Skip to main content
. 2022 Nov 3;4(6):fcac257. doi: 10.1093/braincomms/fcac257

Table 3.

Effect of PSQI measures on CSF biomarkers at baseline

Variables log(CSF Aβ42) log(CSF P-tau) log(CSF t-tau)
β Coefficient (95% CI) P-value β Coefficient (95% CI) P-value β Coefficient (95% CI) P-value
Total PSQI score −0.003 (−0.012 0.006) 0.809 0.002 (−0.005 0.008) 0.609 0.003 (−0.003 0.008) 0.344
Dichotomized PSQI score (ref. Total PSQI ≤5) −0.013 (−0.064 0.039) 0.63 0.039 (−0.002 0.081) 0.062 0.044 (0.007 0.08) 0.018
Sleep latency (ref. ≤ 15 min)
 16–30 min −0.039 (−0.094 0.015) 0.153 0.007 (−0.038 0.053) 0.759 0.001 (−0.039 0.041) 0.946
 31–60 min 0.033 (−0.041 0.107) 0.382 −0.007 (−0.068 0.054) 0.824 −0.008 (−0.062 0.046) 0.767
 >60 min 0.058 (−0.05 0.165) 0.29 −0.02 (−0.108 0.069) 0.665 −0.014 (−0.092 0.064) 0.728
Sleep duration (ref. > 7 h)
 6–7 h −0.034 (−0.089 0.022) 0.232 0.069 (0.023 0.115) 0.003 0.057 (0.016 0.097) 0.006
 5–6 h −0.04 (−0.125 0.045) 0.356 0.019 (−0.052 0.09) 0.597 0.036 (−0.026 0.098) 0.259
 <5 h 0.097 (−0.076 0.27) 0.272 −0.01 (−0.151 0.131) 0.89 0.014 (−0.11 0.138) 0.824
Sleep efficiency (ref. > 85%)
 75–84% −0.036 (−0.093 0.021) 0.212 0.015 (−0.033 0.062) 0.545 0.015 (−0.027 0.057) 0.481
 65–74% −0.039 (−0.118 0.04) 0.335 0.029 (−0.036 0.094) 0.385 0.024 (−0.034 0.081) 0.418
 <65% −0.002 (−0.086 0.082) 0.961 0.039 (−0.03 0.108) 0.267 0.046 (−0.015 0.107) 0.138
Sleep disturbance (ref. 0)a
 1–9 −0.125 (−0.219 −0.031) 0.009 0.016 (−0.063 0.094) 0.695 0.034 (−0.035 0.104) 0.334
 >9 −0.121 (−0.23 -0.012) 0.03 0.016 (−0.074 0.107) 0.722 0.028 (−0.052 0.108) 0.49
Daytime dysfunction (ref. 0)b
 1–2 −0.037 (−0.09 0.016) 0.173 0.018 (−0.025 0.06) 0.411 0.014 (−0.023 0.052) 0.448
 >2 −0.032 (−0.15 0.087) 0.6 −0.085 (−0.182 0.011) 0.084 −0.06 (−0.145 0.025) 0.165

ref.: Level of reference. aCategories corresponding to scores of ‘10–18’ and ‘19–27’ have been collapsed due to <20 observations in one category. bCategories corresponding to scores of ‘3–4’ and ‘5–6’ have been collapsed due to <20 observations in one category. All models with log(CSF Aβ42) as an outcome are adjusted by age, sex, site of data collection, APOE-ɛ4 carriership, anxiety (State-Trait Anxiety Inventory) and log(CSF p-tau) levels. All models with log(p-tau) and log(CSF t-tau) levels as outcomes are adjusted by age, sex, site of data collection, APOE-ɛ4 carriership, body mass index, physical activity and log(CSF Aβ42) levels.