Using Birth Cohort Data to Assess the Impact of the UK 2008–2010 Economic Recession on Smoking During Pregnancy

Using Birth Cohort Data to Assess the Impact of the UK 2008–2010 Economic Recession on Smoking... Abstract Introduction Despite the well-known link between stress and smoking, evidence for associations between economic recession, financial stress, and smoking is contradictory. In this study, we assess whether women were more likely to continue smoking during pregnancy if they were exposed to the UK 2008–2010 economic recession during pregnancy than those who were unexposed, and whether this relationship is mediated by financial stress. Methods We used cross-sectional data on 2775 pregnant women who were regular smokers before pregnancy and who were enrolled in the UK Born in Bradford cohort study between March 2007 and December 2010. The cutoff date for exposure to recession was set as August 1, 2008, based on local and national economic data. Multivariable logistic regression analysis included potential confounders: maternal age, parity, cohabitation, ethnicity, and maternal age. The mediating role of financial stress was analyzed using “worse off financially” and a “difficult financial situation” as indicators of financial stress in Sobel–Goodman mediation tests with bootstrap resampling. Results After taking into account potential confounders, exposure to recession was associated with continued smoking during pregnancy (OR = 1.19, 95% CI = 1.01 to 1.41, p = 0.03). A worse financial situation and a difficult financial situation were identified as mediators, explaining 8.4% and 17.6%, respectively, of the relationship between exposure to recession and smoking during pregnancy. Conclusions Smoking during pregnancy is associated with exposure to the UK 2008–2010 economic recession during pregnancy, and this relationship is partly mediated by financial stress. Implications Health inequalities in smoking during pregnancy are affected by economic recession, as those who are most likely to smoke are also most likely to experience the financial stress resulting from economic recession. Socioeconomic conditions at the societal and individual level are important targets when aiming to reduce rates of smoking during pregnancy. Introduction Social Gradients in Tobacco Smoking Tobacco smoking in pregnancy is more prevalent among those who have a lower socioeconomic status as indicated by a range of socioeconomic measures ranging from education and employment status to poor childhood circumstances, housing tenure, and job strain.1–4 In an analysis of continued smoking during pregnancy in 15 European countries, Smedberg and colleagues2 found an average prevalence rate of 26.2%, compared with 43.2% in the lowest educational category. In England, a combination of socioeconomic indicators showed a very steep social gradient with 6% of women smoking during pregnancy in the highest and 62% of women smoking in the lowest socioeconomic group.4 An analysis of data from the Millennium Cohort Study showed that smoking and heavy smoking during pregnancy were associated with a range of factors more common among lower socioeconomic groups, including leaving school before eighteen, not having a bank account, and financial difficulties.5 Tobacco Smoking and Financial Stress Financial stress appears to play a significant role in explaining these social gradients, although most of the evidence comes from smoking in the general population. In Australia, financial stress according to an eight-item scale relating to shortage of money was associated with a lower likelihood of quitting and a higher likelihood of relapse.6 This finding was confirmed in a survey with data from the United States, United Kingdom, Canada, and Australia. Smokers experiencing financial stress had an increased interest in quitting but were less likely to make an attempt, and consequently their odds of succeeding was half that of the group without financial stress.7 Involuntary job loss among older workers was associated with increased odds of a relapse in smoking, defined as reported smoking point prevalence after quitting.8 A US community-based sample was used to demonstrate that women are less likely to quit smoking in response to an adverse financial event, and more likely to relapse as a result of such an event.9 Tobacco Smoking During the 2008–2010 Economic Recession Given that financial stress affects those of lower socioeconomic status disproportionally, and macro-economic circumstances such as economic recession can evoke financial stress, it is reasonable to assume that the economic recession, which hit Europe between 2008 and 2010, would have increased rates of smoking during pregnancy. Women who were regular smokers may have been less able to quit smoking as financial strain placed a burden on their cognitive, social, and emotional capacities. The evidence, however, which is largely based on tobacco use in the general population, is contradictory. Based on US economic data, Ruhm concluded that smokers, especially heavy smokers, cut down on smoking during economic downturns.10 Others have showed that increased wages lead to increased cigarette use,11 and larger than expected declines in smoking during the economic recession have been found in Iceland12,13 and in Greece.14 These findings have come with nuances and reservations. The authors of the Greek study pointed out that more effective enforcement of tobacco control legislation may partly explain their findings.14 Gallus and colleagues argued that the number of unemployed people grows during a recession, so that a small decrease in smoking is seen at the ecological level because the downward trend in smoking prevalence for employed people is slightly larger than the increase in smoking among the unemployed during economic recession.15 In a study of UK parents, financial strain during the recession was associated with an increased risk of persistent tobacco use and relapse.16 In summary, most of the literature would suggest that experiences of financial stress, which are likely to result from an economic recession, might increase the likelihood of continued smoking during pregnancy. Although these associations are plausible, the literature on recession and smoking suggests an effect in the opposite direction. An analysis specifically testing associations between economic recession and smoking during pregnancy, including the mediating effect of financial stress, might clarify previous findings. Hypothesis In a deprived UK community hit hard by the UK 2008–2010 economic recession, women who smoke regularly will be more likely to continue smoking during pregnancy if they were exposed to the recession during pregnancy compared with those unexposed, and this relationship is mediated by financial stress. Methods This study follows guidance on reporting set out in the STROBE statement (Supplementary Table A1). Data The Born in Bradford (BiB) cohort was set up to examine the impact of social, environmental, psychological, behavioral, and biological factors on maternal and child health and well-being.17 Pregnant women (N = 12450) were recruited around 26 to 28 weeks of pregnancy at Bradford’s only maternity unit when attending universal screening for gestational diabetes, between March 2007 and December 2010.18 Ethical approval for data collection was granted by Bradford Research Ethics Committee (Ref 07/H1302/112). After excluding pregnancies without baseline questionnaire data (N = 2377), stillbirths (N = 61), second and third pregnancies to the same mother in the cohort (N = 1323), and twins and triplets (N = 142), a dataset of 10035 mother–infant pairs was obtained. A dataset with complete data for the exposure variable and covariates was used. This dataset of 8952 pregnant women was used to explore relationships between exposure to recession and financial stress. In this dataset, 2775 women were regular smokers and reported information on smoking during pregnancy and relevant covariates. Setting Bradford is a deprived and ethnically diverse city in the North of England with a population of over half a million.19 Employment is more reliant on manufacturing industry than the UK average and this sector has been in decline for decades, making the city particularly vulnerable to the effects of economic recession. There was a sharp increase in the proportion of claimants for Job-Seeker’s Allowance (JSA) (the main form of financial support for unemployed people identified as actively seeking work) in Bradford from August 2008, with a peak in JSA claimants at 5.0% in September 2009 (Figure 1).20 In November 2007, JSA-claimant-rates in Bradford were 133% higher than for England as a whole; by November 2015, rates were 181% higher in Bradford than the average for England. Figure 1. View largeDownload slide Percentage JSA claimants as a proportion of the resident population aged 16–64. Figure 1. View largeDownload slide Percentage JSA claimants as a proportion of the resident population aged 16–64. Main Exposure The definition of recession as “A period of temporary economic decline during which trade and industrial activity are reduced, generally identified by a fall in GDP in two successive quarters”21 may not correspond with the lived experience of recession. We used reports in national and local media22 (Supplementary Box A1) in combination with data on unemployment related benefit claims21 (Figure 1) to inform the decision of a cutoff date for recession start date to be the first of August 2008; an estimated conception date from August 01, 2008 onwards was therefore classified as “exposed.” No “end of exposure” cutoff was used as Figure 1 indicates the economic impact of the recession lasted well beyond the study recruitment period. The percentage of study participants reporting to be financially worse off than a year ago increases from 15.3% (N = 60) at the beginning of May 2008 to 24.2% (N = 88) around the beginning of June 2008 (Figure 2). In an exploration of different measures of exposure, we found no evidence for an annual worsening of individual’s financial situation (analyses available upon request). A binary cutoff before and since August 01, 2008 was therefore considered an appropriate measure of exposure to recession. Figure 2. View largeDownload slide Percentage of pregnant women reporting to be worse off financially by year and quarter of recruitment. Figure 2. View largeDownload slide Percentage of pregnant women reporting to be worse off financially by year and quarter of recruitment. Mediating Factors Financial stress was operationalized through two variables; perceived financial situation and change in perceived financial situation. Women were asked how well they and their partner were managing financially, with response categories ranging from “living comfortably” to “difficult” (“very difficult” and “just about getting by” were merged with “difficult”). Change in perceived financial situation was measured with the question “Compared to a year ago, how would you say you and your husband/partner are doing financially now?,” with answer categories “better off,” “about the same” or “worse off.” Outcomes Data on maternal smoking were obtained from the study baseline questionnaire. A binary variable of continued smoking during pregnancy for those who reported to smoke regularly was derived from questions on smoking during pregnancy 3 months before pregnancy, in the first 3 months of pregnancy and since the beginning of the fourth month. Women who reported to have stopped smoking in the first month of pregnancy and did not report smoking any cigarettes since the beginning of the fourth month were counted as nonsmokers during pregnancy. Covariates The Directed Acyclic Graph created in DAGgity V2.3 and published on the Daggity Web site (http://dagitty.net/mLWfIRw) shows the hypothesized causal relationships between exposure to recession, continued smoking during pregnancy and covariates.23 Parity (nulliparous vs. other), maternal age, cohabitation with a partner (yes/no) ethnic group (White British, Pakistani, other), and education of the mother (<5 GCSE, 5 GCSE, A level, >A level, other, or equivalents of these qualifications) were identified as potential confounders. Statistical Analyses Firstly, sample characteristics were explored and differences tested between the subsample exposed to recession during pregnancy and the unexposed group. Chi-square tests were used for binary variables and t-tests for continuous variables. Multivariable logistic regression analysis was used to model the relationship between exposure to economic recession and continued smoking during pregnancy, after which mediation analysis was performed to assess the role of financials stress. The mediating role of financial stress was tested with Sobel–Goodman mediation analyses in Stata 12, using the “sgmediation” and “bootstrap” commands.24 This method has been found to be more rigorous and more likely to identify a true mediation effect than the widely used causal steps outlined by Baron and Kenny.25–27 We avoided violation of the normality assumption of the Sobel–Goodman test by performing a bootstrap analysis with 5000 sampling repetitions.26 Results The sample consists of 2775 women who were regular smokers before pregnancy, and Table 1 shows an overview of their characteristics. As opposed to those recruited before the economic recession, pregnant women who were recruited after the start of the recession and were thus exposed were more likely to be nulliparous (first child), to have a higher level of education, to be in a difficult financial situation, and to be worse off than a year ago. Prevalence rates of smoking during pregnancy varied from 40.0% among those who reported to be managing well financially, to 65.2% among mothers who reported their financial situation was difficult/very difficult. Table 1. Participant Characteristics by Exposure Group Participant characteristics  Unexposed N = 1362  Exposed N = 1413  Total N = 2775  p value test of difference between groups*  Maternal age (years)  25.79  26.13  25.97  0.06  Parity        0.03   Nulliparous  50.44%  54.56%  52.57%   Multiparous  49.56%  45.44%  47.43%  Ethnicity        0.35   White British  77.97%  76.57%  77.26%   Pakistani  12.19%  11.89%  12.04%   Other  9.84%  11.54%  10.70%  Maternal education        <0.01   <5 GCSE  25.40%  23.50%  24.43%   5 GCSE  38.11%  36.73%  37.41%   A level  13.44%  18.75%  16.14%   >A level  14.54%  15.64%  15.10%   Other  8.52%  5.38%  6.92%  Cohabiting        0.11   Yes  69.53%  66.67%  68.07%   No  30.47%  33.33%  31.93%  Financial situation        0.02   Comfortable  23.79%  21.80%  22.77%   Doing alright  40.53%  37.30%  38.88%   Difficult  35.68%  40.91%  38.34%  Change in financial situation        <0.01   Same/better off  74.23%  69.57%  71.86%   Worse off  25.77%  30.43%  28.14%  Continued smoking during pregnancy        0.14   Yes  53.52%  56.33%  54.95%   No  46.48%  43.67%  45.05%  Participant characteristics  Unexposed N = 1362  Exposed N = 1413  Total N = 2775  p value test of difference between groups*  Maternal age (years)  25.79  26.13  25.97  0.06  Parity        0.03   Nulliparous  50.44%  54.56%  52.57%   Multiparous  49.56%  45.44%  47.43%  Ethnicity        0.35   White British  77.97%  76.57%  77.26%   Pakistani  12.19%  11.89%  12.04%   Other  9.84%  11.54%  10.70%  Maternal education        <0.01   <5 GCSE  25.40%  23.50%  24.43%   5 GCSE  38.11%  36.73%  37.41%   A level  13.44%  18.75%  16.14%   >A level  14.54%  15.64%  15.10%   Other  8.52%  5.38%  6.92%  Cohabiting        0.11   Yes  69.53%  66.67%  68.07%   No  30.47%  33.33%  31.93%  Financial situation        0.02   Comfortable  23.79%  21.80%  22.77%   Doing alright  40.53%  37.30%  38.88%   Difficult  35.68%  40.91%  38.34%  Change in financial situation        <0.01   Same/better off  74.23%  69.57%  71.86%   Worse off  25.77%  30.43%  28.14%  Continued smoking during pregnancy        0.14   Yes  53.52%  56.33%  54.95%   No  46.48%  43.67%  45.05%  *χ2 test for categorical variables and t-test for continuous variables. View Large Table 1. Participant Characteristics by Exposure Group Participant characteristics  Unexposed N = 1362  Exposed N = 1413  Total N = 2775  p value test of difference between groups*  Maternal age (years)  25.79  26.13  25.97  0.06  Parity        0.03   Nulliparous  50.44%  54.56%  52.57%   Multiparous  49.56%  45.44%  47.43%  Ethnicity        0.35   White British  77.97%  76.57%  77.26%   Pakistani  12.19%  11.89%  12.04%   Other  9.84%  11.54%  10.70%  Maternal education        <0.01   <5 GCSE  25.40%  23.50%  24.43%   5 GCSE  38.11%  36.73%  37.41%   A level  13.44%  18.75%  16.14%   >A level  14.54%  15.64%  15.10%   Other  8.52%  5.38%  6.92%  Cohabiting        0.11   Yes  69.53%  66.67%  68.07%   No  30.47%  33.33%  31.93%  Financial situation        0.02   Comfortable  23.79%  21.80%  22.77%   Doing alright  40.53%  37.30%  38.88%   Difficult  35.68%  40.91%  38.34%  Change in financial situation        <0.01   Same/better off  74.23%  69.57%  71.86%   Worse off  25.77%  30.43%  28.14%  Continued smoking during pregnancy        0.14   Yes  53.52%  56.33%  54.95%   No  46.48%  43.67%  45.05%  Participant characteristics  Unexposed N = 1362  Exposed N = 1413  Total N = 2775  p value test of difference between groups*  Maternal age (years)  25.79  26.13  25.97  0.06  Parity        0.03   Nulliparous  50.44%  54.56%  52.57%   Multiparous  49.56%  45.44%  47.43%  Ethnicity        0.35   White British  77.97%  76.57%  77.26%   Pakistani  12.19%  11.89%  12.04%   Other  9.84%  11.54%  10.70%  Maternal education        <0.01   <5 GCSE  25.40%  23.50%  24.43%   5 GCSE  38.11%  36.73%  37.41%   A level  13.44%  18.75%  16.14%   >A level  14.54%  15.64%  15.10%   Other  8.52%  5.38%  6.92%  Cohabiting        0.11   Yes  69.53%  66.67%  68.07%   No  30.47%  33.33%  31.93%  Financial situation        0.02   Comfortable  23.79%  21.80%  22.77%   Doing alright  40.53%  37.30%  38.88%   Difficult  35.68%  40.91%  38.34%  Change in financial situation        <0.01   Same/better off  74.23%  69.57%  71.86%   Worse off  25.77%  30.43%  28.14%  Continued smoking during pregnancy        0.14   Yes  53.52%  56.33%  54.95%   No  46.48%  43.67%  45.05%  *χ2 test for categorical variables and t-test for continuous variables. View Large Multivariable Analysis After taking into account potential confounders, exposure to recession was associated with continued smoking during pregnancy (OR = 1.19, 95% CI = 1.01 to 1.41, p = 0.03) (Table 2). All other factors being equal, this represents a 3.7% difference in smoking prevalence, with 53.1% of women in the unexposed and 56.8% of women in the exposed group smoking during pregnancy. Table 2. Exposure to Recession in Relation to Smoking During Pregnancy   Model 1. Bivariate “exposure recession” N = 2775  Model 2. Multivariate “exposure recession” N = 2775  OR  95% CI  OR  95% CI  Exposure recession  1.12  0.96; 1.30  1.19  1.01;1.41*  Maternal age      0.93  0.92; 0.95***  Parity           Multiparous      1.37  1.14;1.65**  Ethnic group (vs. White British)           Pakistani      0.61  0.48;0.79***   Other      0.79  0.61;1.04  Maternal education (vs. <5 GCSE)           5 GCSE      0.50  0.40;0.63***   A level      0.40  0.31;0.53***   >A level      0.19  0.14;0.25***   Other      0.33  0.24;0.47***  Cohabitation           Yes      0.53  0.44;0.64***    Model 1. Bivariate “exposure recession” N = 2775  Model 2. Multivariate “exposure recession” N = 2775  OR  95% CI  OR  95% CI  Exposure recession  1.12  0.96; 1.30  1.19  1.01;1.41*  Maternal age      0.93  0.92; 0.95***  Parity           Multiparous      1.37  1.14;1.65**  Ethnic group (vs. White British)           Pakistani      0.61  0.48;0.79***   Other      0.79  0.61;1.04  Maternal education (vs. <5 GCSE)           5 GCSE      0.50  0.40;0.63***   A level      0.40  0.31;0.53***   >A level      0.19  0.14;0.25***   Other      0.33  0.24;0.47***  Cohabitation           Yes      0.53  0.44;0.64***  *p < .05, ** p < .01, *** p < .001 View Large Table 2. Exposure to Recession in Relation to Smoking During Pregnancy   Model 1. Bivariate “exposure recession” N = 2775  Model 2. Multivariate “exposure recession” N = 2775  OR  95% CI  OR  95% CI  Exposure recession  1.12  0.96; 1.30  1.19  1.01;1.41*  Maternal age      0.93  0.92; 0.95***  Parity           Multiparous      1.37  1.14;1.65**  Ethnic group (vs. White British)           Pakistani      0.61  0.48;0.79***   Other      0.79  0.61;1.04  Maternal education (vs. <5 GCSE)           5 GCSE      0.50  0.40;0.63***   A level      0.40  0.31;0.53***   >A level      0.19  0.14;0.25***   Other      0.33  0.24;0.47***  Cohabitation           Yes      0.53  0.44;0.64***    Model 1. Bivariate “exposure recession” N = 2775  Model 2. Multivariate “exposure recession” N = 2775  OR  95% CI  OR  95% CI  Exposure recession  1.12  0.96; 1.30  1.19  1.01;1.41*  Maternal age      0.93  0.92; 0.95***  Parity           Multiparous      1.37  1.14;1.65**  Ethnic group (vs. White British)           Pakistani      0.61  0.48;0.79***   Other      0.79  0.61;1.04  Maternal education (vs. <5 GCSE)           5 GCSE      0.50  0.40;0.63***   A level      0.40  0.31;0.53***   >A level      0.19  0.14;0.25***   Other      0.33  0.24;0.47***  Cohabitation           Yes      0.53  0.44;0.64***  *p < .05, ** p < .01, *** p < .001 View Large Mediation Analysis Results of the mediation analysis are summarized in Table 3. Table 3. Results of Sobel–Goodman Test With Bootstrap Analysis   Coefficient  SE  95% CI  Model with mediator “worse financial situation” (N = 2775)   Indirect effect*  0.003  0.001  0.000 to 0.001   Direct effect  0.031  0.017  −0.003 to 0.066  Model with mediator “difficult financial situation” (N = 2775)   Indirect effect**  0.006  0.002  0.002 to 0.011   Direct effect  0.028  0.018  −0.007 to 0.063    Coefficient  SE  95% CI  Model with mediator “worse financial situation” (N = 2775)   Indirect effect*  0.003  0.001  0.000 to 0.001   Direct effect  0.031  0.017  −0.003 to 0.066  Model with mediator “difficult financial situation” (N = 2775)   Indirect effect**  0.006  0.002  0.002 to 0.011   Direct effect  0.028  0.018  −0.007 to 0.063  *Proportion of the effect that is mediated: 0.084. **Proportion of the effect that is mediated: 0.176. View Large Table 3. Results of Sobel–Goodman Test With Bootstrap Analysis   Coefficient  SE  95% CI  Model with mediator “worse financial situation” (N = 2775)   Indirect effect*  0.003  0.001  0.000 to 0.001   Direct effect  0.031  0.017  −0.003 to 0.066  Model with mediator “difficult financial situation” (N = 2775)   Indirect effect**  0.006  0.002  0.002 to 0.011   Direct effect  0.028  0.018  −0.007 to 0.063    Coefficient  SE  95% CI  Model with mediator “worse financial situation” (N = 2775)   Indirect effect*  0.003  0.001  0.000 to 0.001   Direct effect  0.031  0.017  −0.003 to 0.066  Model with mediator “difficult financial situation” (N = 2775)   Indirect effect**  0.006  0.002  0.002 to 0.011   Direct effect  0.028  0.018  −0.007 to 0.063  *Proportion of the effect that is mediated: 0.084. **Proportion of the effect that is mediated: 0.176. View Large When a worse financial situation than a year ago is used as the indication of financial stress, bootstrap results for the Sobel–Goodman test suggests a modest mediation effect (observed coefficient = 0.003, 95% CI = 0.00 to 0.01). Of the total relationship between exposure to recession and continued smoking during pregnancy, 8.4% is estimated to be mediated by the effect of a worse financial situation. When financial stress is operationalized as a current difficult financial situation (compared to comfortable/doing alright), a modest mediation effect is again present (observed coefficient = 0.006, 95% CI = 0.00 to 0.01). The proportion of the relationship between exposure to recession and continued smoking during pregnancy mediated by the effect of a difficult financial situation is 17.6%. By taking into account the influence of financial stress, either through the inclusion of a worse or difficult financial situation variable, the direct effect of exposure to the economic recession on smoking during pregnancy is no longer statistically significant. Conclusions This study demonstrates that exposure to economic recession is associated with continued smoking during pregnancy for regular smokers and that this relationship is partly mediated by financial stress. Exposure to the UK 2008–2010 economic recession during pregnancy was associated with a difficult financial situation and being worse off than a year ago for women in the Born in Bradford cohort study who were regular smokers. After taking into account key confounders, pregnant women exposed to recession were more likely to continue smoking during pregnancy than those not exposed. Mediation analyses showed that the relationship between exposure to recession and continued smoking during pregnancy was partly explained by financial stress, either measured by a difficult or worsened financial situation of the household. With the inclusion of financial stress in the model, the relationship between exposure to recession and smoking during pregnancy was no longer statistically significant. Limitations Even though covariates were included to adjust for differences between the “exposed” and “unexposed,” there might be alternative explanations for our findings. Couples who get pregnant during an economic recession may be different from those who choose not to. This may explain differences between the unexposed and exposed group shown in Table 1. Economic recession and unemployment have been found to be associated with a decline in fertility rates, particularly for young women having their first child,28,29 and fertility rates in England and Wales declined between 2008 and 2010 in all age groups except for mothers aged 35 and above.30 If couples who were more strongly affected by recession were more likely to delay pregnancies, we have underestimated the relationship between economic recession, financial stress, and smoking during pregnancy. Our study is limited by missing data and by its cross-sectional design. Despite our careful considerations of the cutoff for exposure to recession by using area-specific and national data, it is possible that we did not identify correctly after which date people were likely to experience the economic recession in their lives. In addition to misspecification of the exposure cutoff, data on key covariates may be influenced by date of data collection, as researchers will have gradually become more experienced in the administration of the questionnaire. It is likely that data are not missing at random. Finally, smoking during pregnancy is a stigmatized behavior that is likely to be underreported when relying on self-report measures, which means our prevalence rates of smoking during pregnancy are likely to underestimate true prevalence rates. Implications for Research and Society This study contradicts findings which suggest a beneficial effect of economic recession on smoking, and fits with literature on the role of stress, including financial stress, in a person’s ability to quit smoking and prevent relapse.5–9,16 In our sample, prevalence rates of smoking during pregnancy increased over time, whereas there is a slow but steady decline in smoking during pregnancy nationally.31 In Scotland, the introduction of a smoking ban in public places led to reduced rates of smoking during pregnancy.32 Our data show no such effect on smoking during pregnancy around the time the smoking ban took effect in England in July 2007. Any downward trends in smoking prevalence may have been counteracted by the larger impact of the economic recession. All women in this study were sampled from a UK city characterized by deprivation and health inequalities. The increased burden placed on this already disadvantaged community was by no means unavoidable. Research shows that strong social policies can prevent health impacts of recession.33,34 Our study shows that socioeconomic conditions, both at the societal and individual level, are important targets when aiming to reduce rates of smoking during pregnancy. If there is ever a time to invest in protecting the vulnerable to the advantage of society as a whole, the period during and after recession is such a time. Supplementary Material Supplementary Table A1 and Box A1 can be found online at https://academic.oup.com/gerontologist/. Funding This work was supported by the Born in Bradford study funding. The BiB study presents independent research commissioned by the National Institute for Health Research Collaboration for Applied Health Research and Care (NIHR CLAHRC-2013-10022) and the NIHR Programme Grants for Applied Research funding scheme [grant number RP-PG-0407-10044]. Core support for BiB is also provided by the Wellcome Trust [grant number WT101597MA]. All authors receive funding from the Big Lottery Fund as part of the “A Better Start” programme. Declaration of interests The authors report no conflicts of interest. Acknowledgments Born in Bradford is only possible because of the enthusiasm and commitment of the children and parents in the study. We are grateful to all the participants, health professionals and researchers who make Born in Bradford happen. References 1. Goedhart G, van der Wal MF, Cuijpers P, Bonsel GJ. Psychosocial problems and continued smoking during pregnancy. Addict Behav . 2009; 34( 4): 403– 406. Google Scholar CrossRef Search ADS PubMed  2. Smedberg J, Lupattelli A, Mårdby AC, Nordeng H. Characteristics of women who continue smoking during pregnancy: A cross-sectional study of pregnant women and new mothers in 15 European countries. BMC Pregnancy Childbirth . 2014; 14: 213. Google Scholar CrossRef Search ADS PubMed  3. Graham H, Hawkins SS, Law C. Lifecourse influences on women’s smoking before, during and after pregnancy. Soc Sci Med . 2010; 70( 4): 582– 587. Google Scholar CrossRef Search ADS PubMed  4. Penn G, Owen L. Factors associated with continued smoking during pregnancy: Analysis of socio-demographic, pregnancy and smoking-related factors. Drug Alcohol Rev . 2002; 21( 1): 17– 25. Google Scholar CrossRef Search ADS PubMed  5. Pickett KE, Wilkinson RG, Wakschlag LS. The psychosocial context of pregnancy smoking and quitting in the Millennium Cohort Study. J Epidemiol Community Health . 2009; 63( 6): 474– 480. Google Scholar CrossRef Search ADS PubMed  6. Siahpush M, Carlin JB. Financial stress, smoking cessation and relapse: Results from a prospective study of an Australian national sample. Addiction . 2006; 101( 1): 121– 127. Google Scholar CrossRef Search ADS PubMed  7. Siahpush M, Yong HH, Borland R, Reid JL, Hammond D. Smokers with financial stress are more likely to want to quit but less likely to try or succeed: Findings from the International Tobacco Control (ITC) Four Country Survey. Addiction . 2009; 104( 8): 1382– 1390. Google Scholar CrossRef Search ADS PubMed  8. Falba T, Teng HM, Sindelar JL, Gallo WT. The effect of involuntary job loss on smoking intensity and relapse. Addiction . 2005; 100( 9): 1330– 1339. Google Scholar CrossRef Search ADS PubMed  9. McKee SA, Maciejewski PK, Falba T, Mazure CM. Sex differences in the effects of stressful life events on changes in smoking status. Addiction . 2003; 98( 6): 847– 855. Google Scholar CrossRef Search ADS PubMed  10. Ruhm CJ. Healthy living in hard times. J Health Econ . 2005; 24( 2): 341– 363. Google Scholar CrossRef Search ADS PubMed  11. Xu X. The business cycle and health behaviors. Soc Sci Med . 2013; 77: 126– 136. Google Scholar CrossRef Search ADS PubMed  12. Ásgeirsdóttir TL, Corman H, Noonan K, Ólafsdóttir Þ, Reichman NE. Was the economic crisis of 2008 good for Icelanders? Impact on health behaviors. Econ Hum Biol . 2014; 13: 1– 19. Google Scholar CrossRef Search ADS PubMed  13. McClure CB, Valdimarsdóttir UA, Hauksdóttir A, Kawachi I. Economic crisis and smoking behaviour: prospective cohort study in Iceland. BMJ Open . 2012; 2( 5): e001386. Google Scholar CrossRef Search ADS PubMed  14. Filippidis FT, Schoretsaniti S, Dimitrakaki Cet al.   Trends in cardiovascular risk factors in Greece before and during the financial crisis: The impact of social disparities. Eur J Public Health . 2014; 24( 6): 974– 979. Google Scholar CrossRef Search ADS PubMed  15. Gallus S, Ghislandi S, Muttarak R. Effects of the economic crisis on smoking prevalence and number of smokers in the USA. Tob Control . 2015; 24( 1): 82– 88. Google Scholar CrossRef Search ADS PubMed  16. McKenna CS, Law C, Pearce A. Financial strain, parental smoking, and the Great Recession: An analysis of the UK Millennium Cohort Study. Nicotine Tob Res . 2017; 19(12):1521–1525. 17. Raynor P; Born in Bradford Collaborative Group. Born in Bradford, a cohort study of babies born in Bradford, and their parents: Protocol for the recruitment phase. BMC Public Health . 2008; 8: 327. Google Scholar CrossRef Search ADS PubMed  18. Wright J, Small N, Raynor Pet al.   Cohort profile: The born in bradford multi-ethnic family cohort study. Int J Epidemiol  2013; 42( 4): 978– 991. Google Scholar CrossRef Search ADS PubMed  19. Office for National Statistics (ONS). Jobseeker’s Allowance With Rates and Proportions . Newport: Wales; 2016. https://www.nomisweb.co.uk/. Accessed January 28, 2016. 20. Office for National Statistics (ONS). Population Estimates for UK, England and Wales, Scotland and Northern Ireland . Newport: Wales; 2016. http://bit.ly/1ObV5s7. Accessed June 28, 2016. 21. Oxford English Dictionary . Oxford, UK: Oxford University Press; 2016. www.oed.com. Accessed July 7, 2016. 22. LexisNexis UK; 2016. www.nexis.com. Accessed May 7, 2018. 23. Textor J, Hardt J, Knüppel S. DAGitty: A graphical tool for analyzing causal diagrams. Epidemiology . 2011; 22( 5): 745. Google Scholar CrossRef Search ADS PubMed  24. StataCorp. Stata Statistical Software: Release 12 . College Station, Texas: StataCorp LP; 2011. 25. Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J Pers Soc Psychol . 1986; 51( 6): 1173– 1182. Google Scholar CrossRef Search ADS PubMed  26. Hayes AF. Beyond baron and kenny: Statistical mediation analysis in the new millennium. Commun Monogr . 2009; 76( 4): 408– 420. Google Scholar CrossRef Search ADS   27. Zhao X, Lynch JG, Chen Q. Reconsidering Baron and Kenny: Myths and truths about mediation analysis. J Cons Res  2010; 37( 2): 197– 206. Google Scholar CrossRef Search ADS   28. Sobotka T, Skirbekk V, Philipov D. Economic recession and fertility in the developed world. Popul Dev Rev . 2011; 37( 2): 267– 306. Google Scholar CrossRef Search ADS PubMed  29. Goldstein J, Kreyenfeld M, Jasilioniene A, Örsal DDK. Fertility reactions to the” great recession” in Europe: Recent evidence from order-specific data. Demographic Res  2013; 29: 85– 104. Google Scholar CrossRef Search ADS   30. Office for National Statistics (ONS). Birth Summary Tables, England and Wales: 2014 . Newport: Wales; 2015. http://tinyurl.com/h99xmxq. Accessed July 21, 2016. PubMed PubMed  31. Health & Social Care Information Centre (HSCIC). Statistics on Women’s Smoking Status at Time of Delivery, England - Quarter 4, 2013–14 . Leeds, UK: NHS Digital; 2014. http://www.hscic.gov.uk/catalogue/PUB14258. Accessed July 21, 2016. 32. Mackay DF, Nelson SM, Haw SJ, Pell JP. Impact of Scotland’s smoke-free legislation on pregnancy complications: Retrospective cohort study. PLoS Med . 2012; 9( 3): e1001175. Google Scholar CrossRef Search ADS PubMed  33. Bacigalupe A, Escolar-Pujolar A. The impact of economic crises on social inequalities in health: What do we know so far? Int J Equity Health . 2014; 13: 52. Google Scholar CrossRef Search ADS PubMed  34. Stuckler D, Basu S, Suhrcke M, Coutts A, McKee M. The public health effect of economic crises and alternative policy responses in Europe: An empirical analysis. Lancet . 2009; 374( 9686): 315– 323. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nicotine and Tobacco Research Oxford University Press

Using Birth Cohort Data to Assess the Impact of the UK 2008–2010 Economic Recession on Smoking During Pregnancy

Loading next page...
 
/lp/ou_press/using-birth-cohort-data-to-assess-the-impact-of-the-uk-2008-2010-ePLJC4q1Ek
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
1462-2203
eISSN
1469-994X
D.O.I.
10.1093/ntr/nty083
Publisher site
See Article on Publisher Site

Abstract

Abstract Introduction Despite the well-known link between stress and smoking, evidence for associations between economic recession, financial stress, and smoking is contradictory. In this study, we assess whether women were more likely to continue smoking during pregnancy if they were exposed to the UK 2008–2010 economic recession during pregnancy than those who were unexposed, and whether this relationship is mediated by financial stress. Methods We used cross-sectional data on 2775 pregnant women who were regular smokers before pregnancy and who were enrolled in the UK Born in Bradford cohort study between March 2007 and December 2010. The cutoff date for exposure to recession was set as August 1, 2008, based on local and national economic data. Multivariable logistic regression analysis included potential confounders: maternal age, parity, cohabitation, ethnicity, and maternal age. The mediating role of financial stress was analyzed using “worse off financially” and a “difficult financial situation” as indicators of financial stress in Sobel–Goodman mediation tests with bootstrap resampling. Results After taking into account potential confounders, exposure to recession was associated with continued smoking during pregnancy (OR = 1.19, 95% CI = 1.01 to 1.41, p = 0.03). A worse financial situation and a difficult financial situation were identified as mediators, explaining 8.4% and 17.6%, respectively, of the relationship between exposure to recession and smoking during pregnancy. Conclusions Smoking during pregnancy is associated with exposure to the UK 2008–2010 economic recession during pregnancy, and this relationship is partly mediated by financial stress. Implications Health inequalities in smoking during pregnancy are affected by economic recession, as those who are most likely to smoke are also most likely to experience the financial stress resulting from economic recession. Socioeconomic conditions at the societal and individual level are important targets when aiming to reduce rates of smoking during pregnancy. Introduction Social Gradients in Tobacco Smoking Tobacco smoking in pregnancy is more prevalent among those who have a lower socioeconomic status as indicated by a range of socioeconomic measures ranging from education and employment status to poor childhood circumstances, housing tenure, and job strain.1–4 In an analysis of continued smoking during pregnancy in 15 European countries, Smedberg and colleagues2 found an average prevalence rate of 26.2%, compared with 43.2% in the lowest educational category. In England, a combination of socioeconomic indicators showed a very steep social gradient with 6% of women smoking during pregnancy in the highest and 62% of women smoking in the lowest socioeconomic group.4 An analysis of data from the Millennium Cohort Study showed that smoking and heavy smoking during pregnancy were associated with a range of factors more common among lower socioeconomic groups, including leaving school before eighteen, not having a bank account, and financial difficulties.5 Tobacco Smoking and Financial Stress Financial stress appears to play a significant role in explaining these social gradients, although most of the evidence comes from smoking in the general population. In Australia, financial stress according to an eight-item scale relating to shortage of money was associated with a lower likelihood of quitting and a higher likelihood of relapse.6 This finding was confirmed in a survey with data from the United States, United Kingdom, Canada, and Australia. Smokers experiencing financial stress had an increased interest in quitting but were less likely to make an attempt, and consequently their odds of succeeding was half that of the group without financial stress.7 Involuntary job loss among older workers was associated with increased odds of a relapse in smoking, defined as reported smoking point prevalence after quitting.8 A US community-based sample was used to demonstrate that women are less likely to quit smoking in response to an adverse financial event, and more likely to relapse as a result of such an event.9 Tobacco Smoking During the 2008–2010 Economic Recession Given that financial stress affects those of lower socioeconomic status disproportionally, and macro-economic circumstances such as economic recession can evoke financial stress, it is reasonable to assume that the economic recession, which hit Europe between 2008 and 2010, would have increased rates of smoking during pregnancy. Women who were regular smokers may have been less able to quit smoking as financial strain placed a burden on their cognitive, social, and emotional capacities. The evidence, however, which is largely based on tobacco use in the general population, is contradictory. Based on US economic data, Ruhm concluded that smokers, especially heavy smokers, cut down on smoking during economic downturns.10 Others have showed that increased wages lead to increased cigarette use,11 and larger than expected declines in smoking during the economic recession have been found in Iceland12,13 and in Greece.14 These findings have come with nuances and reservations. The authors of the Greek study pointed out that more effective enforcement of tobacco control legislation may partly explain their findings.14 Gallus and colleagues argued that the number of unemployed people grows during a recession, so that a small decrease in smoking is seen at the ecological level because the downward trend in smoking prevalence for employed people is slightly larger than the increase in smoking among the unemployed during economic recession.15 In a study of UK parents, financial strain during the recession was associated with an increased risk of persistent tobacco use and relapse.16 In summary, most of the literature would suggest that experiences of financial stress, which are likely to result from an economic recession, might increase the likelihood of continued smoking during pregnancy. Although these associations are plausible, the literature on recession and smoking suggests an effect in the opposite direction. An analysis specifically testing associations between economic recession and smoking during pregnancy, including the mediating effect of financial stress, might clarify previous findings. Hypothesis In a deprived UK community hit hard by the UK 2008–2010 economic recession, women who smoke regularly will be more likely to continue smoking during pregnancy if they were exposed to the recession during pregnancy compared with those unexposed, and this relationship is mediated by financial stress. Methods This study follows guidance on reporting set out in the STROBE statement (Supplementary Table A1). Data The Born in Bradford (BiB) cohort was set up to examine the impact of social, environmental, psychological, behavioral, and biological factors on maternal and child health and well-being.17 Pregnant women (N = 12450) were recruited around 26 to 28 weeks of pregnancy at Bradford’s only maternity unit when attending universal screening for gestational diabetes, between March 2007 and December 2010.18 Ethical approval for data collection was granted by Bradford Research Ethics Committee (Ref 07/H1302/112). After excluding pregnancies without baseline questionnaire data (N = 2377), stillbirths (N = 61), second and third pregnancies to the same mother in the cohort (N = 1323), and twins and triplets (N = 142), a dataset of 10035 mother–infant pairs was obtained. A dataset with complete data for the exposure variable and covariates was used. This dataset of 8952 pregnant women was used to explore relationships between exposure to recession and financial stress. In this dataset, 2775 women were regular smokers and reported information on smoking during pregnancy and relevant covariates. Setting Bradford is a deprived and ethnically diverse city in the North of England with a population of over half a million.19 Employment is more reliant on manufacturing industry than the UK average and this sector has been in decline for decades, making the city particularly vulnerable to the effects of economic recession. There was a sharp increase in the proportion of claimants for Job-Seeker’s Allowance (JSA) (the main form of financial support for unemployed people identified as actively seeking work) in Bradford from August 2008, with a peak in JSA claimants at 5.0% in September 2009 (Figure 1).20 In November 2007, JSA-claimant-rates in Bradford were 133% higher than for England as a whole; by November 2015, rates were 181% higher in Bradford than the average for England. Figure 1. View largeDownload slide Percentage JSA claimants as a proportion of the resident population aged 16–64. Figure 1. View largeDownload slide Percentage JSA claimants as a proportion of the resident population aged 16–64. Main Exposure The definition of recession as “A period of temporary economic decline during which trade and industrial activity are reduced, generally identified by a fall in GDP in two successive quarters”21 may not correspond with the lived experience of recession. We used reports in national and local media22 (Supplementary Box A1) in combination with data on unemployment related benefit claims21 (Figure 1) to inform the decision of a cutoff date for recession start date to be the first of August 2008; an estimated conception date from August 01, 2008 onwards was therefore classified as “exposed.” No “end of exposure” cutoff was used as Figure 1 indicates the economic impact of the recession lasted well beyond the study recruitment period. The percentage of study participants reporting to be financially worse off than a year ago increases from 15.3% (N = 60) at the beginning of May 2008 to 24.2% (N = 88) around the beginning of June 2008 (Figure 2). In an exploration of different measures of exposure, we found no evidence for an annual worsening of individual’s financial situation (analyses available upon request). A binary cutoff before and since August 01, 2008 was therefore considered an appropriate measure of exposure to recession. Figure 2. View largeDownload slide Percentage of pregnant women reporting to be worse off financially by year and quarter of recruitment. Figure 2. View largeDownload slide Percentage of pregnant women reporting to be worse off financially by year and quarter of recruitment. Mediating Factors Financial stress was operationalized through two variables; perceived financial situation and change in perceived financial situation. Women were asked how well they and their partner were managing financially, with response categories ranging from “living comfortably” to “difficult” (“very difficult” and “just about getting by” were merged with “difficult”). Change in perceived financial situation was measured with the question “Compared to a year ago, how would you say you and your husband/partner are doing financially now?,” with answer categories “better off,” “about the same” or “worse off.” Outcomes Data on maternal smoking were obtained from the study baseline questionnaire. A binary variable of continued smoking during pregnancy for those who reported to smoke regularly was derived from questions on smoking during pregnancy 3 months before pregnancy, in the first 3 months of pregnancy and since the beginning of the fourth month. Women who reported to have stopped smoking in the first month of pregnancy and did not report smoking any cigarettes since the beginning of the fourth month were counted as nonsmokers during pregnancy. Covariates The Directed Acyclic Graph created in DAGgity V2.3 and published on the Daggity Web site (http://dagitty.net/mLWfIRw) shows the hypothesized causal relationships between exposure to recession, continued smoking during pregnancy and covariates.23 Parity (nulliparous vs. other), maternal age, cohabitation with a partner (yes/no) ethnic group (White British, Pakistani, other), and education of the mother (<5 GCSE, 5 GCSE, A level, >A level, other, or equivalents of these qualifications) were identified as potential confounders. Statistical Analyses Firstly, sample characteristics were explored and differences tested between the subsample exposed to recession during pregnancy and the unexposed group. Chi-square tests were used for binary variables and t-tests for continuous variables. Multivariable logistic regression analysis was used to model the relationship between exposure to economic recession and continued smoking during pregnancy, after which mediation analysis was performed to assess the role of financials stress. The mediating role of financial stress was tested with Sobel–Goodman mediation analyses in Stata 12, using the “sgmediation” and “bootstrap” commands.24 This method has been found to be more rigorous and more likely to identify a true mediation effect than the widely used causal steps outlined by Baron and Kenny.25–27 We avoided violation of the normality assumption of the Sobel–Goodman test by performing a bootstrap analysis with 5000 sampling repetitions.26 Results The sample consists of 2775 women who were regular smokers before pregnancy, and Table 1 shows an overview of their characteristics. As opposed to those recruited before the economic recession, pregnant women who were recruited after the start of the recession and were thus exposed were more likely to be nulliparous (first child), to have a higher level of education, to be in a difficult financial situation, and to be worse off than a year ago. Prevalence rates of smoking during pregnancy varied from 40.0% among those who reported to be managing well financially, to 65.2% among mothers who reported their financial situation was difficult/very difficult. Table 1. Participant Characteristics by Exposure Group Participant characteristics  Unexposed N = 1362  Exposed N = 1413  Total N = 2775  p value test of difference between groups*  Maternal age (years)  25.79  26.13  25.97  0.06  Parity        0.03   Nulliparous  50.44%  54.56%  52.57%   Multiparous  49.56%  45.44%  47.43%  Ethnicity        0.35   White British  77.97%  76.57%  77.26%   Pakistani  12.19%  11.89%  12.04%   Other  9.84%  11.54%  10.70%  Maternal education        <0.01   <5 GCSE  25.40%  23.50%  24.43%   5 GCSE  38.11%  36.73%  37.41%   A level  13.44%  18.75%  16.14%   >A level  14.54%  15.64%  15.10%   Other  8.52%  5.38%  6.92%  Cohabiting        0.11   Yes  69.53%  66.67%  68.07%   No  30.47%  33.33%  31.93%  Financial situation        0.02   Comfortable  23.79%  21.80%  22.77%   Doing alright  40.53%  37.30%  38.88%   Difficult  35.68%  40.91%  38.34%  Change in financial situation        <0.01   Same/better off  74.23%  69.57%  71.86%   Worse off  25.77%  30.43%  28.14%  Continued smoking during pregnancy        0.14   Yes  53.52%  56.33%  54.95%   No  46.48%  43.67%  45.05%  Participant characteristics  Unexposed N = 1362  Exposed N = 1413  Total N = 2775  p value test of difference between groups*  Maternal age (years)  25.79  26.13  25.97  0.06  Parity        0.03   Nulliparous  50.44%  54.56%  52.57%   Multiparous  49.56%  45.44%  47.43%  Ethnicity        0.35   White British  77.97%  76.57%  77.26%   Pakistani  12.19%  11.89%  12.04%   Other  9.84%  11.54%  10.70%  Maternal education        <0.01   <5 GCSE  25.40%  23.50%  24.43%   5 GCSE  38.11%  36.73%  37.41%   A level  13.44%  18.75%  16.14%   >A level  14.54%  15.64%  15.10%   Other  8.52%  5.38%  6.92%  Cohabiting        0.11   Yes  69.53%  66.67%  68.07%   No  30.47%  33.33%  31.93%  Financial situation        0.02   Comfortable  23.79%  21.80%  22.77%   Doing alright  40.53%  37.30%  38.88%   Difficult  35.68%  40.91%  38.34%  Change in financial situation        <0.01   Same/better off  74.23%  69.57%  71.86%   Worse off  25.77%  30.43%  28.14%  Continued smoking during pregnancy        0.14   Yes  53.52%  56.33%  54.95%   No  46.48%  43.67%  45.05%  *χ2 test for categorical variables and t-test for continuous variables. View Large Table 1. Participant Characteristics by Exposure Group Participant characteristics  Unexposed N = 1362  Exposed N = 1413  Total N = 2775  p value test of difference between groups*  Maternal age (years)  25.79  26.13  25.97  0.06  Parity        0.03   Nulliparous  50.44%  54.56%  52.57%   Multiparous  49.56%  45.44%  47.43%  Ethnicity        0.35   White British  77.97%  76.57%  77.26%   Pakistani  12.19%  11.89%  12.04%   Other  9.84%  11.54%  10.70%  Maternal education        <0.01   <5 GCSE  25.40%  23.50%  24.43%   5 GCSE  38.11%  36.73%  37.41%   A level  13.44%  18.75%  16.14%   >A level  14.54%  15.64%  15.10%   Other  8.52%  5.38%  6.92%  Cohabiting        0.11   Yes  69.53%  66.67%  68.07%   No  30.47%  33.33%  31.93%  Financial situation        0.02   Comfortable  23.79%  21.80%  22.77%   Doing alright  40.53%  37.30%  38.88%   Difficult  35.68%  40.91%  38.34%  Change in financial situation        <0.01   Same/better off  74.23%  69.57%  71.86%   Worse off  25.77%  30.43%  28.14%  Continued smoking during pregnancy        0.14   Yes  53.52%  56.33%  54.95%   No  46.48%  43.67%  45.05%  Participant characteristics  Unexposed N = 1362  Exposed N = 1413  Total N = 2775  p value test of difference between groups*  Maternal age (years)  25.79  26.13  25.97  0.06  Parity        0.03   Nulliparous  50.44%  54.56%  52.57%   Multiparous  49.56%  45.44%  47.43%  Ethnicity        0.35   White British  77.97%  76.57%  77.26%   Pakistani  12.19%  11.89%  12.04%   Other  9.84%  11.54%  10.70%  Maternal education        <0.01   <5 GCSE  25.40%  23.50%  24.43%   5 GCSE  38.11%  36.73%  37.41%   A level  13.44%  18.75%  16.14%   >A level  14.54%  15.64%  15.10%   Other  8.52%  5.38%  6.92%  Cohabiting        0.11   Yes  69.53%  66.67%  68.07%   No  30.47%  33.33%  31.93%  Financial situation        0.02   Comfortable  23.79%  21.80%  22.77%   Doing alright  40.53%  37.30%  38.88%   Difficult  35.68%  40.91%  38.34%  Change in financial situation        <0.01   Same/better off  74.23%  69.57%  71.86%   Worse off  25.77%  30.43%  28.14%  Continued smoking during pregnancy        0.14   Yes  53.52%  56.33%  54.95%   No  46.48%  43.67%  45.05%  *χ2 test for categorical variables and t-test for continuous variables. View Large Multivariable Analysis After taking into account potential confounders, exposure to recession was associated with continued smoking during pregnancy (OR = 1.19, 95% CI = 1.01 to 1.41, p = 0.03) (Table 2). All other factors being equal, this represents a 3.7% difference in smoking prevalence, with 53.1% of women in the unexposed and 56.8% of women in the exposed group smoking during pregnancy. Table 2. Exposure to Recession in Relation to Smoking During Pregnancy   Model 1. Bivariate “exposure recession” N = 2775  Model 2. Multivariate “exposure recession” N = 2775  OR  95% CI  OR  95% CI  Exposure recession  1.12  0.96; 1.30  1.19  1.01;1.41*  Maternal age      0.93  0.92; 0.95***  Parity           Multiparous      1.37  1.14;1.65**  Ethnic group (vs. White British)           Pakistani      0.61  0.48;0.79***   Other      0.79  0.61;1.04  Maternal education (vs. <5 GCSE)           5 GCSE      0.50  0.40;0.63***   A level      0.40  0.31;0.53***   >A level      0.19  0.14;0.25***   Other      0.33  0.24;0.47***  Cohabitation           Yes      0.53  0.44;0.64***    Model 1. Bivariate “exposure recession” N = 2775  Model 2. Multivariate “exposure recession” N = 2775  OR  95% CI  OR  95% CI  Exposure recession  1.12  0.96; 1.30  1.19  1.01;1.41*  Maternal age      0.93  0.92; 0.95***  Parity           Multiparous      1.37  1.14;1.65**  Ethnic group (vs. White British)           Pakistani      0.61  0.48;0.79***   Other      0.79  0.61;1.04  Maternal education (vs. <5 GCSE)           5 GCSE      0.50  0.40;0.63***   A level      0.40  0.31;0.53***   >A level      0.19  0.14;0.25***   Other      0.33  0.24;0.47***  Cohabitation           Yes      0.53  0.44;0.64***  *p < .05, ** p < .01, *** p < .001 View Large Table 2. Exposure to Recession in Relation to Smoking During Pregnancy   Model 1. Bivariate “exposure recession” N = 2775  Model 2. Multivariate “exposure recession” N = 2775  OR  95% CI  OR  95% CI  Exposure recession  1.12  0.96; 1.30  1.19  1.01;1.41*  Maternal age      0.93  0.92; 0.95***  Parity           Multiparous      1.37  1.14;1.65**  Ethnic group (vs. White British)           Pakistani      0.61  0.48;0.79***   Other      0.79  0.61;1.04  Maternal education (vs. <5 GCSE)           5 GCSE      0.50  0.40;0.63***   A level      0.40  0.31;0.53***   >A level      0.19  0.14;0.25***   Other      0.33  0.24;0.47***  Cohabitation           Yes      0.53  0.44;0.64***    Model 1. Bivariate “exposure recession” N = 2775  Model 2. Multivariate “exposure recession” N = 2775  OR  95% CI  OR  95% CI  Exposure recession  1.12  0.96; 1.30  1.19  1.01;1.41*  Maternal age      0.93  0.92; 0.95***  Parity           Multiparous      1.37  1.14;1.65**  Ethnic group (vs. White British)           Pakistani      0.61  0.48;0.79***   Other      0.79  0.61;1.04  Maternal education (vs. <5 GCSE)           5 GCSE      0.50  0.40;0.63***   A level      0.40  0.31;0.53***   >A level      0.19  0.14;0.25***   Other      0.33  0.24;0.47***  Cohabitation           Yes      0.53  0.44;0.64***  *p < .05, ** p < .01, *** p < .001 View Large Mediation Analysis Results of the mediation analysis are summarized in Table 3. Table 3. Results of Sobel–Goodman Test With Bootstrap Analysis   Coefficient  SE  95% CI  Model with mediator “worse financial situation” (N = 2775)   Indirect effect*  0.003  0.001  0.000 to 0.001   Direct effect  0.031  0.017  −0.003 to 0.066  Model with mediator “difficult financial situation” (N = 2775)   Indirect effect**  0.006  0.002  0.002 to 0.011   Direct effect  0.028  0.018  −0.007 to 0.063    Coefficient  SE  95% CI  Model with mediator “worse financial situation” (N = 2775)   Indirect effect*  0.003  0.001  0.000 to 0.001   Direct effect  0.031  0.017  −0.003 to 0.066  Model with mediator “difficult financial situation” (N = 2775)   Indirect effect**  0.006  0.002  0.002 to 0.011   Direct effect  0.028  0.018  −0.007 to 0.063  *Proportion of the effect that is mediated: 0.084. **Proportion of the effect that is mediated: 0.176. View Large Table 3. Results of Sobel–Goodman Test With Bootstrap Analysis   Coefficient  SE  95% CI  Model with mediator “worse financial situation” (N = 2775)   Indirect effect*  0.003  0.001  0.000 to 0.001   Direct effect  0.031  0.017  −0.003 to 0.066  Model with mediator “difficult financial situation” (N = 2775)   Indirect effect**  0.006  0.002  0.002 to 0.011   Direct effect  0.028  0.018  −0.007 to 0.063    Coefficient  SE  95% CI  Model with mediator “worse financial situation” (N = 2775)   Indirect effect*  0.003  0.001  0.000 to 0.001   Direct effect  0.031  0.017  −0.003 to 0.066  Model with mediator “difficult financial situation” (N = 2775)   Indirect effect**  0.006  0.002  0.002 to 0.011   Direct effect  0.028  0.018  −0.007 to 0.063  *Proportion of the effect that is mediated: 0.084. **Proportion of the effect that is mediated: 0.176. View Large When a worse financial situation than a year ago is used as the indication of financial stress, bootstrap results for the Sobel–Goodman test suggests a modest mediation effect (observed coefficient = 0.003, 95% CI = 0.00 to 0.01). Of the total relationship between exposure to recession and continued smoking during pregnancy, 8.4% is estimated to be mediated by the effect of a worse financial situation. When financial stress is operationalized as a current difficult financial situation (compared to comfortable/doing alright), a modest mediation effect is again present (observed coefficient = 0.006, 95% CI = 0.00 to 0.01). The proportion of the relationship between exposure to recession and continued smoking during pregnancy mediated by the effect of a difficult financial situation is 17.6%. By taking into account the influence of financial stress, either through the inclusion of a worse or difficult financial situation variable, the direct effect of exposure to the economic recession on smoking during pregnancy is no longer statistically significant. Conclusions This study demonstrates that exposure to economic recession is associated with continued smoking during pregnancy for regular smokers and that this relationship is partly mediated by financial stress. Exposure to the UK 2008–2010 economic recession during pregnancy was associated with a difficult financial situation and being worse off than a year ago for women in the Born in Bradford cohort study who were regular smokers. After taking into account key confounders, pregnant women exposed to recession were more likely to continue smoking during pregnancy than those not exposed. Mediation analyses showed that the relationship between exposure to recession and continued smoking during pregnancy was partly explained by financial stress, either measured by a difficult or worsened financial situation of the household. With the inclusion of financial stress in the model, the relationship between exposure to recession and smoking during pregnancy was no longer statistically significant. Limitations Even though covariates were included to adjust for differences between the “exposed” and “unexposed,” there might be alternative explanations for our findings. Couples who get pregnant during an economic recession may be different from those who choose not to. This may explain differences between the unexposed and exposed group shown in Table 1. Economic recession and unemployment have been found to be associated with a decline in fertility rates, particularly for young women having their first child,28,29 and fertility rates in England and Wales declined between 2008 and 2010 in all age groups except for mothers aged 35 and above.30 If couples who were more strongly affected by recession were more likely to delay pregnancies, we have underestimated the relationship between economic recession, financial stress, and smoking during pregnancy. Our study is limited by missing data and by its cross-sectional design. Despite our careful considerations of the cutoff for exposure to recession by using area-specific and national data, it is possible that we did not identify correctly after which date people were likely to experience the economic recession in their lives. In addition to misspecification of the exposure cutoff, data on key covariates may be influenced by date of data collection, as researchers will have gradually become more experienced in the administration of the questionnaire. It is likely that data are not missing at random. Finally, smoking during pregnancy is a stigmatized behavior that is likely to be underreported when relying on self-report measures, which means our prevalence rates of smoking during pregnancy are likely to underestimate true prevalence rates. Implications for Research and Society This study contradicts findings which suggest a beneficial effect of economic recession on smoking, and fits with literature on the role of stress, including financial stress, in a person’s ability to quit smoking and prevent relapse.5–9,16 In our sample, prevalence rates of smoking during pregnancy increased over time, whereas there is a slow but steady decline in smoking during pregnancy nationally.31 In Scotland, the introduction of a smoking ban in public places led to reduced rates of smoking during pregnancy.32 Our data show no such effect on smoking during pregnancy around the time the smoking ban took effect in England in July 2007. Any downward trends in smoking prevalence may have been counteracted by the larger impact of the economic recession. All women in this study were sampled from a UK city characterized by deprivation and health inequalities. The increased burden placed on this already disadvantaged community was by no means unavoidable. Research shows that strong social policies can prevent health impacts of recession.33,34 Our study shows that socioeconomic conditions, both at the societal and individual level, are important targets when aiming to reduce rates of smoking during pregnancy. If there is ever a time to invest in protecting the vulnerable to the advantage of society as a whole, the period during and after recession is such a time. Supplementary Material Supplementary Table A1 and Box A1 can be found online at https://academic.oup.com/gerontologist/. Funding This work was supported by the Born in Bradford study funding. The BiB study presents independent research commissioned by the National Institute for Health Research Collaboration for Applied Health Research and Care (NIHR CLAHRC-2013-10022) and the NIHR Programme Grants for Applied Research funding scheme [grant number RP-PG-0407-10044]. Core support for BiB is also provided by the Wellcome Trust [grant number WT101597MA]. All authors receive funding from the Big Lottery Fund as part of the “A Better Start” programme. Declaration of interests The authors report no conflicts of interest. Acknowledgments Born in Bradford is only possible because of the enthusiasm and commitment of the children and parents in the study. We are grateful to all the participants, health professionals and researchers who make Born in Bradford happen. References 1. Goedhart G, van der Wal MF, Cuijpers P, Bonsel GJ. Psychosocial problems and continued smoking during pregnancy. Addict Behav . 2009; 34( 4): 403– 406. Google Scholar CrossRef Search ADS PubMed  2. Smedberg J, Lupattelli A, Mårdby AC, Nordeng H. Characteristics of women who continue smoking during pregnancy: A cross-sectional study of pregnant women and new mothers in 15 European countries. BMC Pregnancy Childbirth . 2014; 14: 213. Google Scholar CrossRef Search ADS PubMed  3. Graham H, Hawkins SS, Law C. Lifecourse influences on women’s smoking before, during and after pregnancy. Soc Sci Med . 2010; 70( 4): 582– 587. Google Scholar CrossRef Search ADS PubMed  4. Penn G, Owen L. Factors associated with continued smoking during pregnancy: Analysis of socio-demographic, pregnancy and smoking-related factors. Drug Alcohol Rev . 2002; 21( 1): 17– 25. Google Scholar CrossRef Search ADS PubMed  5. Pickett KE, Wilkinson RG, Wakschlag LS. The psychosocial context of pregnancy smoking and quitting in the Millennium Cohort Study. J Epidemiol Community Health . 2009; 63( 6): 474– 480. Google Scholar CrossRef Search ADS PubMed  6. Siahpush M, Carlin JB. Financial stress, smoking cessation and relapse: Results from a prospective study of an Australian national sample. Addiction . 2006; 101( 1): 121– 127. Google Scholar CrossRef Search ADS PubMed  7. Siahpush M, Yong HH, Borland R, Reid JL, Hammond D. Smokers with financial stress are more likely to want to quit but less likely to try or succeed: Findings from the International Tobacco Control (ITC) Four Country Survey. Addiction . 2009; 104( 8): 1382– 1390. Google Scholar CrossRef Search ADS PubMed  8. Falba T, Teng HM, Sindelar JL, Gallo WT. The effect of involuntary job loss on smoking intensity and relapse. Addiction . 2005; 100( 9): 1330– 1339. Google Scholar CrossRef Search ADS PubMed  9. McKee SA, Maciejewski PK, Falba T, Mazure CM. Sex differences in the effects of stressful life events on changes in smoking status. Addiction . 2003; 98( 6): 847– 855. Google Scholar CrossRef Search ADS PubMed  10. Ruhm CJ. Healthy living in hard times. J Health Econ . 2005; 24( 2): 341– 363. Google Scholar CrossRef Search ADS PubMed  11. Xu X. The business cycle and health behaviors. Soc Sci Med . 2013; 77: 126– 136. Google Scholar CrossRef Search ADS PubMed  12. Ásgeirsdóttir TL, Corman H, Noonan K, Ólafsdóttir Þ, Reichman NE. Was the economic crisis of 2008 good for Icelanders? Impact on health behaviors. Econ Hum Biol . 2014; 13: 1– 19. Google Scholar CrossRef Search ADS PubMed  13. McClure CB, Valdimarsdóttir UA, Hauksdóttir A, Kawachi I. Economic crisis and smoking behaviour: prospective cohort study in Iceland. BMJ Open . 2012; 2( 5): e001386. Google Scholar CrossRef Search ADS PubMed  14. Filippidis FT, Schoretsaniti S, Dimitrakaki Cet al.   Trends in cardiovascular risk factors in Greece before and during the financial crisis: The impact of social disparities. Eur J Public Health . 2014; 24( 6): 974– 979. Google Scholar CrossRef Search ADS PubMed  15. Gallus S, Ghislandi S, Muttarak R. Effects of the economic crisis on smoking prevalence and number of smokers in the USA. Tob Control . 2015; 24( 1): 82– 88. Google Scholar CrossRef Search ADS PubMed  16. McKenna CS, Law C, Pearce A. Financial strain, parental smoking, and the Great Recession: An analysis of the UK Millennium Cohort Study. Nicotine Tob Res . 2017; 19(12):1521–1525. 17. Raynor P; Born in Bradford Collaborative Group. Born in Bradford, a cohort study of babies born in Bradford, and their parents: Protocol for the recruitment phase. BMC Public Health . 2008; 8: 327. Google Scholar CrossRef Search ADS PubMed  18. Wright J, Small N, Raynor Pet al.   Cohort profile: The born in bradford multi-ethnic family cohort study. Int J Epidemiol  2013; 42( 4): 978– 991. Google Scholar CrossRef Search ADS PubMed  19. Office for National Statistics (ONS). Jobseeker’s Allowance With Rates and Proportions . Newport: Wales; 2016. https://www.nomisweb.co.uk/. Accessed January 28, 2016. 20. Office for National Statistics (ONS). Population Estimates for UK, England and Wales, Scotland and Northern Ireland . Newport: Wales; 2016. http://bit.ly/1ObV5s7. Accessed June 28, 2016. 21. Oxford English Dictionary . Oxford, UK: Oxford University Press; 2016. www.oed.com. Accessed July 7, 2016. 22. LexisNexis UK; 2016. www.nexis.com. Accessed May 7, 2018. 23. Textor J, Hardt J, Knüppel S. DAGitty: A graphical tool for analyzing causal diagrams. Epidemiology . 2011; 22( 5): 745. Google Scholar CrossRef Search ADS PubMed  24. StataCorp. Stata Statistical Software: Release 12 . College Station, Texas: StataCorp LP; 2011. 25. Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J Pers Soc Psychol . 1986; 51( 6): 1173– 1182. Google Scholar CrossRef Search ADS PubMed  26. Hayes AF. Beyond baron and kenny: Statistical mediation analysis in the new millennium. Commun Monogr . 2009; 76( 4): 408– 420. Google Scholar CrossRef Search ADS   27. Zhao X, Lynch JG, Chen Q. Reconsidering Baron and Kenny: Myths and truths about mediation analysis. J Cons Res  2010; 37( 2): 197– 206. Google Scholar CrossRef Search ADS   28. Sobotka T, Skirbekk V, Philipov D. Economic recession and fertility in the developed world. Popul Dev Rev . 2011; 37( 2): 267– 306. Google Scholar CrossRef Search ADS PubMed  29. Goldstein J, Kreyenfeld M, Jasilioniene A, Örsal DDK. Fertility reactions to the” great recession” in Europe: Recent evidence from order-specific data. Demographic Res  2013; 29: 85– 104. Google Scholar CrossRef Search ADS   30. Office for National Statistics (ONS). Birth Summary Tables, England and Wales: 2014 . Newport: Wales; 2015. http://tinyurl.com/h99xmxq. Accessed July 21, 2016. PubMed PubMed  31. Health & Social Care Information Centre (HSCIC). Statistics on Women’s Smoking Status at Time of Delivery, England - Quarter 4, 2013–14 . Leeds, UK: NHS Digital; 2014. http://www.hscic.gov.uk/catalogue/PUB14258. Accessed July 21, 2016. 32. Mackay DF, Nelson SM, Haw SJ, Pell JP. Impact of Scotland’s smoke-free legislation on pregnancy complications: Retrospective cohort study. PLoS Med . 2012; 9( 3): e1001175. Google Scholar CrossRef Search ADS PubMed  33. Bacigalupe A, Escolar-Pujolar A. The impact of economic crises on social inequalities in health: What do we know so far? Int J Equity Health . 2014; 13: 52. Google Scholar CrossRef Search ADS PubMed  34. Stuckler D, Basu S, Suhrcke M, Coutts A, McKee M. The public health effect of economic crises and alternative policy responses in Europe: An empirical analysis. Lancet . 2009; 374( 9686): 315– 323. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Nicotine and Tobacco ResearchOxford University Press

Published: May 7, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off