension 0.037 0.5663
 Heart disease 0.038 0.7150
 Hyperlipidemia −0.016 0.8530
 Stroke 0.349 0.1682
Supplement intake
 Vitamin D (Reference: No)
  Yes 0.006 0.9631
 Other vitamins (Reference: No)
  Yes 0.104 0.2160
 Calcium (Reference: No)
  Yes −0.105 0.0852

*P< 0.05. Adjusted for age, education, BMI, smoking, drinking, exercise, vegetarian diet, vitamin D, other vitamins, calcium, blood type, fasting blood glucose, creatinine, uric acid, total cholesterol, triglycerides, HDL, LDL, GOT, GPT, SBP, DBP, waist circumference, waist-hip ratio, body fat, diabetes, hypertension, heart disease, hyperlipidemia and stroke