Table 4 Logistic regression model summaries for associations between customised questions around fear related to safety during sleep (independent variable) and self-reported sleep characteristics (dependent variable) in women (n = 234).

From: Associations between fears related to safety during sleep and self-reported sleep in men and women living in a low-socioeconomic status setting

Fear item

Excessive daytime sleepiness (ESS > 10)

Poor sleep quality (PSQI > 5)

Clinically significant insomnia symptoms (ISI > 14)

PSQI disturbance item

PSQI daytime dysfunction item

PSQI time-in-bed

PSQI total sleep time

“Awakened by strange noises”

2.62 (1.45–4.72), p = 0.001

3.45 (1.82–6.55), p < 0.001

4.98 (1.04–23.86), p = 0.045

1.53 (0.92–2.55), p = 0.105

1.00 (0.61–1.62), p = 0.988

1.36 (0.79–2.33), p = 0.261

0.91 (0.54–1.54), p = 0.735

“Dreams about past traumatic experiences”

2.05 (1.12–3.74), p = 0.020

2.18 (1.18–4.05), p = 0.013

2.44 (0.75–7.97), p = 0.140

2.11 (1.21–3.70), p = 0.009

1.15 (0.68–1.93), p = 0.612

1.67 (0.94–2.99), p = 0.081

1.13 (0.65–1.99), p = 0.661

“Fear of being attacked during sleep”

3.42 (1.81–6.43), p < 0.001

2.13 (1.12–4.03), p = 0.020

2.30 (0.70–7.58), p = 0.170

1.28 (0.72–2.28), p = 0.408

1.15 (0.66–1.98), p = 0.627

2.01 (1.12–3.91), p = 0.021

1.58 (0.86–2.88), p = 0.138

“Fear of not being safe during sleep”

2.43 (1.29–4.57), p = 0.006

3.61 (1.80–6.54), p < 0.001

3.77 (1.14–12.48), p = 0.030

1.30 (0.72–2.33), p = 0.375

1.29 (0.74–2.25), p = 0.365

2.16 (1.14–4.09), p = 0.018

1.31 (0.71–2.40), p = 0.382

“Fear of falling asleep”

3.44 (1.13–5.28), p = 0.023

3.30 (1.53–7.12), p = 0.002

3.74 (1.04–13.44), p = 0.043

1.38 (0.67–2.82), p = 0.384

1.25 (0.66–2.37), p = 0.488

1.76 (0.82–3.77), p = 0.147

0.94 (0.45–1.97), p = 0.865

“Sleeping with the light on to feel safe”

1.91 (0.91–4.00), p = 0.086

1.57 (0.73–3.37), p = 0.246

5.92 (1.54–22.77), p = 0.010

0.80 (0.40–1.60), p = 0.530

0.95 (0.49–1.85), p = 0.890

1.11 (0.54–2.29), p = 0.776

1.08 (0.54–2.15), p = 0.820

  1. Data are presented as odds ratio (95% confidence interval), p value. Logistic regression models were adjusted for age a priori and then adjusted for covariates which had a p value < 0.15 in univariate analyses (Supplementary Table S4). Additional covariates were as follows: for excessive daytime sleepiness model: smoking, alcohol consumption, household density and level of education; for poor sleep quality model: alcohol consumption per week, level of education and presence of young children in the house; for insomnia symptoms model: alcohol consumption and level of education; for the time-in-bed models: BMI, employment status, presence of young children and household density and for the total sleep time models: BMI and employment status. Significant values are in bold.
Back to article page