Smoking Cessation and Changes in Body Mass Index Among Middle Aged and Older Adults

Abstract

Objective: This study contributes to the body of literature examining smoking cessation and body mass index (BMI) for adults aged 50 and older. Method: Utilizing the 2004 and 2010 waves of the RAND Health and Retirement Study, this analysis utilized Fixed Effects (FE) regression on a sample of 1,316 adults aged 50 and older. Results: Older adults undergo a small change in BMI after a transition from smoking to nonsmoking during a 6-year period, and this occurs after accounting for individual-level unobserved heterogeneity. More specifically, men experience a BMI gain of 1.24 (p< .01) and women experience a BMI gain of 1.58 (p< .01). Discussion: Gerontologists/health professionals can use these results to inform older adults about the potential for a small increase in BMI and, in the process, assuage any apprehensions about excessive weight gain. This insight may encourage a greater number of older adults to cease smoking.

Keywords

older adult smoking smoking cessation body mass index smoking weight

Introduction

Background

Smoking remains a key public health issue because many individuals continue to smoke every day despite the detrimental impact of tobacco use on health, excess medical costs, and economic/productivity loss (Miller, Ernst, & Collin, 1999; U.S. Department of Health and Human Services [HHS], 2014; Warner, Hodgson, & Carroll, 1999; Weng, Ali, & Leonardi-Bee, 2013; Xu, Bishop, Kennedy, Simpson, & Pechacek, 2014). As early as 1964, the Surgeon General’s Report linked cigarette use with lung cancer and heart disease (Terry, 1973; United States, 1964). Another precursor is the World Health Organization Chronicle from 1970, which put forth that cigarette use plays a major role in ischemic heart disease, lung cancer, chronic bronchitis, emphysema, and disability, findings supported by subsequent Surgeon General’s Reports during the past 50 years. (Fletcher & Horn, 1970; U.S. Department of HHS, 1982, 1984, 1988, 2004, 2006, 2010, 2014). Given this widespread and long-standing information about the consequences of tobacco use, smokers would be quick to cease consumption. However, smokers face challenges to cessation, and one unique hurdle is apprehension surrounding weight gain (Clark et al., 2006; Luostarinen et al., 2013; Orleans, Jepson, Resch, & Rimer, 1994; Orleans, Rimer, Cristinzio, Keintz, & Fleisher, 1991; Treviño et al., 2014; Veldheer et al., 2014). The present study examined whether such a gain occurs after smoking cessation and if so, how much? This is a noteworthy contribution because some studies have found minimal weight gain, while others have found in excess of 20 pounds (Aubin et al., 2012; Bush, 2014; Kasteridis & Yen, 2012; Veldheer, 2015). In addition, this research applied data specifically for adults aged 50 and older, which is an important consideration given the vast majority of smoking literature focuses either on (a) adolescents or (b) the general adult population.

Importance of Smoking for Older Adults

Smoking is of particular concern among older adults. Ten-year trends from the National Health Interview Survey (NHIS) show no significant changes in the prevalence of smoking from 2002 to 2011 for older adults (i.e., aged 65+) with estimates of 9.3% and 7.9%, respectively (Agaku, King, & Dube, 2014; Dube & Wu, 2015). Equally worrisome, estimates of the prevalence of daily cigarette use show no statistically significant changes for older adults with 8.5% in 2002 and 6.6% in 2011. These estimates are consistent with other surveys. For example, Garrett, Dube, Winder, and Caraballo (2013) utilized the National Survey on Drug Use and Health to estimate smoking prevalence for the U.S. population from 2006 to 2008 and 2009 to 2010. Their analysis revealed that 9.4% of older adults smoked during the first time period and 9.2% smoked during the second time period, suggesting almost no change. A longer time horizon paints the same picture for all adults—NHIS data show 10% of women smoked cigarettes in 1965, and this reduced slightly to 9% in 2010 (Federal Interagency Forum, 2012). Rates for men have decreased, but the overall percentage still remains high at 10% for 2010.

The lack of a statistically significant decrease in prevalence rates despite numerous education and outreach efforts for smoking cessation is worrisome and particularly so for older adults. These individuals face greater risks for disease and mortality due to a longer exposure period (W. S. Choi, Harris, Okuyemi, & Ahluwalia, 2003; Freedman, Nelson, & Freedman, 2012; Gilpin, Choi, Berry, & Pierce, 1999; Reidpath, Davey, Kadirvelu, Soyiri, & Allotey, 2014; Trinidad, Gilpin, Lee, & Pierce, 2004; Unger & Chen, 1999; U.S. Department of HHS, 2014). Gellert, Schöttker, and Brenner’s (2012) systematic review and meta-analysis using MEDLINE, EMBASE, and ISI Web of Knowledge place this in perspective. These scholars focus exclusively on all-cause mortality for older adults (i.e., at least 60 years of age) from smoking. Their analysis of 17 studies from seven countries from 1987 to 2011 revealed 83% increased mortality for smokers, and such a robust estimate was applicable for the oldest age groups, as well as across gender and geography. For the United States, current smokers were estimated to have a relative mortality of 1.87, and smoking is the leading cause in nearly one-in-five deaths (approximately 480,000 deaths annually) according to recent estimates from the U.S. Department of HHS (2014).

Smoking and Weight Gain

Given numerous comorbidities associated with tobacco, smoking cessation is a positive initiative for improving health. However, individuals express concerns about weight gain from quitting (Jeffery, Hennrikus, Lando, Murray, & Liu, 2000; Meyers et al., 1997; Perkins, Levine, Marcus, & Shiffman, 1997; Veldheer, 2014; Williamson et al., 1991). According to a national mail survey conducted with American Association of Retired Persons (AARP) members, nearly 50% of the 6,000 respondents aged 50 to 74 years listed weight gain as a barrier (Orleans et al., 1991). This commonly cited obstacle was more problematic for heavy smokers (i.e., more than 25 cigarettes/day). Another national survey using the Adult Use of Tobacco data for adults aged 50 to 74 years revealed weight gain as an important factor for a return to smoking, second only to irritability (Orleans et al., 1994). A more recent study utilizing a randomized trial found 50% of females and 26% of males expressed weight concerns from quitting (Clark et al., 2006). Scholars have long recognized such a concern as a potential barrier, but what is less understood is how much weight gain occurs as a direct result from cessation and what is the outcome for older adults. For example, an early comprehensive literature review examined (a) the relationship between smoking and body weight, (b) possible mechanisms for weight gain, and (c) ways to maintain weight gain utilizing 29 cross-sectional studies from 1971 to 1987 and 41 prospective studies from 1970 to 1989. With a sample size of more than 350,000, these scholars concluded smoking cessation was associated with weight gain, but the mechanisms underlying such a gain are not well established (Klesges, Meyers, Klesges, & LaVasque, 1989). This is supported by Flegal, Troiano, Pamuk, Kuczmarski, and Campbell’s (1995) analysis of the third wave of the National Health and Nutritional Examination Survey, which found a gain of nearly 12 pounds from smoking cessation over a 10-year period for adults aged 35 to 70+ years. However, Klesges et al. (1997) caution some studies suggest only a modest weight gain of four pounds, while others have found as much as 18 pounds (also see Veldheer, 2015).

As for recent studies, Bush et al. (2014) surveyed quitline participants in five states from August to December 2010 with baseline as well as 3- and 6-month follow-up weight gathered from telephone questionnaires. Their analyses revealed no consistent patterns for weight gain and such an outcome also held for (a) treatment engagement and (b) treatment effectiveness. However, these scholars made a point to cite Kasteridis and Yen’s (2012) findings, which suggested, “quitting smoking resulted in minimal weight gain but the magnitude of weight gain varied by age and gender” (p. 214). A large-scale studying examining smoking cessation and weight gain utilizing a sample of 300,000 volunteers from 1992 to 2000 found quitters gained an average of less than one pound annually and nearly six pounds over a 5-year period (Travier et al., 2012). These studies contrast with Aubin et al.’s (2012) meta-analysis of 62 studies using a random effects (RE) inverse variance model. These scholars found statistically significant weight gain for both untreated and treated quitters at 1-, 2-, 3-, 6-, and 12-month follow-ups. Aubin et al. concluded smoking cessation was associated with a weight gain of approximately 11 pounds, and most of this gain occurred within 3 months. In addition, they also found variation and noted some quitters gained over 22 pounds.

Although some of the previous literature found a statistically significant and positive association for changes in weight after smoking cessation, the aforementioned studies raise interesting questions regarding the extent of weight gained. This study attempts to reconcile this uncertainty for middle aged and older adults. More specifically,

A: the amount of increase in body mass index due to a transition from smoking to nonsmoking.

B: the amount of increase in body mass index by gender.

These objectives are important to test because (a) a small number of studies have utilized the advantages of panel data in explicating the effects of smoking cessation on body mass index (BMI) and weight while accounting for economic and social characteristics that have become salient with older adults. More specifically, the increased number of older adults living in a divorced or widowed state and a greater percentage of older adults with limited financial resources due to the 2007-2009 recession (Burgard, Ailshire, & Kalousova, 2013; Fenge et al., 2012; Lin & Brown, 2012). Testing this hypothesis is also noteworthy because (b) some studies have concluded minimal weight gain, while others have found substantial differences. As such, understanding the amount of change can provide gerontologists/health professionals useful information in tailoring weight management programs for groups at higher risk.

Method

Conceptual Framework

In terms of an underlying framework, this article applies a modified version of the Health Utility Model (Muurinen & Le Grand, 1985). More formally, this study posits individuals undertaking a health-related behavioral change (i.e., smoking cessation) attempt to optimize their utility function, as depicted below (Case & Deaton, 2005; Grossman, 1972a, 1972b, 2000; Muurinen, 1982; Muurinen & Le Grand, 1985):

U = \sum_{0}^{T} {(1 + ρ)}^{t} υ (c_{t}, H_{t})

H_{t + 1} = Θ m_{t} + (1 - δ_{t}) H_{t}

This utility function, U, shows an individual’s time preference ( $ρ$ ), consumption ( $c_{t}$ ), and stock of health ( $H_{t}$ ). The individual attempts to maximize this over his or her lifetime or T. By altering health behavior (changing $m_{t}$ in the current time period with Θ as the rate of efficiency), the individual is able to decrease/increase the stock of health in the future time period or H_t+₁. If the behavior is detrimental (e.g., excessive drinking, drug abuse, smoking, etc.), then the stock of health will be reduced in future time periods. However, if the behavior is salubrious (e.g., exercise, better nutrition, smoking cessation, etc.), then the stock of health will be increased or, at the very least, not decrease significantly in future time periods. Health is understood to decrease or depreciate at the rate given by δ_t. This study posits that since individuals incur significant monetary and nonmonetary costs (i.e., time, effort, disruption of a lifestyle) when engaging in health behavior changes, they will attempt to maximize the perceived/derived benefits from such a change. Some studies provide support for this. For example, a noteworthy study using the first four waves of the Health and Retirement Study (HRS), which is the same data set for this study, found forward-looking models for addictive behaviors better predicted the data for adults aged 50 and older and prediction improved when the sample was restricted to adults aged 70 and older (Arcidiacono, Sieg, & Sloan, 2007). In the context of this study, the perceived/derived benefit is not only the overall improvement in health from smoking cessation but also the objective of minimizing weight gain. That is, older adults behave in a manner that is consistent with the expectation that gains in health achieved from smoking cessation will be offset if negative behaviors are undertaken (excessive drinking, eating, etc.) This framework stems from the Health Belief Model and the Demand for Health (Becker, 1974; Grossman, 1972a, 1972b, 2000; Janz & Becker 1984; Rosenstock, 1974). For older adults, some studies provide support for the applicability of this general framework (N. G. Choi & DiNitto, 2015; Rosenstock, 1990/2008; Strecher, DeVellis, Becker, & Rosenstock, 1986). Utilizing an extension of this model, the present study examines the following:

Hypothesis 1: After accounting for individual-level unobserved heterogeneity, there is a small increase in BMI due to a transition from smoking to nonsmoking.

Hypothesis 2: The small increase in BMI occurs for both older men and older women.

Data

This study utilized multiple waves of the RAND HRS. Supported by the National Institute on Aging (NIA U01AG009740) and the Social Security Administration, HRS is conducted by the Institute for Social Research at the University of Michigan and explores labor force participation and health transitions for individuals pre- and post-retirement (Karp, 2007). This widely used data set is a longitudinal panel study of more than 26,000 older individuals in the United States (RAND, Version L, Clair et al., 2011).

Analytic Sample

This research utilized the 2004 and 2010 panel (i.e., Waves 7 and 10, respectively). The following surveys/cohorts were included: HRS, War Babies, and Early Baby Boomers. As the focus is on older adults, individuals younger than 50 years of age in 2004 were excluded. The 2004 sample was selected because it captures some of the earliest Baby Boomers (i.e., those born in mid-1940s). The 2010 sample was selected because a 6-year time horizon can capture long-term changes in BMI.

The initial data set of smokers only and smokers who transition contains 3,044 observations. Next, respondent identification numbers are examined to ensure a balanced panel consisting of the same individuals for the two time periods. This produces a working data set with 2,982 individuals (1,491 for 2004 and 1,491 for 2010). However, not all observations have complete data for the independent variables of interest. Once this is accounted, the final data set contains 2,632 observations or 1,316 observations for each year. Categorization by smoking status shows 379 transition to nonsmoking, while 937 remain as smokers for the total sample in 2010. In terms of gender, 196 women transition while 530 remain as smokers, and 183 men transition while 407 remain as smokers.

Dependent Variable

The dependent variable is BMI. It is a continuous variable and calculated by the formula: $\frac{weight in kilograms}{height in meter s^{2}}$ (Panel NHLBI 1998). BMI is frequently used in health studies because it is convenient, comparable, and cost-effective. Although public health scholars have expressed concerns regarding its use as a measure of adiposity, this reservation is not as problematic for this study (Freedman, Thornton, et al., 2012; Shah & Braverman, 2012). For one, this research does not evaluate changes in fat composition from smoking cessation. Rather, the focus is changes in BMI and weight. Second, the goal of this study is to allow comparisons between past and current reports (i.e., BMI is frequently reported). Finally, this measure is readily available in a large number of data sets should one want to extend this study.

Time-Variant Independent Variables

The primary independent variable corresponds to nonsmoking status, a dummy variable for those who transition out of smoking from 2004 to 2010. Three important clarifications were made: (a) Nonsmoking represents individuals who transition from smoking in 2004 and remain as nonsmokers in 2010, (b) continuously smoking is the reference group, and (c) smoking status is self-reported with the following questions: “Have you ever smoked cigarettes?” and “Do you smoke cigarettes now?” (Clair et al., 2011). Self-reporting can be a major limitation. In some instances, an individual may report discontinued use when such is not the case. Although previous research shows moderate to high levels of concordance between self-reporting and cotinine verification, underreporting has the potential to bias the estimator (Caraballo, Giovino, Pechacek, & Mowery, 2001; Gorber, Schofield-Hurwitz, Hardt, Levasseur, & Tremblay, 2009; Patrick et al., 1994). Age is a continuous variable, and because its effect can be nonlinear, it is reformulated into four splines: 50-59 for Age 1, 60-69 for Age 2, 70-79 for Age 3, and 80+ for Age 4. As this variable is not a dummy, all categories are retained. Given that labor/employment considerations play a significant role in influencing health and health-related behaviors, this grouping includes employment status, value of bank accounts, supplemental security income (SSI), and earnings. Because this analysis uses the 2004 and 2010 waves, all dollar-value variables are indexed to 2010, and this dollar-value reflects the U.S. currency. The monetary values are also log transformed.

As mentioned in the “Introduction” section, this research also accounts for health attributes. This grouping consists of doctor visits, government health insurance, long-term care insurance, any walking difficulties, number of drinks consumed, and feelings of happiness. While most of these variables are self-explanatory, feelings of happiness should be clarified. This variable asks the respondent to answer as yes/no if such an emotion was present much of the time during the week before the interview. Government health insurance is set to equal one for those receiving TRI-CARE or CHAMPUS and zero otherwise. Long-term care insurance is also a dummy variable, and it accounts for older adults who maintain insurance for accidents, illnesses, or aging-related conditions requiring care.

Time-Invariant Independent Variables

Variables that do not vary across time are dropped from the FE model. Whereas such variables are retained in RE, this is a limitation in using FE. For this analysis, time-invariant variables include education, gender, never married, race, and veteran status. Because there is no change in such variables from 2004 to 2010 for this sample of older adults, estimates cannot be calculated under FE. However, studies have found variation in BMI changes by gender from smoking cessation. To explicate gender effects, separate models are estimated for men and women.

Statistical Analysis

The econometric framework utilizes advantages inherent with the panel nature of HRS. Although panel data are more complex to manage due to repeated observations, its structure provides considerable advantages, such as reducing problems due to unobserved heterogeneity, increasing efficiency, and uncovering important relationships that vary over time (Baltagi, 2009). To ensure a balanced panel, only individuals alive in 2004 and 2010 are included. This forms the basis for the FE framework:

y_{it} = α_{i} + β_{it} + γ Z_{it} + δ C_{i} + ε_{it}, where ε_{it} = μ_{i} | V_{it}

where x denotes the primary variable of interest, z represents time-varying controls, c signifies time-invariant controls, and ε represents the error term. The error term also contains the time-invariant and time-variant elements. As this analysis only examines 2004 and 2010 (i.e., two time periods), estimates from the FE framework are identical to the first-difference (FD) framework (Wooldridge, 2010). The model can be resummarized as follows:

Δ y_{i} = α + βΔ X_{i} + Δ μ_{i} where Δ μ_{i} = Δ ε_{i}

Now, Δx_i signifies a change from smoking to nonsmoking for individual i, while β represents the estimate on BMI from smoking cessation, as compared with individuals who remain as smokers from 2004 to 2010.

Specification Test

To determine whether FE or RE needs to be used, one should conduct a Hausman test. More specifically, FE assumes E(ε_it|α_i,x_it) = 0 while RE posits the stronger assumption of E(ε_it|α_i,x_it) = E(ε_it| x_it). For the Hausman test, the Null states individual effects are random and the preferred model is RE, while the Alternative Hypothesis states these effects are systematically different and the preferred model is FE. Results suggest the Null can be rejected in favor of the Alternative and FE is the appropriate model choice (χ² = 74.30, p = .001). For nontechnical readers, this test is important because improper model specification can result in inconsistent estimators.

Sensitivity Analyses

Given the possibility of extreme values for BMI, this study conducts two types of sensitivity analyses on the lower and upper tails of the distribution. The first is known as the Winsor process and it is applied to 2.5% of the extreme sample. These observations are not removed. Instead, BMI values at the tails are replaced with data at the 2.5 percentile and 97.5 percentile, respectively. This provides more robust estimators for the FE analysis. The second is known as the Grubbs test and it is conducted by removing extreme values utilizing an iterative process (Grubbs, 1969). This exclusion of outlier observations for BMI can be acceptable given such values are typically indicative of a subpopulation (i.e., extreme obesity). For this analysis, the Grubbs test did not identify any outliers for BMI.

Results

Descriptive Analysis

In terms of a transition, 379 individuals cease smoking while 937 remain as smokers (Table 1). Relative to nonsmokers, smokers are a few years younger (63 vs. 68) and a larger percentage is employed (34% vs. 20%). Smokers also have greater earnings and real estate valuations. However, nonsmokers maintain greater bank account valuations and slightly larger Supplementary Security Income. Not surprising given nonsmokers are older, this group has a much higher percentage receiving Medicare (69% vs. 47%). A greater percentage of nonsmokers also suffer from walking difficulty (42% vs. 31%) and maintain more frequent doctor visits (12 vs. 9). There are minimal differences for marital status. As for BMI, smokers approximate 26.67, while nonsmokers equal 27.81.

Table 1.

Mean Values for Smokers/Nonsmokers for Total Sample (2010).

	Smoker		Nonsmoker		Total
	M	SD	M	SD	M	SD
Nonsmoker	0.00	0.00	1.00	0.00	0.14	0.35
BMI	26.67	5.12	27.81	6.06	26.84	5.29
Age	63.19	7.22	67.59	6.66	63.83	7.31
Married	59%		60%		59%
Divorced	24%		22%		24%
Widowed	17%		18%		17%
Family size	2.26	1.21	2.23	1.21	2.25	1.21
Real estate	US$22,135.70	US$142,086.69	US$16,544.85	US$78,636.07	US$21,330.64	US$134,806.11
Accounts	US$21,884.00	US$100,021.25	US$26,957.99	US$71,292.39	US$22,614.64	US$96,418.47
Employed	34%		20%		32%
Supplementary income	US$797.90	US$3,094.70	US$821.93	US$3,298.80	US$801.36	US$3,124.27
Earnings	US$14,466.29	US$26,695.62	US$10,918.04	US$25,184.74	US$13,955.35	US$26,508.04
Medicare	47%		69%		50%
Government health	53%		72%		56%
LTC insurance	8%		10%		8%
Drinks	1.08		0.79		1.04
Walking difficulty	31%		42%		32%
Was happy	83%		86%		83%
No. of doctor visits	9.25	16.69	12.21	24.87	9.68	18.12
n	937		379		1,316

Note. BMI = body mass index; LTC = Long-term care.

For women, 196 cease smoking, while 530 remain as smokers (Table 2). As compared with nonsmokers, women smokers are younger by 5 years, greater number are still married, and much higher percentage is employed, which also reflects over US$3,500 in earnings. Meanwhile, nonsmokers maintain higher valuations for bank accounts and real estate. Nonsmokers also have a greater percentage receiving Medicare and also have more doctor visits. As for BMI, smokers approximate 26.50 while nonsmokers estimate 27.56.

Table 2.

Mean Values for Smokers/Nonsmokers by Gender (2010).

	Women			Men
	Smoker	Nonsmoker	Total	Smoker	Nonsmoker	Total
	M	M	M	M	M	M
Nonsmoker	0.00	1.00	0.13	0.00	1.00	0.16
BMI	26.50	27.56	26.64	26.89	28.47	27.14
Age	62.98	67.64	63.60	63.47	67.55	64.11
Married	50%	45%	50%	71%	75%	72%
Divorced	25%	27%	26%	22%	16%	21%
Widowed	24%	28%	25%	7%	9%	7%
Family size	2.21	2.19	2.20	2.32	2.27	2.31
Real estate	US$17,075.67	US$20,464.29	US$17,527.48	US$28,720.47	US$12,346.99	US$26,141.86
Accounts	US$22,754.82	US$24,080.29	US$22,931.55	US$20,750.78	US$30,040.12	US$22,213.73
Employed	32%	16%	30%	35%	24%	33%
Supplementary Income	US$819.73	US$520.70	US$779.86	US$769.48	US$1,144.55	US$828.55
Earnings	US$11,543.14	US$8,006.83	US$11,071.63	US$18,270.26	US$14,036.07	US$17,603.43
Medicare	45%	71%	49%	49%	66%	52%
Government health	52%	76%	55%	54%	68%	57%
LTC insurance	9%	11%	9%	7%	9%	8%
Drinks	0.69	0.55	0.67	1.59	1.04	1.50
Walking difficulty	34%	42%	35%	27%	43%	29%
Was happy	81%	84%	82%	84%	89%	85%
No. of doctor visits	10.55	14.22	11.04	7.56	10.05	7.96
n	530	196	726	407	183	590

Note. BMI = body mass index; LTC = Long-term care.

For men, 183 cease smoking, while 407 remain as smokers. Smokers are younger by 4 years and a higher percentage is employed with greater earnings. Men smokers also maintain significantly greater real estate valuations, while nonsmokers have a considerable advantage with bank account valuations. Once again, a greater percentage of nonsmokers receive Medicare, have a walking difficulty, and maintain additional doctor visits. As for BMI, the difference is substantial with smokers at 26.89 and nonsmokers at 28.47.

Fixed Effects (FE) Analyses: Total Sample

Model 1 shows the estimate for a transition to nonsmoking for the total sample (Table 3). Without controlling for any demographic, economic, or health attributes, a transition is associated with an increase of 1.32 for BMI. After adding Age/Marital Status/Family in Model 2, the estimate for nonsmoking increases to 1.47. The estimate remains robust in Model 3 with the addition of economic attributes and in Model 4 (final model) with the addition of health attributes. All things held equal, a transition to nonsmoking is associated with an increase of 1.44 for BMI in the final model (p< .01). For the average older adult with a height of 66 inches and weight of 178 pounds, this translates to an increase of over eight pounds (Ogden, 2004).

Table 3.

BMI Changes From Smoking Transition for Total Sample.

	(1)	(2)	(3)	(4)
	BMI	BMI	BMI	BMI
	b/SE	b/SE	b/SE	b/SE
Nonsmoke	1.32*** (0.21)	1.47*** (0.24)	1.44*** (0.24)	1.44*** (0.24)
Age/marital/family
Age 1: (50-59)		0.05* (0.03)	0.04 (0.03)	0.04 (0.04)
Age 2: (60-69)		0.03 (0.02)	0.02 (0.04)	0.02 (0.05)
Age 3: (70-79)		−0.15*** (0.03)	−0.13*** (0.03)	−0.14*** (0.03)
Age 4: (80+)		−0.25 (0.14)	−0.25 (0.17)	−0.25 (0.19)
Divorced		−0.20 (0.43)	−0.17 (0.47)	−0.17 (0.48)
Widowed		−0.86** (0.40)	−0.88** (0.41)	−0.86** (0.40)
Family size		−0.07 (0.08)	−0.10 (0.08)	−0.10 (0.09)
Economic
Ln real estate			−0.00 (0.02)	−0.00 (0.03)
Ln accounts			0.07*** (0.02)	0.07** (0.03)
Employed			0.12 (0.21)	0.13 (0.23)
Ln SSI			0.10** (0.05)	0.11* (0.06)
Ln earnings			−0.02 (0.02)	−0.02 (0.02)
Medicare			−0.10 (0.46)	−0.10 (0.46)
Government health			0.24 (0.40)	0.24 (0.40)
LTC insurance			−0.12 (0.31)	−0.14 (0.31)
Health
Drinks				0.03 (0.05)
Walk difficulty				0.24 (0.23)
Was happy (last 30 days)				0.20 (0.25)
Doctor visits				−0.00 (0.00)
Constant	26.67*** (0.05)	23.75*** (2.00)	24.29*** (2.18)	24.14*** (2.19)
Observations	2,632	2,632	2,632	2,632
Adjusted R²	.81	.81	.81	.82

Note. Adjusted R² obtained from areg command. BMI = body mass index; LTC = Long-term care; SSI = supplemental security income.

p< .10. **p< .05. ***p< .01.

Of all the age splines, only the third is significant and it suggests adults aged 70 to 79 are associated with a 0.14 decrease in BMI (p< .01). However, this estimate is practically insignificant. In other words, this amounts to a miniscule increase of approximately one pound. With respect to marital states, only widowed is significant and it is associated with a decrease of 0.86 in BMI. For the average older adult, this translates to a decrease of approximately five pounds. As for economic attributes, accounts are positively associated with an increase in BMI, but this estimate is practically insignificant. Although the number of drinks and walking difficulty variables are in the expected direction, they are not statistically significant.

FE Analyses: By Gender

FE estimates show older women who transition to nonsmoking are associated with an increase of 1.58 for BMI (Table 4). For the average woman aged 60+ with a height of 63 inches and weight of 165 pounds, this translates to over eight pounds (Ogden, 2004). This moderate increase is highly significant (p< .01) and occurs after controlling for unobserved individual-level heterogeneity. For this sample of older women, Age spline 3 (70-79) is associated with a slight decrease of 0.19 in BMI. The other age splines are not statistically significant. None of the marital states are significant nor is family size. However, an increase in accounts is associated with a very small increase in BMI.

Table 4.

BMI Changes From Smoking Transition by Gender.

	Women	Men
	BMI	BMI
	b/SE	b/SE
Nonsmoke	1.58*** (0.36)	1.24*** (0.31)
Age/Marital/Family
Age 1: (50-59)	0.04 (0.05)	0.03 (0.05)
Age 2: (60-69)	0.10 (0.08)	−0.04 (0.06)
Age 3: (70-79)	−0.19*** (0.05)	−0.07 (0.04)
Age 4: (80+)	0.31 (0.42)	−0.44** (0.18)
Divorced	0.35 (0.72)	−1.13** (0.44)
Widowed	−0.66 (0.47)	−1.09 (0.72)
Family size	−0.04 (0.12)	−0.25** (0.12)
Economic
Ln real estate	0.02 (0.04)	−0.04 (0.04)
Ln accounts	0.08* (0.04)	0.05 (0.03)
Employed	0.10 (0.34)	0.21 (0.29)
Ln SSI	0.14 (0.10)	0.08 (0.06)
Ln earnings	−0.03 (0.03)	−0.00 (0.03)
Medicare	−0.25 (0.72)	0.20 (0.52)
Government health	−0.41 (0.62)	0.92** (0.42)
LTC insurance	−0.01 (0.42)	−0.65* (0.38)
Health
Drinks	0.04 (0.11)	0.03 (0.05)
Walking difficulty	0.06 (0.31)	0.49 (0.33)
Was happy (last 30 days)	0.08 (0.35)	0.41 (0.34)
Doctor visits	−0.00 (0.01)	0.01 (0.01)
Constant	23.62*** (3.20)	24.97*** (2.97)
Observations	1,470	1,162
Adjusted R²	.816	.823

Note. Adjusted R² obtained from areg command. BMI = body mass index; LTC = Long-term care; SSI = supplemental security income.

p< .10. **p< .05. ***p< .01.

FE estimates for men are more interesting. A transition to nonsmoking is associated with an increase of 1.24 for BMI (Table 4). For the average man aged 60+ with a height of 69 inches and weight of 191 pounds, this translates to approximately eight pounds (Ogden, 2004). This moderate increase is highly significant (p< .01) and occurs after controlling for unobserved individual-level heterogeneity. For this sample of older men, Age spline 4 (80-89) is associated with a decrease of 0.44 in BMI. The other age splines are not statistically significant. As for marital states, divorced men experience a 1.13 decrease in BMI, which translates to approximately six pounds. The positive association for government health insurance is significant with an estimate of 0.92. Meanwhile, LTC insurance is associated with a decrease of 0.65, but it is only statistically significant at α = .10. The health attributes are in the expected direction, but insignificant.

FE Analyses: By Gender for Winsor Sample

The Winsor sample provides more robust estimators as it accounts for outliers. However, the overall impact and interpretation for women and men does not change substantially (Table 5). The Age/Marital/Family, Economic, and Health variables maintain direction and significance. One notable change occurs with the nonsmoke estimator for women. In the initial model, this variable equals 1.58 but reduces to 1.46 (p< .01). This suggests women with a transition from smoking to nonsmoking gain approximately seven pounds, which is less than the over eight pounds from the initial model (i.e., full sample without replacement of outliers). Please note: The Winsor sample BMI has a mean of 26.7 with a standard deviation of 4.9 (range = 18.8-39.5) as compared with the Total sample BMI mean of 26.8 with a standard deviation of 5.3 (range = 13.1-57.4).

Table 5.

BMI Changes for Smoking Transition by Gender (2.5% Winsor Sample).

	Women	Men
	BMI	BMI
	b/SE	b/SE
Nonsmoke	1.46*** (0.31)	1.26*** (0.29)
Age/marital/family
Age 1: (50-59)	0.04 (0.05)	0.02 (0.05)
Age 2: (60-69)	0.12* (0.07)	−0.05 (0.06)
Age 3: (70-79)	−0.18*** (0.05)	−0.08* (0.04)
Age 4: (80+)	0.32 (0.42)	−0.43** (0.18)
Divorced	0.20 (0.61)	−0.98** (0.42)
Widowed	−0.77* (0.41)	−0.97 (0.73)
Family size	−0.07 (0.11)	−0.24** (0.12)
Economic
Ln real estate	0.01 (0.04)	−0.02 (0.03)
Ln accounts	0.07* (0.04)	0.04 (0.03)
Employed	0.10 (0.31)	0.12 (0.28)
Ln SSI	0.16** (0.08)	0.06 (0.06)
Ln earnings	−0.03 (0.03)	−0.00 (0.03)
Medicare	−0.56 (0.58)	0.16 (0.52)
Government health	−0.15 (0.49)	0.82** (0.41)
LTC insurance	−0.11 (0.40)	−0.30 (0.37)
Health
Drinks	0.02 (0.09)	0.03 (0.05)
Walking difficulty	0.09 (0.30)	0.35 (0.30)
Was happy	0.04 (0.33)	0.38 (0.33)
Doctor visits	−0.00 (0.01)	0.01 (0.01)
Constant	23.91*** (2.95)	25.22*** (2.76)
Observations	1,470	1,162
Adjusted R²	.820	.819

Note. Adjusted R² obtained from areg command. BMI = body mass index; ; LTC = Long-term care; SSI = supplemental security income.

p< .10. **p< .05. ***p< .01.

Discussion

Contributions of This Study

Results from this article suggest smoking cessation is positively associated with a small change in BMI. More specifically, older adults experience a gain of approximately eight pounds. Some may not view this as a small amount. However, readers should recall this sample consists of middle aged and older adults where 42% of former smokers suffer from a walking difficulty. As well, a larger share of previous smokers is older by 4 years (Mage of approximately 68 years) and a smaller percentage is employed. As Models 2 and 3 show, after accounting for both demographic and economic attributes, the estimate for nonsmokers remains highly significant and robust. In other words, the estimator remains stable after adjusting for marital states, assets/earnings, and insurance. This also applies to Model 4. After adjusting for health attributes (drinking, walking difficulty, doctor visits), the estimator still maintains significance and size. These results support Hypothesis 1: After accounting for individual-level unobserved heterogeneity, there is a small increase in BMI due to a transition from smoking to nonsmoking.

This research also uncovered a slight gain in BMI occurs for both men and women. That is, unlike previous studies, there are no differences by age or gender for this sample of older adults. More specifically, men and women gain nearly eight pounds. These results support Hypothesis 2: The small increase in BMI occurs for both older men and older women. Such an outcome is plausible for this older age group given their strong desire to improve and maintain health due to new family roles (grandparents), long-awaited retirement (travel/leisure), and avoidance of expensive medical treatments (Luo, LaPierre, Hughes, & Waite, 2012; Syse, Veenstra, Furunes, Mykletun, & Solem, 2017). A study examining perceptions of successful aging found an overwhelming majority (nearly 90%) of adults aged 65 and older rated (1) remaining in good health, (2) staying involved with the world and people around them, and (3) remaining free of chronic diseases as important (Phelan, Anderson, Lacroix, & Larson, 2004). During the past decade, a greater segment of older adults has expressed a stronger desire to age successfully and, when possible, take the necessary steps to achieve this. A possible explanation for this may be that older adults are forward-looking. Unlike myopic behavior where individuals primarily emphasize the current time period, older adults may place considerable value on future time periods (Arcidiacono et al., 2007). With increased life spans and significant advancements in geriatric medicine during the past half-century, older adults may discount the future at a lower rate.

Intervention: Self-Awareness and Public Awareness

Given that some middle aged and older adults may be willing to give greater consideration to future time periods, encouraging this group to cease smoking is a logical starting point. One practical and cost-effective public health effort relates to self-awareness. That is, a recognition that discontinued tobacco use may result in a small increase in BMI, but this should not be a source of anxiety or worry. In addition, an increase in BMI may not be entirely negative. A yearlong longitudinal study of a smoking cessation program found evidence for weight gain and this increase was not only attributed to fat but also (a) muscle mass, (b) muscle strength, and (c) bone density (Rom, Reznick, Keidar, Karkabi, & Aizenbud, 2015). Self-awareness is an important consideration for another reason: Individuals who gain weight on a previous attempt may have reservations and lack confidence in future attempts (Veldheer et al., 2014). Therefore, educating smokers about the possibility for and acceptance of weight gain could result in greater cessation rates and better views of body image.

Public Health Efforts

This research utilized the advantages of panel data to show increased BMI from smoking cessation. However, the amount of gain for both men and women is moderate, at best. After accounting for individual-level unobserved heterogeneity and controlling for demographic, economic, and health attributes, middle aged and older adults gain nearly eight pounds. This is less than the approximately 11 pounds found by Veldheer et al. (2015), Gennuso et al. (2014), Rom et al. (2014), Aubin et al. (2012), and significantly less than the over 15 pounds found by Lycett, Munafò, Johnstone, Murphy, and Aveyard (2011). Gerontologists/public health workers can take proactive steps to ensure older adults are aware of this. Informing older adults about the trade-offs from smoking cessation is worthwhile because many of the diseases and deaths attributable to tobacco use are preventable (Danaei et al., 2009; U.S. Department of HHS, 1990). Utilizing the Global Burden of Disease Study from 2010, Lim et al. (2013) found tobacco smoking as one of the three leading factors for disease in 1990 and again in 2010. These researchers also noted 31% of disability adjusted life years for ischemic heart disease can be attributed to tobacco smoking as a risk factor. Equally worrisome, tobacco smoking was among the top five leading risk factors for disease burden in most of the world in 2010. Given the aforementioned concerns regarding health, medical costs, and economic losses, the U.S. Department of Public Health is wise in continuing to educate and campaign against smoking. This worthwhile public health investment is noted in Healthy People 2020, which lists one major goal being a reduction in the percentage of adult smokers (i.e., aged 18 and above) from 18.1%, as of 2012, to no more than 12% by 2020 (Agaku, King, Dube, & CDC, 2014).

Strengths of Study

This research contributes to an improved understanding of smoking cessation and BMI/weight gain by utilizing a nationally representative and well-defined population. Whereas most studies focus on the entire adult population or a population with a large age bracket (e.g., 30+), this study sample consists only of adults aged 50 and older. Another noteworthy attribute is the use of panel data and FE regression to account for unobserved heterogeneity, efficiency, and changes in smoking cessation and BMI over time. Finally, this study advances the field of gerontology by utilizing interdisciplinary frameworks from economics, psychology, and public health (i.e., utility models, forward-looking behavior).

Limitations of Study

There are a few limitations with this research. For one, individuals with a transition to nonsmoking status in 2010 may have initiated multiple attempts. This is not observed in the data. Along the same lines, exactly when an individual ceased smoking is not observed. In other words, an individual may have ceased at any point after 2004 but before 2010. Although not critical given the stability of the estimates (see Table 3), having these covariates as controls would provide greater insight into the behavioral aspect of cessation. That is, how often do individuals fail and what prompts an additional effort. Multiple cessation attempts can also impact changes in BMI. Perhaps, each failed attempt with an increase in BMI may, when aggregated, result in significant weight gain. As stated previously, self-reporting in another major limitation and underreporting has the potential to bias the estimator. Finally, this study could benefit from a larger sample. Having an additional 200 or so observations would introduce greater variability in the data. This remains a challenge for researchers studying an older adult population.

Conclusion

This study is among a select group to examine the association between smoking cessation and changes in BMI for adults 50 years and older. Utilizing the HRS, fixed-effects analyses show a small change in BMI after a transition from smoking to nonsmoking during a 6-year period, and this occurs after accounting for individual-level unobserved heterogeneity. More specifically, men experience a BMI gain of 1.24 and women experience a BMI gain of 1.58. Health practitioners/gerontologists can use this finding to assuage fears surrounding weight gain from smoking cessation and encourage older adults to consider discontinuing tobacco use. Future research can expand this area of inquiry by conducting a similar analysis by investigating race. Blacks, Hispanics, and Whites will have different experiences from smoking cessation and may require different intervention strategies. Such an undertaking would help smoking cessation specialists develop targeted solutions and, in the process, help achieve the goals of Health People 2020.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Author Biography

Andy Sharma is a political economist with research interests in aging, health disparities, later-life migration, spatial analysis, and quantiative methods.

References

Agaku

I. T.

King

B. A.

Dube

S. R.

, & Centers for Disease Control and Prevention (CDC). (2014). Current cigarette smoking among adults—United States, 2005–2012. Morbidity and Mortality Weekly Report, 63(2), 29-34.

Arcidiacono

Sieg

Sloan

(2007). Living rationally under the volcano? An empirical analysis of heavy drinking and smoking. International Economic Review, 48, 37-65.

Aubin

H. J.

Farley

Lycett

Lahmek

Aveyard

(2012). Weight gain in smokers after quitting cigarettes: Meta-analysis. British Medical Journal, 345, Article e4439.

Baltagi

(2008). Econometric analysis of panel data. John Wiley & Sons.

Becker

M. H.

(1974). The health belief model and personal health behavior. Thorofare, NJ: Slack.

Burgard

S. A.

Ailshire

J. A.

Kalousova

(2013). The great recession and health people, populations, and disparities. The ANNALS of the American Academy of Political and Social Science, 650, 194-213.

Bush

T. M.

Levine

M. D.

Magnusson

Cheng

Chen

Mahoney

Zbikowski

S. M.

(2014). Impact of baseline weight on smoking cessation and weight gain in quitlines. Annals of Behavioral Medicine, 47, 208-217.

Bush

T. M.

Levine

M. D.

Magnusson

Cheng

Chen

Mahoney

. . . Zbikowski

S. M.

(2014). Impact of baseline weight on smoking cessation and weight gain in quitlines. Annals of Behavioral Medicine, 47(2), 208-217.

Caraballo

R. S.

Giovino

G. A.

Pechacek

T. F.

Mowery

P. D.

(2001). Factors associated with discrepancies between self-reports on cigarette smoking and measured serum cotinine levels among persons aged 17 years or older: Third National Health and Nutrition Examination Survey, 1988-1994. American Journal of Epidemiology, 153, 807-814.

10.

Case

Deaton

A. S.

(2005). Broken down by work and sex: How our health declines. In David

(Ed.), Analyses in the economics of aging (pp. 185-212). Chicago, IL: University of Chicago Press.

11.

Choi

N. G.

DiNitto

D. M.

(2015). Role of new diagnosis, social isolation, and depression in older adults’ smoking cessation. The Gerontologist, 55, 793-801.

12.

Choi

W. S.

Harris

K. J.

Okuyemi

Ahluwalia

J. S.

(2003). Predictors of smoking initiation among college-bound high school students. Annals of Behavioral Medicine, 26, 69-74.

13.

Clair

P. S.

Bugliari

Campbell

Chien

Hayden

Hurd

Pantoja

(2011). RAND HRS data documentation, version L (Technical report). Santa Monica, CA: Labor & Population Program, RAND Center for the Study of Aging.

14.

Clark

M. M.

Hurt

R. D.

Croghan

I. T.

Patten

C. A.

Novotny

Sloan

J. A.

Bernath

(2006). The prevalence of weight concerns in a smoking abstinence clinical trial. Addictive Behaviors, 31, 1144-1152.

15.

Danaei

Ding

E. L.

Mozaffarian

Taylor

Rehm

Murray

C. J.

Ezzati

(2009). The preventable causes of death in the United States: Comparative risk assessment of dietary, lifestyle, and metabolic risk factors. PLoS Medicine, 6(4), e1000058.

16.

Dube

S. R.

L. T.

(2015). Cigarette smoking among adults aged 45 and older in the United States, 2002–2011. In Substance use and older people (pp. 57-74).Chichester, UK: Wiley-Blackwell.

17.

Fenge

L. A.

Hean

Worswick

Wilkinson

Fearnley

Ersser

(2012). The impact of the economic recession on well-being and quality of life of older people. Health and Social Care in the Community, 20, 617-624.

18.

Flegal

K. M.

Troiano

R. P.

Pamuk

E. R.

Kuczmarski

R. J.

Campbell

S. M.

(1995). The influence of smoking cessation on the prevalence of overweight in the United States. New England Journal of Medicine, 333, 1165-1170.

19.

Fletcher

C. M.

Horn

(1970, August). Smoking and health. WHO Chronicle, 24(8), 345-370.

20.

Freedman

D. S.

Thornton

J. C.

Pi-Sunyer

F. X.

Heymsfield

S. B.

Wang

Pierson

R. N.

. . . Gallagher

(2012). The body adiposity index (hip circumference ÷ height(1.5)) is not a more accurate measure of adiposity than is BMI, waist circumference, or hip circumference. Obesity, 20, 2438-2444.

21.

Freedman

K. S.

Nelson

N. M.

Feldman

L. L.

(2012). Smoking initiation among young adults in the United States and Canada, 1998-2010: A systematic review. Preventing Chronic Disease, 9, p. E05.Article 110037.

22.

Garrett

B. E.

Dube

S. R.

Winder

Caraballo

R. S.

(2013). Cigarette Smoking—United States, 2006-2008 and 2009-2010. Morbidity and Mortality Weekly Report, 62(3), 81-84.

23.

Gellert

Schöttker

Brenner

(2012). Smoking and all-cause mortality in older people: Systematic review and meta-analysis. Archives of Internal Medicine, 172, 837-844.

24.

Gennuso

K. P.

Thraen-Borowski

K. M.

Schlam

T. R.

LaRowe

T. L.

Fiore

M. C.

Baker

T. B.

Colbert

L. H.

(2014). Smokers’ physical activity and weight gain one year after a successful versus unsuccessful quit attempt. Preventive Medicine, 67, 189-192.

25.

Gilpin

E. A.

Choi

W. S.

Berry

Pierce

J. P.

(1999). How many adolescents start smoking each day in the United States? Journal of Adolescent Health, 25, 248-255.

26.

Gorber

S. C.

Schofield-Hurwitz

Hardt

Levasseur

Tremblay

(2009). The accuracy of self-reported smoking: A systematic review of the relationship between self-reported and cotinine-assessed smoking status. Nicotine & Tobacco Research, 11, 12-24.

27.

Grossman

(1972a). The demand for health: A theoretical and empirical investigation. New York, NY: National Bureau of Economic Research.

28.

Grossman

(1972b). On the concept of health capital and the demand for health. The Journal of Political Economy, 80, 223-255.

29.

Grossman

(2000). The human capital model. Handbook of Health Economics, 1, 347-408.

30.

Grubbs

F. E.

(1969). Procedures for detecting outlying observations in samples. Technometrics, 11, 1-21.

31.

Janz

N. K.

Becker

M. H.

(1984). The health belief model: A decade later. Health Education & Behavior, 11, 1-47.

32.

Jeffery

R. W.

Hennrikus

D. J.

Lando

H. A.

Murray

D. M.

Liu

J. W.

(2000). Reconciling conflicting findings regarding postcessation weight concerns and success in smoking cessation. Health Psychology, 19, 242-246.

33.

Kasteridis

Yen

S. T.

(2012). Smoking cessation and body weight: Evidence from the Behavioral Risk Factor Surveillance Survey. Health Services Research, 47, 1580-1602.

34.

Karp

(2007).Growing older in America: The health and retirement study. National Institute on Aging, National Institutes of Health, US Department of Health and Human Services.

35.

Klesges

R. C.

Meyers

A. W.

Klesges

L. M.

LaVasque

M. E.

(1989). Smoking, body weight, and their effects on smoking behavior: A comprehensive review of the literature. Psychological Bulletin, 106, 204-230.

36.

Klesges

R. C.

Winders

S. E.

Meyers

A. W.

Eck

L. H.

Ward

K. D.

Hultquist

C. M.

Shadish

W. R.

(1997). How much weight gain occurs following smoking cessation? A comparison of weight gain using both continuous and point prevalence abstinence. Journal of Consulting and Clinical Psychology, 65, 286-291.

37.

Lim

S. S.

Vos

Flaxman

A. D.

Danaei

Shibuya

Adair-Rohani

Davis

(2013). A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: A systematic analysis for the Global Burden of Disease Study 2010. The Lancet, 380, 2224-2260.

38.

Lin

I. F.

Brown

S. L.

(2012). Unmarried boomers confront old age: A national portrait. The Gerontologist, 52, 153-165.

39.

Luo

LaPierre

T. A.

Hughes

M. E.

Waite

L. J.

(2012). Grandparents providing care to grandchildren a population-based study of continuity and change. Journal of Family Issues, 33, 1143-1167.

40.

Luostarinen

Tuovinen

E. L.

Saarni

S. E.

Kinnunen

Hukkinen

Haukkala

Korhonen

(2013). Weight concerns among Finnish ever-smokers: A population-based study. Nicotine & Tobacco Research, 15, 1696-1704.

41.

Lycett

Munafò

Johnstone

Murphy

Aveyard

(2011). Associations between weight change over 8 years and baseline body mass index in a cohort of continuing and quitting smokers. Addiction, 106, 188-196.

42.

Meyers

A. W.

Klesges

R. C.

Winders

S. E.

Ward

K. D.

Peterson

B. A.

Eck

L. H.

(1997). Are weight concerns predictive of smoking cessation? A prospective analysis. Journal of Consulting and Clinical Psychology, 65, 448-452.

43.

Miller

V. P.

Ernst

Collin

(1999). Smoking-attributable medical care costs in the USA. Social Science & Medicine, 48, 375-391.

44.

Muurinen

J. M.

(1982). Demand for health: A generalised Grossman model. Journal of Health Economics, 1, 5-28.

45.

Muurinen

J. M.

Le Grand

(1985). The economic analysis of inequalities in health. Social Science & Medicine, 20, 1029-1035.

46.

National Center for Health Statistics (2012). Older Americans 2012: Key indicators of wellbeing. Federal Interagency Forum on Aging Related Statistics, p. 176.

47.

Ogden

C. L.

(2004). Mean body weight, height, and body mass index, United States 1960–2002. Washington, DC: Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics.

48.

Orleans

C. T.

Jepson

Resch

Rimer

B. K.

(1994). Quitting motives and barriers among older smokers. The 1986 Adult Use of Tobacco Survey revisited. Cancer, 74(S7), 2055-2061.

49.

Orleans

C. T.

Rimer

B. K.

Cristinzio

Keintz

M. K.

Fleisher

(1991). A national survey of older smokers: Treatment needs of a growing population. Health Psychology, 10, 343-351.

50.

Patrick

D. L.

Cheadle

Thompson

D. C.

Diehr

Koepsell

Kinne

(1994). The validity of self-reported smoking: A review and meta-analysis. American Journal of Public Health, 84, 1086-1093.

51.

Perkins

K. A.

Levine

M. D.

Marcus

M. D.

Shiffman

(1997). Addressing women’s concerns about weight gain due to smoking cessation. Journal of Substance Abuse Treatment, 14, 173-182.

52.

Phelan

E. A.

Anderson

L. A.

Lacroix

A. Z.

Larson

E. B.

(2004). Older adults’ views of “successful aging”—How do they compare with researchers’ definitions? Journal of the American Geriatrics Society, 52, 211-216.

53.

Reidpath

D. D.

Davey

T. M.

Kadirvelu

Soyiri

I. N.

Allotey

(2014). Does one cigarette make an adolescent smoker, and is it influenced by age and age of smoking initiation? Evidence of association from the U.S. Youth Risk Behavior Surveillance System (2011). Preventive Medicine, 59, 37-41.

54.

Rom

Reznick

A. Z.

Keidar

Karkabi

Aizenbud

(2015). Smoking cessation-related weight gain—Beneficial effects on muscle mass, strength and bone health. Addiction, 110, 326-335.

55.

Rosenstock

I. M.

(1974). The health belief model and preventive health behavior. Health Education Monographs, 2, 354-386.

56.

Rosenstock

I. M.

(2008). The health belief model: Explaining health behavior through expectancies. In Glanz

Rimer

B. K.

Viswanath

(Eds.), Health behavior and health education: Theory, research, and practice (pp. 39-62). San Francisco, CA: John Wiley. (Original work published 1990)

57.

Shah

N. R.

Braverman

E. R.

(2012). Measuring adiposity in patients: The utility of body mass index (BMI), percent body fat, and leptin. PLoS ONE, 7(4), e33308.

58.

Strecher

V. J.

DeVellis

B. M.

Becker

M. H.

Rosenstock

I. M.

(1986). The role of self-efficacy in achieving health behavior change. Health Education & Behavior, 13, 73-92.

59.

Syse

Veenstra

Furunes

Mykletun

R. J.

Solem

P. E.

(2017). Changes in health and health behavior associated with retirement. Journal of Aging and Health, 29(1), 99-127.

60.

Terry

(1973). Combating the health hazards of smoking. CA: A Cancer Journal for Clinicians, 23, 362-366.

61.

Travier

Agudo

May

A. M.

Gonzalez

Luan

J. A.

Wareham

N. J.

Clavel-Chapelon

(2012). Longitudinal changes in weight in relation to smoking cessation in participants of the EPIC-PANACEA study. Preventive Medicine, 54, 183-192.

62.

Treviño

L. A.

Baker

McIntosh

Mustian

Seplaki

C. L.

Guido

J. J.

Ossip

D. J.

(2014). Physical activity as a coping strategy for smoking cessation in mid-life and older adults. Addictive Behaviors, 39, 885-888.

63.

Trinidad

D. R.

Gilpin

E. A.

Lee

Pierce

J. P.

(2004). Do the majority of Asian-American and African-American smokers start as adults? American Journal of Preventive Medicine, 26, 156-158.

64.

Unger

J. B.

Chen

(1999). The role of social networks and media receptivity in predicting age of smoking initiation: A proportional hazards model of risk and protective factors. Addictive Behaviors, 24, 371-381.

65.

United States. (1964). Surgeon General’s Advisory Committee on Smoking and Health: Smoking and health—Report of the Advisory Committee to the Surgeon General of the public health service. Washington, DC: U.S. Public Health Service.

66.

U.S. Department of Health and Human Services. (1982). Surgeon General’s Advisory Committee on Smoking and Health: The health consequences of smoking—A report of the Surgeon General. Washington D.C.

67.

U.S. Department of Health and Human Services. (1988). The health consequences of smoking.

68.

U.S. Department of Health and Human Services. (1990). The health benefits of smoking cessation: A report of the Surgeon General (Publication No. 90-8416).Washington, DC.

69.

U.S. Department of Health and Human Services. (1994). Preventing tobacco use among young people: A report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.

70.

U.S. Department of Health and Human Services. (2004). The health consequences of smoking: A report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.

71.

U.S. Department of Health and Human Services. (2006). The health consequences of involuntary exposure to tobacco smoke: A report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.

72.

US Department of Health and Human Services. (2010). How tobacco smoke causes disease: the biology and behavioral basis for smoking-attributable disease: a report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2.

73.

U.S. Department of Health and Human Services. (2014). The health consequences of smoking—50 years of progress: A report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health

74.

Veldheer

Yingst

Zhu

Foulds

(2015). Ten-year weight gain in smokers who quit, smokers who continued smoking and never smokers in the United States, NHANES 2003–2012. International Journal of Obesity, 39(12), 1727-1732.

75.

Veldheer

Yingst

Foulds

Hrabovsky

Berg

Sciamanna

Foulds

(2014). Once bitten, twice shy: Concern about gaining weight after smoking cessation and its association with seeking treatment. International Journal of Clinical Practice, 68, 388-395.

76.

Veldheer

Yingst

Zhu

Foulds

(2015). Ten-year weight gain in smokers who quit, smokers who continued smoking and never smokers in the United States, NHANES 2003-2012. International Journal of Obesity, 39, 1727-1732.

77.

Warner

K. E.

Hodgson

T. A.

Carroll

C. E.

(1999). Medical costs of smoking in the United States: Estimates, their validity, and their implications. Tobacco Control, 8, 290-300.

78.

Wooldridge

J. M.

(2010). Econometric analysis of cross section and panel data. MIT press.

79.

Weng

S. F.

Ali

Leonardi-Bee

(2013). Smoking and absence from work: Systematic review and meta-analysis of occupational studies. Addiction, 108, 307-319.

80.

Williamson

D. F.

Madans

Anda

R. F.

Kleinman

J. C.

Giovino

G. A.

Byers

(1991). Smoking cessation and severity of weight gain in a national cohort. New England Journal of Medicine, 324, 739-745.

81.

Bishop

E. E.

Kennedy

S. M.

Simpson

S. A.

Pechacek

T. F.

(2014). Annual healthcare spending attributable to cigarette smoking: An update. American Journal of Preventive Medicine, 48, 326-333.