Furthermore, it is likely that under-reporting of smoking is likely to be higher in areas with lower overall prevalence. We were also limited to measuring retailer density at one point in time However, evidence suggests that there has been little change in retailer densities over time in Scotland in the period —, and little in the way of public health interventions to restrict tobacco retailers before this period. Furthermore, any unmeasured temporal variation that is present is likely to be randomly distributed with respect to this study and is therefore unlikely to significantly affect the results.
Despite the strengths of the analytical design of the study, there remain other limitations with our analysis. First, we assume that postcode at delivery was the same as postcode throughout the 9-month period of pregnancy and, second, the analysis was unable to incorporate analyses of consumption patterns or purchasing behaviour and how these may be influenced by the retailer environment. Although the analytical design of the study fully controls for confounding at the individual level for factors that do not vary over time, there remains the potential for confounding from factors that change over time that may also be related to both propensity to smoke as well as residential movement to areas with greater densities of tobacco retailers.
For a variable to be a confounder, it must be time varying and it must be associated with changes in tobacco density exposure and changes in smoking outcome. For example, stressful life events and changes in life circumstances such as unemployment or relationship breakdown may be associated with both smoking behaviour and movement to an area with higher tobacco outlet densities. Although changes in these unmeasured variables contain a direct pathway to smoking, they are not directly connected to tobacco outlet density apart from via a pathway through changes in area income deprivation through residential relocation.
Adjusting for changes in area income deprivation therefore means the observed association between outlet density and smoking is independent of changes in the unobserved variables. While other sources of time varying confounding may exist, they are unlikely to account for the associations we have observed. We also conceptualise neighbourhood smoking prevalence as a confounder. Areas with higher smoking prevalence are likely to be identified by retailers as areas of higher demand increasing the likelihood of locating there.
This creates a confounding pathway between outlet density and smoking. However, it is equally likely that area smoking prevalence may in part be determined by outlet density and therefore may constitute a mediator rather than a confounder of the outlet density association. Thus, adjustment for area smoking prevalence may constitute overadjustment of the outlet density association.
Smoking during pregnancy: A population-based study
Although our results show that both exposures play a role in mediating smoking behaviour independently of each other with area prevalence associations being the stronger of the two , policy interventions that tackle the supply of tobacco may in time reduce area prevalence by reducing the effects associated with retailing density and supply. To the best of our knowledge, this is the first study to examine the influence of neighbourhood tobacco availability on smoking behaviour during pregnancy and one of the few studies of tobacco availability to incorporate a longitudinal analysis with a large dataset.
Our results show the significant role of the tobacco retailer environment in influencing smoking during pregnancy. The findings are the strongest indication yet of the need for policy to tackle the supply side of tobacco usage as a means of meeting the commitments laid out in national tobacco control strategies and of reducing smoking prevalence.
Get this Publication
The density of tobacco retailers in residential, school and other settings has been shown to be associated with the risk of being a smoker. A causal interpretation is limited because previous evidence is almost entirely cross-sectional and may suffer from issues of reverse causality and residual confounding. Among existing studies of tobacco retailer density and smoking behaviour, none have examined the relationship with smoking during pregnancy and none have controlled for a measure of neighbourhood maternal smoking prevalence.
With full population data, we examine within-individual changes in pregnancy smoking in relation to changes in exposure to neighbourhood tobacco availability controlling for neighbourhood maternal smoking prevalence. Risks of pregnancy smoking increased significantly when the pregnancy occurred in areas with the highest availability of tobacco compared with a pregnancy to the same individual in an area with zero tobacco availability.
The findings are unique with respect to smoking during pregnancy and the availability of tobacco and affirm and extend previous cross-sectional evidence from the general population and other demographic groups. National tobacco strategies should incorporate policies restricting tobacco supply as a means of tackling population tobacco use. Contributors TC conceived the study and methodology, conducted the analysis, assembled the data and wrote the first draft of the paper. All authors contributed to final manuscript editing and writing.
Provenance and peer review Not commissioned; externally peer reviewed.
Epigenetic Link Identified Between Maternal Smoking During Pregnancy and Offspring Obesity
Forgot your log in details? Register a new account? Forgot your user name or password? Search for this keyword. Advanced search.
Latest content Current issue Archive Authors About. Log in via Institution. You are here Home Online First Neighbourhood tobacco supply and individual maternal smoking during pregnancy: a fixed-effects longitudinal analysis using routine data. Email alerts. Article Text. Article menu. Research paper.
- powerful ways to end an essay;
- Smoking in Pregnancy: An Ongoing Challenge | Nicotine & Tobacco Research | Oxford Academic.
Neighbourhood tobacco supply and individual maternal smoking during pregnancy: a fixed-effects longitudinal analysis using routine data. Statistics from Altmetric. Methods Study design In addition to standard maternal clinical and demographic information, administrative maternity records in Scotland record smoking behaviour during pregnancy together with residential location at the time of delivery.
Tobacco outlets and neighbourhood maternal pregnancy smoking prevalence Assessment of the tobacco retail environment is based on the same method and data as adopted in earlier work in Scotland. Statistical methods We estimated conditional logistic regression models with individual mother-level fixed effects to examine the risks of smoking during pregnancy adjusting for potentially confounding factors.
Covariates The fixed-effects approach removes the need for adjustment of time-invariant individual level characteristics but the possibility of residual confounding by factors that vary over time remains. Supplementary data [tobaccocontrolSP1.
Smoking Policies and Birth Outcomes: Estimates From a New Era
View this table: View inline View popup. Table 1 Pregnancy-level descriptive statistics for both exposures retailing density and neighbourhood maternal smoking prevalence broken down by smoking status.
Table 2 Descriptive statistics for continuous exposure measures. Discussion This study has found strong and significant positive associations between residential density of tobacco retailers and the risk of maternal smoking during pregnancy. What this paper adds What is already known on this subject The density of tobacco retailers in residential, school and other settings has been shown to be associated with the risk of being a smoker. These associations are consistent across different sociodemographic groups. What important gaps in knowledge exist on this topic A causal interpretation is limited because previous evidence is almost entirely cross-sectional and may suffer from issues of reverse causality and residual confounding.
What this paper adds With full population data, we examine within-individual changes in pregnancy smoking in relation to changes in exposure to neighbourhood tobacco availability controlling for neighbourhood maternal smoking prevalence. Ending the tobacco problem: a blueprint for the nation : National Academies Press. Accessed 27 Apr Cohen JE , Anglin L. Outlet density: a new frontier for tobacco control. Addiction ; : 2 — 3. The neighbourhood effects of geographical access to tobacco retailers on individual smoking behaviour. J Epidemiol Community Health ; 63 : 69 — Second, the data on maternal smoking were based on maternal reports.
It is quite possible that some mothers would be reluctant to admit to smoking during pregnancy because of the social stigma surrounding the act.