The alcohol harm paradox: using a national survey to explore how alcohol may disproportionately impact health in deprived individuals

Bellis, Mark A.; Hughes, Karen; Nicholls, James; Sheron, Nick; Gilmore, Ian; Jones, Lisa

doi:10.1186/s12889-016-2766-x

Table 3 Logistic regression: relationships between alcohol paradox variables and deprivation alone and interacting with alcohol consumption

From: The alcohol harm paradox: using a national survey to explore how alcohol may disproportionately impact health in deprived individuals

Alcohol harm paradox variables	Deprivation (ref non-deprived)			Deprived v. non-deprived within alcohol consumption category (ref minimal consumer^a)
	Deprived				Lower risk			Increased risk			Higher risk			Never			Ex-drinker
	AOR +/−95%CI		P	P ^b	AOR +/−95%CI		P	AOR +/−95%CI		P	AOR +/−95%CI		P	AOR +/−95%CI		P	AOR +/−95%CI		P
Combined health challenges
Current smoker	1.86	1.52–2.28	***	*	1.23	0.82–1.85	ns	1.86	1.09–3.17	*	2.23	1.01–4.89	*	0.57	0.26–1.25	ns	0.94	0.61–1.47	ns
< =1 portion fruit-veg/day	1.87	1.58–2.22	***	ns	1.37	0.85–2.20	ns	1.99	0.99–3.98	ns	2.12	0.93–4.85	ns	0.95	0.48–1.90	ns	1.07	0.65–1.76	ns
< 1 exercise session/week	1.38	1.22–1.56	***	ns	0.89	0.64–1.24	ns	0.93	0.57–1.53	ns	1.49	0.70–3.16	ns	0.60	0.35–1.03	ns	0.90	0.63–1.29	ns
Overweight BMI >25	1.44	1.28–1.63	***	ns	0.88	0.64–1.20	ns	0.96	0.61–1.51	ns	0.60	0.29–1.22	ns	0.86	0.50–1.47	ns	0.80	0.55–1.14	ns
Drinking patterns
Average session binge	1.73	1.44–2.07	***	ns	0.88	0.33–2.36	ns	0.97	0.35–2.73	ns	0.8	0.18–3.48	ns
Typically drink wine	0.42	0.34–0.53	***	**	0.65	0.48–0.89	**	0.55	0.34–0.86	*	0.30	0.14–0.65	**
Typically drink spirits	1.43	1.21–1.68	***	ns	1.04	0.72–1.50	ns	0.96	0.57–1.63	ns	1.00	0.46–2.14	ns
Typically drink beer/cider	1.57	1.34–1.85	***	ns	1.36	0.92–2.00	ns	1.99	1.18–3.36	*	1.16	0.52–2.61	ns
> =5units/week underestimate	0.79	0.65–0.96	*	ns	1.58	0.71–3.52	ns	1.27	0.54–3.00	ns	1.85	0.65–5.26	ns
Drinking histories
Regular drinker <18 years	0.99	0.83–1.18	ns	ns	1.35	0.97–1.87	ns	1.15	0.73–1.80	ns	1.27	0.62–2.63	ns				1.17	0.78–1.75	ns
Drunk <18 years	1.00	0.82–1.20	ns	ns	1.01	0.71–1.45	ns	0.90	0.56–1.45	ns	0.92	0.43–1.98	ns				0.83	0.53–1.30	ns
Frequent drinking history^c	1.09	0.73–1.63	ns	ns	0.96	0.29–3.19	ns	1.59	0.46–5.53	ns	1.35	0.24–7.68	ns				0.76	0.21–2.78	ns
Drunken/bingeing history^c	1.20	0.95–1.52	ns	*	1.87	1.06–3.31	*	2.60	1.32–5.14	**	1.59	0.64–3.93	ns				1.11	0.56–2.23	ns

Age, sex, alcohol consumption risk category (alone) and ethnicity were all also included in the logistic regression model. However, for reasons of space only the key variables deprivation and the deprivation interaction with alcohol consumption risk category are shown. ^aMinimal is the reference category for interactions between deprivation and alcohol consumption. ^b P values shown refer to the significance of the overall contribution of the interactive term (Alcohol Consumption Risk Category*Deprivation) to the model. ^cFrequent drinking and drunken/bingeing histories are limited to individuals aged 35 years or more at interview (See Methods for more details). For deprivation alone AORs (Adjusted Odds Ratios) use non-deprived as the reference category. 95 % CI = 95 % Confidence Intervals. BMI Body Mass Index. Consumption categories are defined as in Table 1. 1 unit is approximately 8 g of pure alcohol. *p<0.05, **p<0.01, ***p<0.001.

Back to article page

ISSN: 1471-2458

Contact us

Submission enquiries: bmcpublichealth@biomedcentral.com
General enquiries: ORSupport@springernature.com

BMC Public Health

Contact us