Skip to main content
. 2019 Dec 3;18:105. doi: 10.1186/s12940-019-0547-6

Fig. 2.

Fig. 2

Associations between skin care product use and pre- and postmenopausal cancer, endometrial cancer and estrogen receptor positive (ER+) and negative (ER-) breast cancer. Skin care product use is modelled in continuous scale using restricted cubic spline transformations of “% skin covered in cream per day” with 4 knots. The p-value corresponds to the null hypothesis that the regression coefficient for the second and third spline transformations equaled zero, i.e., a test of non-linearity