恶人谷

Elysees wrote:倒地不起。。。这叫人如何是好啊。。。吃。。。还是不吃

The covariates chosen for inclusion in the multivariable
model were based on previously identified risk factors
for breast cancer. Multivariable models were adjusted for age,
education (primary school, high school, and university), family
history of breast cancer (no or yes), history of benign breast
disease (no or yes), parity (nulliparous, 1–2 or >=3 births), age at
first birth (nulliparous or ,26, 26–30, or >=31 y), age at
menarche (<=12, 13, or >=14 y), age at menopause (<51 or
>=51 y), oral contraceptive use (never or ever), postmenopausal
hormone use (never and ,3, 3–5, or >=6 y; or users with
missing data on duration), body mass index (in kg/m2) (,18.5,
18.5–24.9, 25–29.9, or >=30), total physical activity (in quartiles
of metabolic equivalent hours per day), smoking status
(never, past, or current), calcium supplement use (no or yes),
and alcohol intake (nondrinkers or ,3.4, 3.4–9.9, or >=10.0 g
alcohol/d).
We conducted stratified analyses to examine whether the association
between multivitamin use and breast cancer was
modified by the use of postmenopausal hormones (never, quit ,5
or >=5 y ago, or current), smoking status (never, past, or current),
and alcohol intake (0–4, 5–9, or >=10 g alcohol/d). To test the
statistical significance of interactions on a multiplicative scale,
we used the log-likelihood ratio test and compared the models
with or without interaction terms. All statistical analyses were
performed with SAS version 9.1 (SAS Institute Inc, Cary, NC).
All P values were 2-sided.

Jun wrote: 另外一个感慨就是人生无常，所有的控制感都是虚幻的，所有的“我只要努力做好ABC就可以XYZ”念头都是虚幻的，说不定适得其反也有可能，只好及时行乐。