Table 2

Logistica and linearb regression to predict reasons for not using smoking cessation aids from demographic, smoking behaviour, motivational, and health-related variables

Prediction variables

Scale "Social and environmental barriers"

OR (CI 95%)

Scale "SCA unnecessary"

β-weight


univariate

multivariable

univariate

multivariable

    demographic variables

gender

malec

female

.559 (.371–.843)

.620 (.376–1.020)

-.019

-d

age

1.004 (.989–1.019)

-

-.065

-

educational level

< 10 yearsc

= 10 years

.382 (.219–.668)

.500 (.267–.936)

-.013

> 10 years

.220 (.111–.434)

.297 (.139–.632)

-.012

-

marital status

married/living with a partnerc

not married/not living with a partner

.783 (.486–1.262)

-

-.019

-

    smoking behaviour variables

cigarettes per day

1.048 (1.015–1.082)

1.015 (.969–1.063)

.071

-

FTND-Scoree

1.274 (1.098–1.477)

1.206 (.992–1.466)

.041

-

age at onset

< 16 yearsc

16–18 years

1.219 (.758–1.959)

.012

> 18 years

1.091 (.613–1.942)

-

-.070

-

quit attempts within the last 12 months

0c

1

1.379 (.780–2.438)

.011

2–5

1.528 (.852–2.742)

.057

>5

1.051 (.389–2.839)

-

.016

-

smoking intensity

1.012 (1.003–1.021)

1.001 (.990–1.013)

.076

-

    motivational variables

intention to quit within the next six months

noc

yes

1.583 (1.013–2.473)

2.008 (1.207–3.338)

.121**

.113**

pros of non-smoking

1.148 (.957–1.376)

-

.078

-

self-efficacy

.885 (.696–1.125)

-

.063

-

    health-related variables

mental health

.927 (.684–1.257)

-

-.083*

-.073

general health

.989 (.976–1.001)

-

.022

-


All analyses were adjusted for study membership (population-based vs. general practice patient sample) as well as study group membership (intervention vs. control condition).

a Scale „Social and environmental barriers", due to substantial non-normality the scale was dichotomised using 0 = no agreement, 1 = at least minimal agreement

b Scale "SCA unnecessary"

c Reference category

d Not sufficient for multivariable analysis

e Because cigarettes per day were used as separate predictor, the FTND score was computed without cigarettes per day to avoid multi-collinearity

* p < .05; ** p < .01

Gross et al. BMC Public Health 2008 8:129   doi:10.1186/1471-2458-8-129

Open Data