Skip to main content
. 2022 Mar 17;19(3):e1003941. doi: 10.1371/journal.pmed.1003941

Fig 4. HRs, 95% CIs, and p-values for the association of duration of lithium exposure with the development of dementia (and its subtypes) by Cox proportional hazards models.

Fig 4

HRs, 95% CIs, and p-values were extracted from inverse probability weighted Cox regression. Adjusted HRs were adjusted for age, sex, marital status, ethnicity, smoking status, alcohol disorders, antipsychotic use, depression, mania or BPAD, hypertension, central vascular disease, diabetes mellitus, and hyperlipidemias. NA indicates no result (no corresponding cases). The AIC was 14,125 for the model fitted for all dementias, 13,064 for the model fitted for AD, and 13,119 for the model fitted for VD. See Fig Q in S1 Appendix for an equivalent analysis using pooled intermediate durations and Fig R in S1 Appendix for the effects of lithium exposure duration as a continuous variable. AD, Alzheimer disease; AIC, Akaike information criterion; BPAD, bipolar affective disorder/mania; CI, confidence interval; HR, hazard ratio; VD, vascular dementia.