Associations of combined lifestyle index with migraine prevalence and headache frequency: a cross-sectional study from the MECH-HK study

Table 4 Relationships between combined lifestyle index and migraine attack frequency (attacks/month) in Hong Kong Chinese women with migraine (N = 357) (October 2019–December 2020)

Exposure ^a	Univariable analysis		Model I ^b		Model II ^c		Model III ^d
Exposure ^a	β (95% CI)	p	β (95% CI)	p	β (95% CI)	p	β (95% CI)	p
Combined lifestyle index	− 0.46 (− 0.72, − 0.20)	< 0.001	− 0.60 (− 0.94, − 0.27)	< 0.001	− 0.61 (− 0.95, − 0.27)	0.001	− 0.55 (− 0.82, − 0.28)	< 0.001
Smoking	−0.15 (− 1.84, 1.55)	0.864	− 0.37 (− 2.07, 1.34)	0.672	− 0.39 (− 2.13, 1.34)	0.655	− 0.32 (− 2.06, 1.42)	0.718
Physical activity	0.08 (− 0.81, 0.97)	0.862	− 0.19 (− 1.12, 0.74)	0.686	− 0.40 (− 1.35, 0.56)	0.418	−0.51 (− 1.47, 0.46)	0.305
Sleep	−0.69 (− 1.71, 0.33)	0.185	− 0.74 (− 1.76, 0.28)	0.156	− 0.78 (− 1.81, 0.25)	0.138	−0.74 (− 1.78, 0.29)	0.159
Stress	−0.57 (− 1.52, 0.37)	0.234	− 0.56 (− 1.50, 0.38)	0.242	− 0.52 (− 1.46, 0.43)	0.283	−0.49 (− 1.44, 0.46)	0.311
Fatigue	−1.15 (−2.15, − 0.14)	0.026	−1.20 (− 2.21, − 0.20)	0.019	−1.13 (− 2.14, − 0.11)	0.030	−1.12 (− 2.14, − 0.09)	0.033
Diet	0.32 (− 0.52, 1.17)	0.450	0.25 (− 0.59, 1.09)	0.561	0.13 (− 0.74, 0.99)	0.772	0.13 (− 0.76, 1.02)	0.774
Body mass index	−0.96 (− 1.77, − 0.14)	0.022	−0.88 (− 1.70, − 0.07)	0.033	−0.82 (− 1.64, 0.00)	0.051	− 0.84 (− 1.67, − 0.02)	0.046
Alcohol	0.10 (− 0.79, 0.99)	0.828	0.05 (− 0.84, 0.94)	0.907	0.07 (− 0.83, 0.96)	0.881	0.12 (− 0.78, 1.03)	0.789

OR Odds ratio, CI Confidence interval
^aThe β of per point increase in the combined lifestyle index and factors
^bModel I adjusted for age. In the analyses of each individual component, the other components were further adjusted
^cModel II adjusted for age, marital status, living condition, educational level, family income, employment status, menstrual age, and menopause. In the analyses of each individual component, the other components were further adjusted
^dModel III adjusted for variables in Model II, hypertension, diabetes, hyperlipidaemia, myocardial infarction, stroke, and cancer. In the analyses of each individual component, the other components were further adjusted

ISSN: 1129-2377