A Citywide Smoking Ban Reduced Maternal Smoking and Risk for Preterm Births: A Colorado Natural Experiment

Abstract

Background:

Few reports exist on the association of a public smoking ban with fetal outcomes and maternal smoking in the United States. We sought to evaluate the effect of a citywide smoking ban in comparison to a like municipality with no such ban in Colorado on maternal smoking and subsequent fetal birth outcomes.

Methods:

A citywide smoking ban in Colorado provided a natural experiment. The experimental citywide smoking ban site was implemented in Pueblo, Colorado. A comparison community was chosen that had no smoking ban, El Paso County, with similar characteristics of population, size, and geography. The two sites served as their own controls, as each had a preban and postban retrospective observation period: preban was April 1, 2001, to July 1, 2003; postban was April 1, 2004, to July 1, 2006. Outcomes were maternal smoking (self-report), low birth weight (LBW) (defined as <2500 g or as <3000 g), and preterm births (<37 weeks gestation) in singleton births from mothers residing in these cities and reported to the State Department of Public Health. A difference-in-differences estimator was used to account for site and temporal trends in multivariate models.

Results:

Compared to El Paso County preban, the odds of maternal smoking and preterm births were, respectively, 38% (p<0.05) and 23% (p<0.05) lower in Pueblo. The odds for LBW births decreased by 8% for <3000 g and increased by 8.4% for <2500 g; however, neither was significant.

Conclusions:

This is the first evidence in the United States that population-level intervention using a smoking ban improved maternal and fetal outcomes, measured as maternal smoking and preterm births.

Introduction

Although tobacco use is largely preventable, its sequelae remain a leading cause of death worldwide, both from direct and indirect exposure.¹ Within the United States, secondhand tobacco smoke was associated with over 3000 deaths from lung cancer and 46,000 deaths from heart disease in 2005.¹ An especially pernicious result of secondhand smoke is the risk to developing fetuses, which has demonstrated immediate and long-term health consequences. For example, prenatal exposure of tobacco smoke, both primary and environmental, has been associated with both low birth weight (LBW) and premature births.^2
–4 Both LBW and preterm birth are important risk factors for perinatal morbidity and mortality and have even been associated with an increased risk of cardiovascular disease (CVD) later in adulthood.^5,6

Public health education campaigns have been implemented as population-level interventions that aim to increase awareness of the dangers of secondhand smoke. Many jurisdictions worldwide have enacted legislation to restrict smoking in public places and in work places in the hope of influencing smoking behavior of the populations they affect and, ultimately, social norms.⁷ The public health basis for public smoking bans is derived from evidence that secondhand smoke has been implicated in the incidence of cancer, CVD, emphysema, bronchitis, and asthma in both adults and children.³ Additionally, secondhand smoke exposure clearly affects the development of asthma and worsening of respiratory function in children.³ The effect of antismoking policies in pregnant women is of considerable interest from a public health and cost perspective, based on the well-described relationship between maternal smoking and fetal health. Reducing health effects of smoking has the potential to reduce lifelong complications and the economic burden associated with preterm births.⁸ Based on 2006 data from the Institute of Medicine (IOM), the high rate of premature births in the United States constitutes a public health concern that costs society at least $26 billion a year.^7

–12 We seek to add to the literature on public health legislation, maternal smoking, and birth outcomes in the United States.

This article contributes to the debate about the effectiveness of legislative smoking bans on behavioral and health outcomes in three ways: (1) we use a natural experiment to test the effect of a smoking ban in comparison to a like municipality with no ban in Colorado, (2) we obtained health department records to measure both maternal and fetal outcomes in the same time period, and (3) we use a difference-in-differences study design, allowing each site to be its own historical control and to account for temporal trends.

Materials and Methods

Natural experiment: The smoking ban

In 2003, Pueblo citizens voted to implement the Smoke-Free Air Act within city limits. On July 1, 2003, the city of Pueblo, Colorado, adopted a strictly enforced citywide ordinance that prohibited smoking inside the workplace and all buildings open to the public, including restaurants, bars, bowling alleys, and other business establishments.¹³ Pueblo Law Enforcement strongly supported the ordinance and imposed significant fines on violators and on facility owners condoning smoking on their premises. Implementation and enforcement of the ordinance began on July 1, 2003, and continue to the present day.¹⁴

Comparison site

The City of El Paso was chosen as a comparison city in part because it had no smoking ban enacted before and after the experiment periods and because of its relative comparability to the experimental community at the overall population level. These similarities consisted of having a diverse ethnic population, comparable costs of living (e.g, transportation, healthcare, and home utilities), and analogous rates of violent crime.¹⁵ Additionally, both communities have been used in another observational study evaluating the effect of the Pueblo smoking ban on the incidence of myocardial infarction (MI).¹⁶ Although the cities of Colorado Springs and Pueblo are in adjacent counties, they are 45 miles apart, thus representing two completely isolated communities.

The experimental site—the City of Pueblo, Colorado—is a community located in the County of Pueblo in southeastern Colorado, U.S. Mountain West region. Based on 2003 data, Pueblo had a higher percentage of smokers compared to the statewide average (22.6% vs 18.6%).¹⁷ On July 1, 2003, Pueblo County had an estimated 147,751 residents, of whom about 70.2% lived within the city limits (n=103,648).¹⁶ El Paso county, the comparison site, includes the city of Colorado Springs, had an estimated population in 2003 of 550,478, with about 67.3% (n=370,448) residing in Colorado Springs. Unlike Pueblo County, El Paso County residents are employed in military activities and high technology jobs.¹⁶ In 2003, the smoking rate for Colorado Springs was 17.7%, which was slightly lower than Pueblo (22.6%) and state averages (18.6%).¹⁷

Data source

We applied for public access files from the State of Colorado Department of Health, Colorado Birth Registry, and the Infant Mortality Registry. We received a human subject review exemption from the Colorado Multiple Institutional Review Board because of the anonymity of data.

Cohort definitions

Using Health Department records, residence within Pueblo or El Paso County was ascertained using the mother's ZIP code, which was part of her record. Only ZIP codes that fell completely within city limits for both Pueblo and El Paso County were included. Only singleton birth records of babies born to mothers residing inside Pueblo city limits and in El Paso County were included in the cohort. Multiple births were excluded, as they may be a risk factor for both LBW and preterm births and also may be confounded with other maternal factors that we cannot measure.¹⁸ Because the data were de-identified, mothers may have more than one birth record in the cohort.

Main outcomes

The outcomes consisted of proportions of maternal smoking, preterm births, and LBW babies preban and postban. Maternal smoking was evaluated by the self-reported number of cigarettes the mother smoked per day during her lifetime, which was dichotomized ases or no. Preterm births were defined as an estimated gestational age of <37 weeks. The World Health Organization (WHO) designates infants weighing ≤2500 g as LBW; however, data from the Centers for Disease Control and Prevention (CDC) and others suggest poor health outcomes even among infants weighing ≤3000 g.^{19

–27} Therefore, two cutoff points for LBWs were used (e.g., <2500 g or <3000 g).

Covariates

Demographic variables consisted of the mother's mean age (inears), mother's age (categorical, <18ears, 18–34 years, and ≥35 years), white race (yes/no), Hispanic ethnicity (yes/no), high school diploma (yes/no), and married (yes/no). Self-reported health history of the mother and preexisting medical conditions consisted of self-reported smoking history (also used as an independent variable in some models), alcohol consumption (defined as any drinks per week and dichotomized as yes/no), any prenatal care visits (yes or no), previous birth (yes/no), previous preterm or small-for-gestational age (SGA) infant (yes/no), chronic hypertension (yes/no), and preexisting diabetes (yes/no). Gestational age was reported based on an early ultrasound examination and also was verified by the physician attending the birth. Self-reported pregnancy complications (yes/no) consisted of gestational diabetes mellitus (GDM), prenatal hypertension, vaginal bleeding, abruption, and placenta previa. Finally, birth characteristics of the fetus consisted of the male gender (yes/no), cesarean birth (yes/no), and weight of child (in grams). The maternal smoking model was adjusted using the maternal demographic and behavioral choice (e.g., smoking and alcohol consumption) covariates. All modeled covariates were categorical or dichotomous. Covariates were chosen a priori based on risk factors and biologic plausibility for LBW or preterm birth.^{8,19,28
–30} Analyses were performed with SAS version 9.2 (SAS Institute, Cary, NC).

Statistical analysis

A chi-square test was used to compare differences across sites for baseline (preban) categorical variables, and Student's t test was used to compare continuous variables. Characteristics were also compared before and after the ban, within site. Adjusted odds ratios (OR) were obtained using multivariable logistic regression analysis.³¹ Difference-in-difference estimation was used to account for site, time, and site by time, in addition to other covariates.

Results

During the time periods studied, a total of 6717 births were identified in the City of Pueblo and 32,293 births within El Paso County. Table 1 summarizes the baseline characteristics of the women sampled delivering singleton births in both Pueblo and El Paso Counties preban. Comparing mothers residing within Pueblo city limits with those in El Paso County preban, significant differences existed in mean age, race, ethnicity, high school education, alcohol consumption, marital status, and prenatal visits (p<0.05 for all variables). A significant difference was also noted in the following medical conditions: anemia, preexisting diabetes, chronic hypertension, GDM, pregnancy-associated hypertension, and vaginal bleeding (p<0.05). Specifically, mothers in Pueblo were more often Hispanic, only high school educated, and reported a prior birth of LBW or SGA infant compared to mothers in El Paso.

Table 1.

Characteristics for Sampled Women Delivering Singleton Births, Residing Within the City Limits of Pueblo and El Paso County During April 1, 2001, to July 1, 2003

	Inside Pueblo City limits	Inside El Paso County	p value ^a
Births	n=3,421	n=16,348
Mother's mean age, years	24.92	26.73	<0.0001(t test)
Age group
<18 years	229 (6.7)	550 (3.4)	<0.0001
18–34 years	3,421 (85.9)	13,880 (84.9)	0.1444
≥35 years	254 (7.4)	1,918 (11.7)	<0.0001
Mother's race
White	3,272 (95.6)	13,915 (85.1)	<0.0001
Mother's ethnicity
Hispanic	1,994 (58.3)	3,285 (20.1)	<0.0001
High school diploma	2,290 (67.3)	13,681 (84.0)	<0.0001
Alcohol consumption^b	74 (2.2)	90 (0.55)	<0.0001
Married	1,698 (49.6)	12,061 (73.8)	<0.0001
Previous births^c	2,051 (60.0)	9,579 (58.6)	0.1410
Previous preterm or small for gestational age birth	15 (0.44)	49 (0.30)	0.1939
Preexisting medical conditions
Anemia	69 (2.0)	128 (0.78)	<0.0001
Preexisting DM	21 (0.61)	55 (0.34)	0.0171
Chronic HTN	23 (0.67)	59 (0.36)	0.0100
Pregnancy complications
Gestational DM	86 (2.5)	292 (1.8)	0.0047
Pregnancy-related HTN	99 (2.9)	340 (2.1)	0.0033
Vaginal bleeding	4 (0.12)	63 (0.39)	0.0140
Placenta previa	10 (0.29)	42 (0.26)	0.7132
Abruption	23 (0.67)	92 (0.56)	0.4435
No prenatal care visits	39 (1.1)	173 (1.1)	0.6
Birth characteristics
Cesarean birth	622 (18.2)	2,935 (18.0)	0.7516
Male sex	1,747 (51.1)	8,472 (51.8)	0.4211

n (%) unless otherwise indicated.

Results from chi-square test of proportions unless otherwise indicated.

One or more drinks, ever.

One or more live or dead births.

DM, diabetes mellitus; HTN, hypertension; LBW, low birth weight.

Examining percentage of mothers who smoke by month in each location, smoking appeared to increase in El Paso County over time but remained the same or slightly decreased in Pueblo during the same time period (Fig. 1). Examining characteristics by location, preban to postban, the unadjusted percentage of smoking mothers decreased in Pueblo from 16.64% to 15.07%, although this was not significant (p=0.0789) (Table 2). Rates of smoking in El Paso County increased significantly from 8.66% to 11.89% (p<0.0001) during the same time period. When evaluating unadjusted frequency data for singleton mothers residing in the City of Pueblo, nonsignificant decreases from 8.51% to 7.92% and 34.23% to 33.45% were seen for singleton mothers giving birth to a baby <2500 g and <3000 g, respectively. Within El Paso County, 7.85% and 29.78 % of women gave birth to a baby weighing <2500 g and <3000 g, respectively, before the ban. These percentages increased to 8.34% and 32.02% after the ban, respectively. The increase in <3000 g babies was significant (p<0.0001). Unadjusted rates of preterm births did not change over time in Pueblo but increased from 7.93% to 9.23% in El Paso County (p<0.0001).

FIG. 1.

Monthly maternal smoking for sampled women delivering singleton births, residing within the city limits of Pueblo and El Paso County during April 1, 2001, to July 1, 2003, and April 1, 2004, to July 1, 2006.

Table 2.

Percentages of Smoking Mothers, Preterm Births, and Low Birth Weight Babies Preban and Postban, by Location

	Preban (%)	Postban (%)	p value
Smoking status of mother
Pueblo	568 (16.64)	492 (15.07)	0.0786
El Paso County	1,413 (8.66)	1,891 (11.89)	<0.0001
Preterm(<37 weeks)
Pueblo	270 (7.89)	245 (7.44)	0.4796
El Paso County	1,296 (7.93)	1,471 (9.23)	<0.0001
LBW(<2500 g)
Pueblo	291 (8.51)	261 (7.92)	0.3848
El Paso County	1,283 (7.85)	1,329 (8.34)	0.1090
LBW(<3000 g)
Pueblo	1,171 (34.23)	1,102 (33.45)	0.5023
El Paso County	4,867 (29.78)	5,105 (32.02)	<0.0001

Results of the multivariable logistic regression, including the difference-in-differences interaction term, were used to assess whether the odds of LBW births in each area were reduced by the smoking ban (Table 3). Unadjusted odds, as well as three adjusted models, are presented. Although the frequency of birth weights <2500 g changed in both areas, this effect was not significantly modified by the smoking ban. However, for change in frequency of the lower third of birth weights <3000 g, this effect was significantly modified by the ban (p<0.05). Once the model was adjusted for medical conditions and birth characteristics, however, a nonsignificant association among the location, ban, and LBW was found.

Table 3.

Odds of Low Birth Weight and Preterm Birth in Pueblo Postban

Models	Low birth weight
	<2500 g	<3000 g
	OR (95% CI)	OR (95% CI)
Unadjusted^a	0.867 (0.716 - 1.050)	0.870 (0.778 - 0.972)^*
Adjusted^b	1.044 (0.824 - 1.323)	0.916 (0.811 - 1.035)

	Preterm birth
	OR (95% CI)
Unadjusted^a	0.794 (0.653 - 0.966)^*
Adjusted^c	0.769 (0.599 - 0.987)^*

	Maternal smoking
	OR (95% CI)
Unadjusted^a	0.624 (0.537 - 0.726)^*
Adjusted^d	0.620 (0.529 - 0.727)^*

Reference group: El Paso County, preban.

p<0.05.

Modeling time, site, time^*site.

Adjusted for age, white race, Hispanic ethnicity, high school education, married, smoking, alcohol consumption, preterm birth, previous birth, hypertension, prenatal hypertension, gender of baby, cesarean birth, previous preterm or small-for-gestational age infant.

Adjusted for age (categorical; see Table 1, ref=age18–34), white race, Hispanic ethnicity, high school education, married, smoking status, alcohol consumption(y/n), previous birth, preexisting medical conditions (hypertension, diabetes, and anemia), pregnancy complications (prenatal hypertension, gestational diabetes, vaginal bleeding, abruption, placenta previa), gender of baby, cesarean birth, weight of child in grams, previous preterm or small-for-gestational age infant.

Adjusted for age, white race, Hispanic ethnicity, high school education, married, alcohol consumption.

CI, 95% confidence intervals; OR, odds ratio; y/n, yes/no.

For the outcome of preterm births, the unadjusted model resulted in a 21% and the adjusted model in a 23% reduction in the odds of preterm births associated with the smoking ban (p<0.05 for both estimates, respectively) (Table 3). The smoking ban was also associated with a significant 38% reduction in the odds of maternal smoking (p<0.05).

Discussion

This is the first U.S. study to evaluate the association between institution of a citywide smoking ban and maternal smoking, LBW, and preterm births. Using a difference-in-differences framework, the smoking ban was associated with a significant reduction in the odds of maternal smoking (38%) and preterm births (23%), although there was no significant decrease in new singleton LBW babies (p<0.05 for both outcomes). These mixed results for the association between a legislated smoking ban and individual health outcomes is consistent with two published report of a similar nature in Ireland. In a cross-sectional observational study, Kabir et al.³² evaluated singleton live births from Coombe University Maternal Hospital in the years 2003 (n=7,593) and 2005 (n=7,648), before and after a comprehensive Irish workplace smoking ban was introduced (March 2004). The investigators determined maternal smoking rates, mean birth weights, and adjusted ORs of LBW (defined as <2500 g) and preterm births (defined as <37 weeks gestation) before and after the smoking ban. After adjusting for maternal pregnancy and demographic factors and fetal characteristics, such as age, maternal smoking status, working status of the mother, gestational age (for LBW only), and weight of the child (for preterm births), the investigators found a 25% decreased risk of preterm births (OR 0.75, 95% confidence interval [CI] 0.59–0.96) and a 43% increased risk of LBW births (OR 1.43, 95% CI 1.10–1.85), with a 12% decline in maternal smoking rates. That study did not have a comparison site and, therefore, concluded that the significant change in LBW reflected temporal and secular trends in industrial nations rather than a result of the smoking ban.³³ Our study improved on that study design by including a comparison site, so we were able to account for secular trends, essentially allowing each site to be its own control.

One other study has addressed the impact of smoke-free legislation on maternal smoking rates. After implementation of smoke-free legislation in Italy that banned all smoking in enclosed spaces, Franchini et al.³⁴ evaluated active and passive maternal smoking exposure in 979 women at the end of their pregnancy through self-report (39–40 week of pregnancy) and fetal cord serum cotinine concentrations (obtained at the time of delivery). Serum cotinine concentrations are a validated biologic marker for cigarette smoke exposure.³⁴ Based on analysis of self-reports, 47.7% of women who smoked before their pregnancy quit smoking at the beginning of their pregnancy, which was confirmed by cord serum cotinine levels. Mothers who continued to smoke during their pregnancy reduced the number of cigarettes consumed from a mean of 10.7 to 5.1. When considering both self-reports and cord serum cotinine levels, the investigators found that between 54% and 79% of nonsmoking women, respectively, were exposed to cigarette smoke during gestation. These data suggest that smoke-free legislation may act as an effective strategy for deterring exposure to both firsthand and secondhand cigarette smoke in pregnant mothers and their newborns.

The implementation of smoking bans has been used as a public health measure to decrease adverse medical events for over 400 years. In 1604, King James I of England presented the manifesto “A Counterblaste to Tobacco,” characterizing the act of smoking tobacco as unhealthy and barbarous.³⁵As of 2011, 48 states and 2960 cities and counties in the United States currently enforce one or more forms of no-smoking ordinances.³⁶ Only 20 states currently exercise a statewide 100% smoke-free workplace, restaurant, and bar law.³⁶ Following smoking bans, significant reductions in MI and acute coronary events have been noted in multiple studies, across age groups.^9
–11,16 Smoke-free laws have also manifested respiratory benefits, as measured by lung function testing, and systemic inflammatory markers, when evaluated in bar workers preban and postban.³⁷ For example, Mackay et al.¹² found an 18.2% reduction in the mean rate of admissions for childhood asthma per year in all hospitals in Scotland after passage of the Smoking, Health, and Social Care Act, which banned all smoking in enclosed public places and workplaces (p<0.001).¹²

Although our findings suggest a positive effect on maternal smoking and preterm births, we did observe an increased trend in monthly smoking rates of mothers residing in El Paso County during the postban period that was not seen in Pueblo (Fig. 1). While the prevalence of smoking in Colorado dropped from 20% to 17.9% during this period, national smoking prevalence in specific ethnic subgroups actually increased.³⁸ Our study did not aim to explain this trend, but we did search for existing explanations. Data from the National Center for Health Statistics (NCHS) from the CDC suggest that the percentage of Hispanic adults who smoked in 2003 rose from 3.0% to 3.4% in 2004 and again to 4.3% in 2005 but then declined to 2.9% in 2006.³⁸ For African Americans, percentages increased from 4.9% in 2003 to 5.3% in 2004 and 2005, with a decline to 4.9% in 2006. For whites, these percentages remained constant from 2003 to 2005 (e.g., 14.2%–14.9%) but declined to 13.6% in 2006.³⁸ As a high percentage of Hispanic adults reside in El Paso County (20%), this could potentially explain the increase in smoking trends of mothers.

Limitations

Our study has several limitations associated with retrospective, observational data and the ecologic study design. First, we could not measure mothers' exposure to secondhand smoke directly, as our data source answered queries only about their own smoking behavior. Thus, variation in exposure to secondhand smoke was an unmeasured variable in our birth certificate data source. Nevertheless, our ecologic study examined population-level exposure to a ban and a change in outcomes in the population, which fundamentally is the question public health interventionist want to answer. Our study asked: Despite incomplete individual-level data, can we see a positive effect of a smoking ban at the population level?

The Colorado Birth Data Registry data source did not include paternal smoking history, which could underestimate possible secondhand smoke exposure, nor were they asked about smoking in the home. Maternal self-report is also likely underreported, as there is social stigma associated with smoking during pregnancy.^39,40 Because of likely underreporting of smoking behavior, we chose not to stratify the sample of mothers by smoking history. With more opportunities to triangulate measures of smokers and nonsmokers, this would have been a way to assess the effect of the smoking ban via secondhand smoke exposure among nonsmokers. Additionally, because changes were analyzed at the population level, we could not discriminate differing effects in smokers vs. nonsmokers. It is likely that a smoking ban would have the greatest effect on nonsmokers, as the level of secondhand smoke would be reduced, whereas the effect on smokers or mothers who live with a smoker would not be as great. Although these are two important populations who likely differ but could not be analyzed separately, our results may be biased toward the null. Thus, the greater benefit to nonsmokers may be obscured by a lesser benefit in smokers.

A similar limitation is that mothers reported lifetime smoking, not in the last year or something more proximal to the pregnancy. Information in regard to trimester of pregnancy when smoking exposure occurred was also unavailable. In general, tobacco exposure has shown the greatest effect in the last trimester of pregnancy, the period of highest fetal growth.⁴¹ In terms of possible contagion in the design, a mother theoretically could live inside Pueblo city limits and cross outside Pueblo for employment, where she would effectively have reduced exposure to the smoking ban.

When evaluating the unadjusted percentages of maternal smoking and fetal outcomes, no statistically significant difference was seen before and after the smoking ban for Pueblo cohorts, while a trend toward of reduction in each outcome was evident. These findings may be because of the lower number of women in the Pueblo cohort, which could have impacted the power of the analysis.

As the authors of the Irish study conceded, significant reductions in maternal smoking rates do not always correlate with a reduction in LBW prevalence.³³ Unfortunately, we cannot uncover the specific mechanism of action from these data. Furthermore, this particular intervention in our study may represent a relatively distal secondhand smoke exposure for pregnant women and their fetuses, as the reduction in exposure by smoking prohibition at bars, restaurants, worksites, and public places may not represent a significant change in exposure overall for this population. Future evaluations could aim to test these attributes of smoking and fetal outcomes more directly by surveying mothers about exposure to smoke at home, history of personal smoking behaviors (past substance use having a lower threshold of social desirability than current substance use), and any other protective behaviors during pregnancy (avoiding bars or smoky restaurants or trying to quit smoking, for example).

Conclusions

Changes in the risk of preterm births by enforcement of smoke-free ordinances have not been reported previously in the U.S. literature, and our data indicate an important new area for larger investigations to empirically evaluate claims of medical benefits of smoking bans. Smoking, an important risk factor for cardiac and pulmonary disease, also has been identified as a modifiable and significant modulator of fetal and placental health.^19,28,29,42 Public tobacco restrictions, particularly those that are full bans with no exemptions, have demonstrated multiple downstream effects, such as augmentation of smoking cessation attempts, fewer relapses in quitters, and fewer cigarettes smoked per day among populations within their respective communities.^43,44 Although individual effects of these outcomes may be difficult to tease out, these restrictions may serve as important mediators of disease reduction across multiple diseases and age groups on a communitywide level.

This is the first U.S. study with data indicating that pregnant mothers and their fetuses represent an important population to survey further in terms of health and cost effects of smoke-free ordinances, even when purported exposure seems deceptively low. Future studies are needed to evaluate the effects of smoke-free ordinances among various maternal populations based on social, demographic, and medical factors in order to determine when and where such measures potentially will exert their greatest benefits. Future work should also assess a dose-response comparing smoking bans and their enforcement and the longer-term effect of these policies.

Footnotes

Acknowledgments

We acknowledge the Colorado Department of Public Health and Environment Division of Health Statistics; Dennis Lezotte, Ph.D., and Jill Norris, Ph.D., Colorado School of Public Health, for their review of the study design of this article in the doctoral program in Health Services Research; and participants at the annual meeting of the American Public Health Association in Denver, Colorado, for their comments. This study had no funding source.

Disclosure Statement

The authors declare that they have no competing interests.

References

World Health Organization. Report on the global tobacco epidemic, 2009: Implementing smoke-free environments. www.who.int/tobacco/mpower/en/index.html. 2011 September 27.

Misra

, Nguyen

. Environmental tobacco smoke and low birth weight: A hazard in the workplace? Environ Health Perspect, 1999; 107,Suppl 6:897–904.

Health effects of exposure to secondhand smoke. Washington, DC: Environmental Protection Agency www.epa.gov/smokefree/healtheffects.html. 2011 September 27.

Ashford

, Hahn

, Hall

, Rayens

, Noland

, Ferguson

. The effects of prenatal secondhand smoke exposure on preterm birth and neonatal outcomes. JOGNN, 2010; 39:525–535.

Walsh

. Effects of maternal smoking on adverse pregnancy outcomes: Examination of the criteria of causation. Hum Biol, 1994; 66:1059–1092.

Barker

. Fetal origins of coronary heart disease. BMJ, 1995; 311:171–174.

Callinan

, Clarke

, Doherty

, Kelleher

. Legislative smoking bans for reducing secondhand smoke exposure, smoking prevalence and tobacco consumption. Cochrane Database Syst Rev, 4CD005992.

Institute of Medicine. Preterm birth: Causes, consequences, and prevention. Washington, DC www.iom.edu/Reports/2006/Preterm-Birth-Causes-Consequences-and-Prevention.aspx. 2011 September 27.

Dove

MSM

, Dockery

DWS

, Mittleman

MAM

et al. The impact of Massachusetts' Smoke-Free Workplace Laws on acute myocardial infarction deaths. Am J Public Health, 2010; 100:2206–2212.

10.

Herman

, Walsh

. Hospital admissions for acute myocardial infarction, angina, stroke, and asthma after implementation of Arizona's comprehensive statewide smoking ban. Am J Public Health, 2011; 101:491–496.

11.

Meyers

, Neuberger

, He

. Cardiovascular effect of bans on smoking in public places: A systematic review and meta-analysis. J Am Coll Cardiol, 2009; 54:1249–1255.

12.

Mackay

, Haw

, Ayres

, Fischbacher

, Pell

. Smoke-free legislation and hospitalizations for childhood asthma. N Engl J Med, 2010; 363:1139–1145.

13.

Amos

. Pueblo, Colo., smoking ban meets with compliance, defiance. Pueblo Chieftain., January 2003.

14.

Nevin-Woods

, Bartecchi

, Alsever

, Beauvais

. The Pueblo Heart Study: Pueblo, CO www.no-smoke.org/pdf/PuebloHeartStudyInformation.pdf. 2011 December 15.

15.

Harvard School of Public Health. Diveristydata.org. Cambridge, MA diversitydata.sph.harvard.edu/. 2011 September 25.

16.

Bartecchi

, Alsever

, Nevin-Woods

et al. Reduction in the incidence of acute myocardial infarction associated with a citywide smoking ordinance. Circulation, 2006; 114:1490–1496.

17.

Shupe

, Clancy

, Holguin

, Rosenblatt

. Tobacco use in Colorado: Results from the Behavioral Risk Factor Surveillance System. http://www.chd.dphe.state.co.us/Resources/pubs/tobacco3c.pdf. 2012 February 28.

18.

Blondel

, Kogan

, Alexander

et al. The impact of the increasing number of multiple births on the rates of preterm birth and low birthweight: An international study. Am J Public Health, 2002; 92:1323–1330.

19.

Valero De Bernabe

, Soriano

, Albaladejo

et al. Risk factors for low birth weight: A review. Eur J Obstet Gynecol Reprod Biol, 2004; 116:3–15.

20.

Centers for Disease Control and Prevention. National birth weight-specific infant mortality surveillance: Preliminary analysis—United States, 1980. MMWR, 1986; 35:269–273.

21.

Wei

, Li

, Sung

et al. Birth weight correlates differently with cardiovascular risk factors in youth. Obesity, 2007; 15:1609–1616.

22.

Pettitt

, Jovanovic

. Low birth weight as a risk factor for gestational diabetes, diabetes, and impaired glucose tolerance during pregnancy. Diabetes Care, 2007; 30,Suppl 2:S147–149.

23.

Ahmed

, Jaakkola

. Maternal occupation and adverse pregnancy outcomes: A Finnish population-based study. Occup Med, 2007; 57:417–423.

24.

Carlsson

, Persson

, Alvarsson

et al. Low birth weight, family history of diabetes, and glucose intolerance in Swedish middle-aged men. Diabetes Care, 1999; 22:1043–1047.

25.

Haas

, Goldberg

, Ohye

, Mosca

, Bove

. Primary repair of aortic arch obstruction with ventricular septal defect in preterm and low birth weight infants. Eur J Cardiothorac Surg, 2000; 17:643–647.

26.

Rode

, Hegaard

, Kjaergaard

, Moller

, Tabor

, Ottesen

. Association between maternal weight gain and birth weight. Obstet Gynecol, 2007; 109:1309–1315.

27.

Simcox

, Jaakkola

. Does work as a nurse increase the risk of adverse pregnancy outcomes? J Occup Environ Med, 2008; 50:590–592.

28.

Ashdown-Lambert

. A review of low birth weight: Predictors, precursors and morbidity outcomes. J R Soc Health, 2005; 125:76–83.

29.

Williams

, Davenport

, Sterne

, Sivapathasundaram

, Fearne

, Curtis

. Mechanisms of risk in preterm low-birthweight infants. Periodontology 2000, 2000; 23:142–50.

30.

Goldenberg

, Culhane

, Iams

, Romero

. Epidemiology and causes of preterm birth. Lancet, 2008; 371:75–84.

31.

Kleinbaum

, Kupper

, Muller

, Nizam

. Applied regression analysis and multivariable methods, 4th. Pacific Grove, CA: Duxbury, 2008.

32.

Kabir

, Clarke

, Conroy

, McNamee

, Daly

, Clancy

. Low birthweight and preterm birth rates 1 year before and after the Irish workplace smoking ban. Br J Obstet Gynaecol, 2009; 116:1782–1787.

33.

Kabir

, Connolly

, Clancy

, Cohen

, Koh

. Declining maternal smoking prevalence did not change low birthweight prevalence in Massachusetts from 1989 to 2004. Eur J Public Health, 2009; 19:65–68.

34.

Franchini

, Caruso

, Perico

et al. Assessment of foetal exposure to cigarette smoke after recent implementations of smoke-free policy in Italy. Acta Paediatr, 2008; 97:546–550.

35.

Eckert

. James I of England: Counterblaste to tobacco (1604) Cancer Bull, 1964; 16:52–54.

36.

Americans for Nonsmokers Rights. Summary of 100% smoke free state laws and population protected by 100% U.S. smokefree laws. Berkley, CA www.no-smoke.org/goingsmokefree.php?id=519. 2011 December 15.

37.

Menzies

, Nair

, Williamson

et al. Respiratory symptoms, pulmonary function, and markers of inflammation among bar workers before and after a legislative ban on smoking in public places. JAMA, 2006; 296:1742–1748.

38.

American Lung Association. Research and Program Services Epidemiology and Statistics Unit. Trends in tobacco use, 2011

Washington, DC

www.lungusa.org/finding-cures/our-research/trend-reports/Tobacco-Trend-Report.pdf. 2011 December 15.

39.

Ford

, Tappin

, Schluter

, Wild

. Smoking during pregnancy: How reliable are maternal self-reports in New Zealand? J Epidemiol Community Health, 1997; 51:246–251.

40.

Klebanoff

, Levine

, Morris

et al. Accuracy of self-reported cigarette smoking among pregnant women in the 1990s. Paediatr Perinat Epidemiol, 2001; 15:140–143.

41.

Reeves

, Bernstein

. Effects of maternal tobacco-smoke exposure on fetal growth and neonatal size. Expert Rev Obstet Gynecol, 2008; 3:719–730.

42.

Andres

, Day

. Perinatal complications associated with maternal tobacco use. Semin Neonatol, 2000; 5:231–241.

43.

Nagelhout

, Willemsen

, de Vries

. The population impact of smoke-free workplace and hospitality industry legislation on smoking behaviour. Findings from a national population survey. Addiction, 2011; 106:816–823.

44.

Naiman

, Glazier

, Moineddin

. Is there an impact of public smoking bans on self-reported smoking status and exposure to secondhand smoke? BMC Public Health, 2011; 11:146.