Skip to main content
. Author manuscript; available in PMC: 2016 Nov 3.
Published in final edited form as: Osteoporos Int. 2014 Feb 25;25(5):1473–1481. doi: 10.1007/s00198-014-2649-x

Table 2.

Results of univariate and multivariable Cox models using alternative approaches to represent SSRI + SNRI exposure

Models Unadjusted HR (95 % CI) Adjusted HR (95 % CI)
SSRI/SNRI current use—first definitiona 1.88 (1.48–2.39) 1.68 (1.32–2.15)b
SSRI/SNRI current use—second definitionc 1.58 (1.17–2.14) 1.47 (1.09–1.99)d
SSRI/SNRI stepwise 1.83 (1.35–2.47) 1.69 (1.28–2.22)e
SSRI/SNRI recent/former
 Former use 1.70 (1.08–2.65) 1.42 (0.87–2.31)
 Recent use 1.80 (1.34–2.42) 1.61 (1.19–2.19)f
a

SSRI/SNRI use was assumed to be constant for the period covered by each interview

b

Adjusted for Charlson score, age, sex, education level, fall in previous month, physical activity, depressive symptoms (MCS scale), BMD hip and lumbar, thiazolidinedione use, vitamin D, and previous vertebral deformity (Genant grade ≥2)

c

SSRI/SNRI use was assumed to continue during the 5-year period only if the subject reported being on these drugs both at the beginning and at the end of the period

d

Adjusted for the same variables as above, except smoking

e

Adjusted for Charlson score, age, sex, education level, fall in previous month, physical activity, depressive symptoms (MCS scale), BMD hip and lumbar, thiazolidinedione and antihypertensive use, vitamin D, and previous vertebral deformity (Genant grade ≥2)

f

Adjusted for the same variables as above, except education