Skip to main content
NCBI home page
PLOS One logoLink to PLOS One
. 2025 Apr 30;20(4):e0319560. doi: 10.1371/journal.pone.0319560

Determining whether weight status mediates the association between number of cigarettes smoked per day and all-cause mortality among US adults who smoke cigarettes

Luis Miguel Mestre 1,2,*, Roger S Zoh 2, Cydne Perry 3, Julia Fukuyama 4, Maria A Parker 2
Editor: Jinyi Wu5
PMCID: PMC12043237  PMID: 40305568

Abstract

Introduction

While there is evidence demonstrating the association between cigarette smoking and weight status, and mortality and weight status, it has not been examined whether weight status is a mediator between number of cigarettes smoked per day (CPD) and all-cause mortality, limiting our knowledge of this association and potential novel approaches to reduce all-cause mortality due to cigarette smoking. We aimed to evaluate whether weight status mediated the association between CPD and mortality.

Methods

We harnessed the 2003-2018 NHANES and the Linkage Mortality Files, which included adults who smoked ≥ 100 lifetime cigarettes (unweighted n = 5,676). A generalized linear model estimated the association between cigarettes smoked per day (CPD) and weight status (e.g., Body Mass Index (BMI) or Waist Circumference (WC)). An Accelerated Failure Time model with a Weibull distribution estimated the association between CPD and all-cause mortality with weight status as a mediator, adjusting for age, SES, alcohol consumption, race/ethnicity, sex/gender, blood pressure, total cholesterol, and physical activity.

Results

Between 2003-2018, the sample’s mean BMI was 27.97 kg/m2, sample’s mean WC was 97.58 cm and mean CPD was 13.21. The total effect in the mediation analysis of WC adjusted by BMI levels in the association between CPD and all-cause mortality was -0.44 (95% CI =  -2.00, -0.20; p =  0.016), the average direct effect was -0.35 (95% CI =  -1.86, -0.10; p =  0.036), and the average indirect effect was -0.10 (95% CI =  -0.23, -0.05; p < 0.001).

Conclusion

WC, as a surrogate measure of weight status, when adjusted by BMI levels, was a partial mediator between CPD and all-cause mortality. Public health interventions aimed to reduce mortality due to cigarette smoking at the population level should consider weight management programs as a harm reduction strategy to reduce mortality.

Introduction

Based on recent statistics, more than 28 million United States (US) adults currently smoke cigarettes [1], while more than 35% have obesity [2], and approximately 2% are classified as underweight [3]. Cigarette smoking and weight status are considered primary public health concerns [4] and major causes of preventable deaths in the US [5]. Although the association between cigarette smoking and weight status is well known [6–27], it is not entirely understood [21], which makes it difficult to prevent death due to combined cigarette smoking and weight status. Multiple studies have shown different directions and magnitudes of the association between cigarette smoking and weight status, especially when using a surrogate measures of weight status such as Body Mass Index (BMI) [6–14,16,18–21].

Individuals who smoke cigarettes tend to have a lower BMI than those who do not smoke cigarettes [6–12,15,19]. However, among those who smoke cigarettes, their waist circumference (WC), and thus central obesity, tend to be higher than those who do not smoke cigarettes [6,11,16–26]. Some studies have found that those who smoke less than 20 CPD (i.e., non-heavy smoking) have a lower BMI than those who do not smoke cigarettes; those who smoke 20 or more CPD (i.e., heavy smoking) have a higher BMI than those who smoke non-heavily or do not smoke cigarettes [6–8,20,21]. This is evidence of a U-shaped association between number of cigarettes smoked per day and BMI at the population level [7]. The reason why individuals who smoke 20 or more CPD tend to have a higher BMI than those with non-heavy smoking or who do not smoke remains unclear [7,11,21]. An accepted hypothesis is that those who heavily smoke also engage in other unhealthy behaviors [12], such as higher alcohol consumption and lower physical activity, behaviors associated with higher levels of BMI [25,28–30]. There is also evidence [10,16,17,22] of a positive dose-response relationship between the CPD and weight status when measured as WC; the higher the CPD, the higher the WC of the person who smokes cigarettes [17,22]. Thus, there is an association between cigarette smoking and weight status, especially central obesity, when approximated as WC [20,23,24,27].

It is well known that cigarette smoking [31–38] and weight status [39–53] are associated with all-cause mortality due to being two of the main risk factors of chronic diseases [35,36,38,45,46,48] and chronic inflammation [11,48]. Individuals who smoke cigarettes have at least a two-fold increasing risk of all-cause mortality [31,34–37] and lose at least ten years of life expectancy [38] than those who do not smoke cigarettes. Among those who smoke cigarettes, there is a dose-response relationship between the number of cigarettes smoked and the risk of all-cause mortality; the higher the number of cigarettes smoked, the higher the risk of all-cause mortality [32,33], even if the smoking behavior is nondaily, or < 1 CPD [33].

Prior studies have shown that individuals who are underweight (BMI <  18.5 kg/m2) or have obesity (BMI ≥  30 kg/m2) experience a higher risk of all-cause mortality than those with a normal weight or who are overweight [39,41–47,50]. In other words, BMI and all-cause mortality have a U-shaped relationship [39,43,44,46,50,54]. Such a U-shaped relationship is moderated by race/ethnicity and sex/gender, indicating that physiological and social factors play a role in the association between BMI and all-cause mortality [39,47]. Cohen et al. [39] found that the Non-Hispanic Black population has lower mortality rates at obesity levels though higher mortality rates at underweight levels than the Non-Hispanic White population; similarly, women are more likely to experience all-cause mortality than men within the same BMI at obesity or underweight levels. WC as a predictor of central obesity is also strongly related to all-cause mortality [48,52,55,56] regardless of BMI levels. There is a dose-response relationship between WC and all-cause mortality; the higher the WC, the higher the risk of all-cause mortality for the same BMI level [51–53,55].

A popular hypothesis to explain the association between cigarette smoking and weight status and the association with all-cause mortality was done by Garrison et al. [57] using the Framingham Study Dataset. Garrison et al. popularized the hypothesis that cigarette smoking was a moderator in the association between weight status and all-cause mortality though this was not replicated by Sempos et al. [58]. Using the same data from Framingham Study, Sempos et al. [58] found that smoking cigarettes is associated with both all-cause mortality and BMI, though not a moderator in the association between BMI and all-cause mortality, a finding that has been replicated and found in other studies in the US adult population [43,59]. Therefore, cigarette smoking and weight status are associated between them and with all-cause mortality though not necessarily through an interaction.

Nonetheless, to our knowledge, weight status as a mediator between cigarette smoking and all-cause mortality, especially CPD and all-cause mortality, has not been evaluated. A mechanism that supports weight status as a potential mediator at the population level in the association between CPD and all-cause mortality is that the higher the CPD, the higher the levels of central obesity, when approximated by WC [10,16,17,20,22,27], regardless of the direction of the association between the CPD and BMI; central obesity as well as CPD are strongly associated with all-cause mortality [48,55]. In this study, we hypothesized that weight status explains at the population level, at least partially, the association between CPD and all-cause mortality in the population of US adults who currently smoke cigarettes.

The objective of this study was to determine whether weight status, assessed by BMI or WC, mediates the association between CPD and all-cause mortality. Using NHANES, a nationally representative, cross-sectional survey data from the US adult population, we specifically aimed to 1) characterize cigarette smoking, weight status, and all-cause mortality and 2) quantify the association between CPD, weight status, and all-cause mortality.

Methods

Study population

We harnessed the 2003-2018 NHANES, which includes US adults (18+). Every NHANES Survey Cycle consists of two years (e.g., 2003-2004), and participants were included by year. NHANES includes data such as biomarkers, demographics, health behaviors (i.e., CPD, alcohol consumption, and physical activity), and measurements (i.e., BMI, WC, and body fat percentage). The mortality data of NHANES’s participants came from the National Center of Health Statistics (NCHS) data linkage, Linked Mortality Files (LMF). The LMF included a limited set of mortality data for adult participants, including follow-up from 2003 to 2019 or until death occurred, whichever was first [60]; multiple survey cycles of NHANES merged with LMF were used to ensure temporality between past 30-day cigarette smoking and all-cause mortality.

We included US adults (+18 years) who had smoked more than one hundred cigarettes and who smoked cigarettes in the past 30 days. We did not include those who formerly smoked cigarettes due to not having data about the CPD in the past 30-days. Fig 1 summarizes the combined NHANES from 2003 to 2018 and LMF of mortality follow-up from 2003 to 2019 of the adult population from the US. We excluded only one participant due to having extreme values in their self-reported physical activity (the participant reported a value of 908 min/day). The other participants were excluded due to missing data from all considered factors (n = 14,565 of adults with past 30-day smoking), resulting in a final analytic sample of n =  5,676, representing 29,862,731 US adults with past 30-day smoking. This study was approved by the IRB of Indiana University (IRB Protocol #16545).

Fig 1. Flow chart of the sample size of adults with past 30-day cigarette smoking in the US from the continuous NHANES, 2003-2018.

Fig 1

Open in a new tab

Scenario

The objectives were accomplished through a mediation analysis in the following scenario: those who currently smoke cigarettes started to smoke cigarettes before the weight status reported at the moment of the interview (verifiable with the comparison of age when starting to smoke cigarettes regularly with the age at the moment of the interview). Therefore, current cigarette smoking occurred first than the weight status at the moment of the interview, and it was assumed, together with weight status, to be fixed during the follow-up time of all-cause mortality.

Measurements

Cigarette smoking (exposure).

CPD was collected through “During the past 30 days, on the days that you smoked, about how many cigarettes did you smoke per day?”

Weight status (mediator).

Weight status, measured by BMI, was calculated as weight (kg)/height2 (m2) by the NHANES staff using in-person weight and height measurements. NHANES staff measured WC (cm) using a steel tape while the participant was standing and holding the gown up; the waist level was defined at the abdomen [61]. We used BMI and WC as surrogate measures of weight status for these analyses because lean mass and visceral tissue are important considerations for health outcomes such as all-cause mortality. Lean mass and visceral tissue are included in BMI and WC but are not measured by body fat percentage [62]. Yet, we also included body fat percentages (Android and Gyroid) as surrogate measures of weight status in our analyses to compare results. Body fat percentages (both Android and Gynoid) were measured using a dual-energy ray absorptiometry (DXA) for those without pregnancy and the bioelectrical impedance analysis (BIA) for those who were pregnant at the moment of the interview [61].

All-cause mortality (outcome).

Mortality status was a time-to-event variable that indicated whether the participant had died in the study period. Together with mortality status, follow-up time (in years) was used. The follow-up time was measured as the time from the moment of the follow-up interview at the mobile examination center until 2019 or until the participant was reported dead, whichever occurred first.

Background characteristics (covariates).

Factors that might influence the association between cigarette smoking and all-cause mortality, especially CPD and all-cause mortality, were included as covariates. Cigarette smoking, weight status, and all-cause mortality vary by sex/gender, age, race/ethnicity, and SES [36,39,43,47,54], as well as physical activity and alcohol consumption [25,28,29,63]. Biomarkers such as blood pressure status and total cholesterol are directly related to all-cause mortality [45,64].

Sex/gender was a dichotomous variable indicating whether the participant self-identified as male or female. Age was a continuous variable that ranged from 18 to 85 years old. The age of participants older than 85 years old was recorded as 85 years old to preserve confidentiality. Race/ethnicity was a polychotomous categorical variable that included the following: Hispanic Mexican, Hispanic Other (i.e., besides Mexican heritage), Non-Hispanic White, Non-Hispanic Black, and Non-Hispanic Other. SES was defined as the ratio of family income to poverty, a continuous variable that ranged from zero to five, where values greater than five were given a value of five by NHANES staff; the closer the value was from zero, the higher the poverty level of the participant [65]. Physical activity was operationalized as the total moderate physical activity per day through the sum of minutes per day of walking or using the bicycle, the minutes per day of moderate physical activity, and the minutes per day of vigorous physical activity times two [66]. Alcohol consumption was defined as the average number of drinks drank per week. Blood pressure and total cholesterol were dichotomous variables that indicated whether the participant had a systolic blood pressure of  ≥  120 mm Hg and a diastolic blood pressure <  80 mm Hg or whether the total cholesterol was  ≥  200 mg/dL, respectively.

Statistical analysis

We did all analyses in R Studio; we used the survey package to adjust data based on their sampling design and weights and to construct the models for the mediation analysis. We used the mediation package for the mediation analysis. Descriptive statistics of the merged NHANES dataset were adjusted by the survey weights of the mobile examination center divided by the number of survey cycles as recommended by NHANES guidelines [67].

The mediation analysis used bootstrapping with 1,000 simulations to obtain uncertainty estimates around the parameter estimates. We ran 1,000 simulations because mediation estimates converged (i.e., there was no meaningful variability among the estimates of interest). Bootstrapping allows to measure the confidence intervals for such statistical estimates without making any assumption about its distribution (i.e., non-parametric), and it is known to be a robust method with a low risk of biased results. An outcome-exposure model and a mediator-exposure model were constructed for the mediation analysis to calculate the indirect, direct, and total effects. We used an AFT model with a Weibull distribution for the outcome-exposure model. We used the AFT Model with a Weibull Distribution because the Cox PH Model assumption, proportionality of hazards, was unmet for some of the covariates, which can produce biased results.

Based on the model selection criterion (AIC and BIC), the AFT Model with a Weibull Distribution was the best fit for these data compared to the Cox PH Model for all surrogate measures of weight status. Results obtained from the AFT Model with a Weibull Distribution can be interpreted as a Cox PH Model. The AFT model included the covariates related to all-cause mortality. Due to the evidence of a U-shaped relation between BMI and the risk of all-cause mortality [11,39,41,43,44,46], a quadratic term of BMI was included when BMI was used as the surrogate measure of weight status. For WC, we did not include a quadratic term due to the dose-response relationship between WC and all-cause mortality being somewhat linear [48,55,56].

We constructed the mediator-exposure model using a generalized linear model with a Gaussian distribution, including covariates associated with both CPD and weight status. Based on Cohen et al. [39], BMI was moderated by race/ethnicity and sex/gender and thus, interactions of these covariates were included; for WC, such interactions were not included. The surrogate measure of weight status, body fat percentages (Android and Gynoid), were also considered for the mediation analyses using the same rationale for WC due to also being related to fat mass.

As in previous studies [20,23,48], we did an analysis to adjust WC by BMI levels. Regardless of the change in BMI by CPD, the WC, and thus central obesity, increases as the CPD increases. Such mediation analysis included a quadratic BMI term in the all-cause mortality-cigarette smoking, though only a BMI term in the waist-circumference-cigarette smoking model. Although BMI and WC are highly correlated, WC for a given BMI is a strong predictor of all-cause mortality [68,69]. Furthermore, WC is a measurement of central obesity, while BMI is a measurement of general obesity. Therefore, a person with a certain weight status (not necessarily obesity) can have central obesity. Thus, by adjusting WC for different BMI levels, we can assess central obesity in different weight statuses by BMI.

For sensitivity analyses, we verified the assumption for mediation analysis, sequential ignorability. First, based on prior knowledge, we identified potential confounders, such as alcohol consumption, sex/gender, age, blood pressure, total cholesterol, race/ethnicity, physical activity, and SES. Using different non-parametric association/correlation tests depending on whether the variables were continuous, categorical, or a mix, it was verified whether such potential confounders were actual confounders.

Results

Descriptive statistics

Table 1 describes some characteristics of the final sample of adults who smoked cigarettes for the past 30 days from NHANES from 2003 to 2018 (n = 5,676). For example, the weighted mean of CPD was 13.21. As expected, the weighted mean of BMI was in the overweight category, and most participants engaged in non-heavy smoking ( < 20 CPD). The mean survival time was 8.00 years among those who died during the study timeframe (i.e., all-cause mortality event occurred). All-cause mortality occurred among less than 10% of the sample. The association between WC and BMI indicates that both surrogate measures are strongly correlated (see S1 File).

Table 1. Descriptive characteristics of sample from NHANES 2003-2018 (n = 5,676).

Sample Characteristics Weighted mean (SE)
Cigarettes smoked per day (CPD) 13.21 (0.245)
BMI (kg/m2) 27.97 (0.125)
Age in years (when interviewed) 42.70 (0.259)
Physical Activity (min/day) 54.58 (1.552)
Alcohol Consumption (Drinks/week) 7.86 (0.292)
SES (USD/poverty measure) 2.35 (0.041)
Mean Survival Time (years) 8.00 (0.139)
Waist Circumference (cm) 97.58 (0.314)
Android Body Fat (%) 34.21 (0.250)
Gynoid Body Fat (%) 34.79 (0.228)
All-cause Mortality n (weighted %)
Alive 5,065 (91.47)
Dead 611 (8.53)
Sex/gender n (weighted %)
Men 3,253 (54.35)
Women 2,423 (45.65)
Race/Ethnicity n (weighted %)
Hispanic (Mexican) 610 (6.10)
Hispanic (Other) 392 (4.22)
Non-Hispanic White 2,855 (70.32)
Non-Hispanic Black 1,353 (12.24)
Non-Hispanic Other 466 (7.12)
Blood Pressure n (weighted %)
Systolic  ≥  120 mm Hg and Diastolic <  80 mm Hg 2,661 (49.81)
Systolic  <  120 mm Hg and Diastolic  ≥  80 mm Hg 3,015 (50.19)
Total Cholesterol n (weighted %)
 ≥  200 mg/dL 2,323 (41.94)
 < 200 mg/dL 3,353 (58.06)
BMI Categories n (weighted %)
Underweight (BMI <  18.5) 170 (2.81)
Normal Weight (18.5  ≤  BMI <  25) 1,908 (34.30)
Overweight (25  ≤  BMI <  30) 1,739 (30.59)
Obesity (BMI  ≥  30) 1,859 (32.29)
Open in a new tab

Mediation analysis

The direct effect and total effect were statistically significant for all surrogate measures (BMI, WC, and body fat percentages), indicating that CPD was directly associated with all-cause mortality, as expected (p-values < 0.05; see Table 2). The indirect effects for all surrogate measures were not significant (see Table 2), indicating that weight status did not explain the association between cigarette smoking and all-cause mortality. However, in the mediation analysis where WC was the mediator adjusted by BMI levels, the WC’s indirect effect was statistically significant (p-value <  0.01; see Table 3 and Fig 2) and 21% of the total effect. Thus, WC may partially explain the association between the CPD and all-cause mortality per BMI levels (see Table 3 and Fig 2).

Table 2. Results of mediation analysis by surrogate measure of weight status between cigarette smoking and all-cause mortality, NHANES 2003-2018.

Effect/Surrogate Measure BMIa Waist Circumference Body Fat % (Android)b Body Fat % (Gynoid)b
Total -0.44 -0.46 -1.39 -1.46
95% CI:
(-2.24, -0.22)		 95% CI:
(-2.15, -0.21)		 95% CI:
(-4.50, -0.64)		 95% CI:
(-5.63, -0.83)
p = 0.0200* p = 0.0180* p = 0.004* p < 0.001*
Indirect -0.014 -0.07 0.004 0.10
95% CI:
(-0.006, 0.050)		 95% CI:
(-0.12, 0.00)		 95% CI:
(-0.01, 0.11)		 95% CI:
(-0.04, 0.23)
p = 0.6900 p = 0.0680c p = 0.7880 p = 0.4300
Direct -0.43 -0.38 -1.39 -1.57
95% CI:
(-2.25, -0.22)		 95% CI:
(-2.11, -0.17)		 95% CI:
(-4.58, -0.67)		 95% CI:
(-5.76, -0.91)
p = 0.0200* p = 0.0200* p = 0.004* p < 0.001*
Open in a new tab
*

p-value <  0.05

aAll-Cause Mortality-Cigarette Smoking Model included a quadratic term for BMI and the interaction of sex/gender- race/ethnicity

bSample size for Mediation analysis was 2,368 due to higher missing values for Body Fat Percentage (Android and Gynoid) and only considering NHANES 2003-2004 and NHANES 2005-2006

cp-value <  0.10

Table 3. Results of mediation analysis of WC between cigarette smoking and all-cause mortality adjusted by BMI, NHANES 2003-2018.

Effect Estimate 95% CI p-value
Total -0.45 (-2.00, -0.20) 0.0160*
Indirect -0.10 (-0.23, -0.05) <0.001*
Direct -0.35 (-1.86, -0.10) 0.0360*
Proportion Mediated 0.21 (0.07, 1.25) 0.0160*
Open in a new tab
*

p-value <  0.05

NOTE: The model for the mediation analysis included BMI as a covariate for the all-cause mortality-cigarette smoking model and the waist circumference-cigarette smoking model. For the all-cause mortality-cigarette smoking model, a quadratic term for BMI was included.

Fig 2. Visualization of mediation analysis with waist circumference.

Fig 2

Open in a new tab

NOTE: values with “*” indicate that estimates are non-linear due to coming from the AFT Model. The “a” value was the estimate effect of CPD of the Waist Circumference- Number of Cigarettes smoked per day model.

Table 3 shows how changes in WC and CPD influence survival time related to all-cause mortality among US adults who currently smoked cigarettes at the moment of interview but died throughout the follow-up. Specifically, the total effect means that for each unit increase in both CPD and WC, the survival time decreases by 0.45 years, by BMI levels and after adjusting for other covariates. The direct effect focuses only on CPD, indicating that for each unit increase in CPD, survival time decreases by 0.35 years when adjusted for other covariates. The indirect effect examines only WC and shows that for each unit increase in WC, survival time decreases by 0.10 years of time, assuming CPD doesn’t change. In summary, increasing both CPD and WC has a notable negative impact on survival time, meaning that participants were more likely to experience all-cause mortality. Fig 2 visualizes mediation effects among CPD, WC, and all-cause mortality by BMI levels. The coefficient “a” is the effect estimate of WC on CPD from the exposure-mediator model. The “a” indicates that for every additional CPD among those who currently smoke cigarettes, on average, their WC will increase by 0.046 cm. However, the mediation effects and the “a” coefficient are not linear.

Discussion

This study found that WC was a partial mediator in the association between CPD and all-cause mortality when adjusted by BMI levels at the population level. To our knowledge, this is the first study to assess whether weight status mediates the association between CPD and all-cause mortality. However, WC did not mediate the association between CPD and all-cause mortality when not adjusted by BMI. These findings have two interpretations: (a) obesity, especially central obesity (when approximated by WC), partially mediates the association between CPD and all-cause mortality when adjusted by BMI levels. (b) High levels of correlation between BMI and WC at the population level might bias the results found in this study. We encourage more research to replicate these findings.

To our knowledge, this is the first study to address whether weight status mediates the association between CPD and all-cause mortality. Still, many findings from this study echo those in the literature. Cigarette smoking has been positively and strongly associated with all-cause mortality; the effect of cigarette smoking on all-cause mortality reduced the survival time of adults who currently smoke cigarettes [31–35,70–73], where the greater the CPD, the higher the risk of all-cause mortality [32,73].

Nonetheless, this study has some strengths and limitations. This study’s greatest strength is its external validity among US adults who smoked in the past 30-days. To our knowledge, it is the first to address whether weight status mediates the association between CPD and all-cause mortality. The statistical analyses of this study were innovative and more robust than the usual mediation analyses with time-to-event data due to the application of recommended models depending on the scenario to deal with time-to-event data by Burgos-Ochoa et al. [74]. This study applied the deletion method with a high missing data rate, which may lead to biased results if the missing data do not occur entirely at random [75]. Another limitation of this study is that it only considers all-cause mortality as an outcome of death instead of specific-cause mortality such as cancer or cardiovascular diseases. Data from NHANES about cigarette smoking, demographics, physical activity, alcohol consumption, and medical history were all self-reported, which could lead to recall bias.

NHANES dataset comes from a cross-sectional study. A causal relationship cannot be inferred due to a lack of temporality. This limits the interpretations of this study only to the association between CPD and all-cause mortality. However, we can use these data, with carefulness, for mediation analysis if we have a scenario where the directions of these associations are clear (i.e., temporality can occur) [76]. In this study, we assumed participants started to smoke cigarettes before their reported surrogate measures of weight status, which we verified by comparing the age of starting to smoke cigarettes with age at interview. This scenario allowed us to see the temporality between cigarette smoking and weight status, assuming such surrogate measures were fixed until the moment of the interview. However, we did not consider changes through time in smoking behaviors or weight changes, which might contribute to the risk of all-cause mortality [77].

Doing a mediation analysis with a time-to-event outcome can apply to scenarios where exposure, mediation, or covariates can be either time-dependent or non-time dependent as done by Lange and Hansen [78], a scenario not evaluated by the work of Maxwell, Cole, and Mitchell [79]. As long as type I error is not inflated, there is a lower risk of bias when estimating a mediation effect when using cross-sectional data, though carefulness in the interpretation of results is encouraged [80]. However, given that the AFT model is known to be robust against measurement error, the type I error is likely not to be inflated, and results are less likely to be biased [81].

The findings from this study support promoting weight management programs as a harm reduction approach to reduce the risk of all-cause mortality due to past 30-day cigarette smoking. Weight management programs should be focused on lowering WC levels and, thus, central obesity through promoting more physical activity and less alcohol consumption, especially among those engaging in heavy smoking [27,82]. In case of severe obesity, bariatric surgery should also be used as a weight management intervention to lower WC levels and, therefore, to reduce all-cause mortality due to the CPD [83].

In conclusion, this study provides evidence that supports WC, when adjusted by BMI levels, is a partial mediator in the association between CPD and all-cause mortality. However, more research is needed. Future work should consider other potential sources of bias and confounders based on prior knowledge, such as changes through time in cigarette smoking and weight using longitudinal data and other substances. We also recommend future longitudinal studies to examine and evaluate potential causes of death related to smoking and weight status using other surrogate measures such as visceral fat.

Implications

Waist circumference, as a surrogate measure of weight status, was a partial mediator between cigarette smoking and all-cause mortality when adjusted by BMI levels. Reducing waist circumference and central obesity through weight loss programs may reduce mortality due to the number of cigarettes smoked per day among adults who smoke.

Supporting information

S1 File. Correlation between body mass index and waist circumference, NHANES 2003-2018.

(TIFF)

pone.0319560.s001.tiff (9.3MB, tiff)

Acknowledgments

None

Abbreviations

AFT

Accelerated Failure Time

BMI

Body Mass Index

CPD

cigarettes per day

LMF

Linkage Mortality Files

NCHS

National Center Health Statistics

NHANES

National Health and Nutrition Examination Survey

SES

Socioeconomic Status

US

United States

WC

Waist Circumference.

Data Availability

The data is publicly available through the CDC webpage (URL: https://wwwn.cdc.gov/nchs/nhanes/default.aspxt). All the used datasets were from NHANES 2003 until 2018.

Funding Statement

This manuscript was supported by the National Institute on Drug Abuse (NIDA) (Award# T32DA019426) and the National Institute on Alcohol Abuse and Alcoholism (NIAAA) under the grant: Minding A Mid-America Gap: NIDA/NIAAA Epidemiology Research Training for Clinical Researchers and Clinicians in Currently Under-Served Areas Role: Pre doctoral trainee Service. Award# R25DA051249 for Luis Miguel Mestre. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

  • 1.Cornelius ME. Tobacco product use among adults – United States, 2021. MMWR Morb Mortal Wkly Rep [Internet]. 2023. [cited 2023 Jul 16];72. Available from: https://www.cdc.gov/mmwr/volumes/72/wr/mm7218a1.htm [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Fryar CD, Caroll MD, Afful J. Products - health E stats - prevalence of overweight, obesity, and extreme obesity among adults aged 20 and over: United States, 1960–1962 through 2017–2018 [Internet]; 2021. [cited 2023 Jul 16]. Available from: https://www.cdc.gov/nchs/data/hestat/obesity-adult-17-18/obesity-adult.htm [Google Scholar]
  • 3.Fryar CD, Caroll MD, Afful J. Health E-stats december 2020 prevalence of underweight among adults aged 20 and over: United States, 1960–1962 through 2017–2018; 2020. [Google Scholar]
  • 4.Bauer U, Briss P, Goodman R, Bowman B. Prevention of chronic disease in the 21st century: elimination of the leading preventable causes of premature death and disability in the USA. Lancet. 2014;384(9937):45–52. [DOI] [PubMed] [Google Scholar]
  • 5.Rostron BL, Chang CM, Pechacek TF. Estimation of cigarette smoking-attributable morbidity in the United States. JAMA Intern Med. 2014;174(12):1922–8. doi: 10.1001/jamainternmed.2014.5219 [DOI] [PubMed] [Google Scholar]
  • 6.Chiolero A, Jacot-Sadowski I, Faeh D, Paccaud F, Cornuz J. Association of cigarettes smoked daily with obesity in a general adult population. Obesity (Silver Spring). 2007;15(5):1311–8. doi: 10.1038/oby.2007.153 [DOI] [PubMed] [Google Scholar]
  • 7.Molarius A, Seidell JC, Kuulasmaa K, Dobson AJ, Sans S. Smoking and relative body weight: an international perspective from the WHO MONICA project. J Epidemiol Community Health. 1997;51(3):252–60. doi: 10.1136/jech.51.3.252 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Istvan JA, Cunningham TW, Garfinkel L. Cigarette smoking and body weight in the cancer prevention study I. Int J Epidemiol. 1992;21(5):849–53. doi: 10.1093/ije/21.5.849 [DOI] [PubMed] [Google Scholar]
  • 9.Rásky É, Stronegger W, Freidl W. The relationship between body weight and patterns of smoking in women and men. Int J Epidemiol. 1996;25(6):1208–12. [DOI] [PubMed] [Google Scholar]
  • 10.Shimokata H, Muller DC, Andres R. Studies in the distribution of body fat. III. Effects of cigarette smoking. JAMA. 1989;261(8):1169–73. [PubMed] [Google Scholar]
  • 11.Audrain-McGovern J, Benowitz N. Cigarette smoking, nicotine, and body weight. Clin Pharmacol Ther. 2011;90(1):164–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Grunberg N. Nicotine, cigarette smoking, and body weight. Br J Addict. 1985;80(4):369–77. [DOI] [PubMed] [Google Scholar]
  • 13.Moffatt RJ, Owens SG. Cessation from cigarette smoking: changes in body weight, body composition, resting metabolism, and energy consumption. Metabolism. 1991;40(5):465–70. doi: 10.1016/0026-0495(91)90225-l [DOI] [PubMed] [Google Scholar]
  • 14.Albanes D, Jones DY, Micozzi MS, Mattson ME. Associations between smoking and body weight in the US population: analysis of NHANES II. Am J Public Health. 1987;77(4):439–44. doi: 10.2105/ajph.77.4.439 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Klesges RC, Meyers AW, Klesges LM, La Vasque ME. Smoking, body weight, and their effects on smoking behavior: a comprehensive review of the literature. Psychol Bull. 1989;106(2):204–30. doi: 10.1037/0033-2909.106.2.204 [DOI] [PubMed] [Google Scholar]
  • 16.Barrett-Connor E, Khaw KT. Cigarette smoking and increased central adiposity. Ann Intern Med. 1989;111(10):783–7. doi: 10.7326/0003-4819-111-10-783 [DOI] [PubMed] [Google Scholar]
  • 17.Clair C, Chiolero A, Faeh D, Cornuz J, Marques-Vidal P, Paccaud F, et al. Dose-dependent positive association between cigarette smoking, abdominal obesity and body fat: cross-sectional data from a population-based survey. BMC Public Health. 2011;1123. doi: 10.1186/1471-2458-11-23 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Caks T, Kos M. Body shape, body size and cigarette smoking relationships. Int J Public Health. 2009;54(1):35–9. [DOI] [PubMed] [Google Scholar]
  • 19.Wee CC, Rigotti NA, Davis RB, Phillips RS. Relationship between smoking and weight control efforts among adults in the united states. Arch Intern Med. 2001;161(4):546–50. doi: 10.1001/archinte.161.4.546 [DOI] [PubMed] [Google Scholar]
  • 20.Morris RW, Taylor AE, Fluharty ME, Bjørngaard JH, Åsvold BO, Elvestad Gabrielsen M, et al. Heavier smoking may lead to a relative increase in waist circumference: evidence for a causal relationship from a Mendelian randomisation meta-analysis. The CARTA consortium. BMJ Open. 2015;5(8):e008808. doi: 10.1136/bmjopen-2015-008808 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Chiolero A, Faeh D, Paccaud F, Cornuz J. Consequences of smoking for body weight, body fat distribution, and insulin resistance. Am J Clin Nutr. 2008;87(4):801–9. doi: 10.1093/ajcn/87.4.801 [DOI] [PubMed] [Google Scholar]
  • 22.Visser M, Launer L, Deurenberg P, Deeg D. Past and current smoking in relation to body fat distribution in older men and women. J Gerontol A Biol Sci Med Sci. 1999;54(6):M293-8. [DOI] [PubMed] [Google Scholar]
  • 23.Carreras-Torres R, Johansson M, Haycock PC, Relton CL, Davey Smith G, Brennan P, et al. Role of obesity in smoking behaviour: Mendelian randomisation study in UK Biobank. BMJ. 2018;361:k1767. doi: 10.1136/bmj.k1767 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Canoy D, Wareham N, Luben R, Welch A, Bingham S, Day N, et al. Cigarette smoking and fat distribution in 21,828 British men and women: a population-based study. Obes Res. 2005;13(8):1466–75. doi: 10.1038/oby.2005.177 [DOI] [PubMed] [Google Scholar]
  • 25.Troisi R, Heinold J, Vokonas P, Weiss S. Cigarette smoking, dietary intake, and physical activity: effects on body fat distribution—the normative aging study. Am J Clin Nutr. 1991;53(5):1104–11. [DOI] [PubMed] [Google Scholar]
  • 26.Akbartabartoori M, Lean MEJ, Hankey CR. Relationships between cigarette smoking, body size and body shape. Int J Obes (Lond). 2005;29(2):236–43. doi: 10.1038/sj.ijo.0802827 [DOI] [PubMed] [Google Scholar]
  • 27.Kim JH, Shim KW, Yoon YS, Lee SY, Kim SS, Oh SW. Cigarette smoking increases abdominal and visceral obesity but not overall fatness: an observational study. PLoS One. 2012;7(9):e45815. doi: 10.1371/journal.pone.0045815 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Oh HS, Seo WS. The compound relationship of smoking and alcohol consumption with obesity. Yonsei Med J. 2001;42(5):480–7. doi: 10.3349/ymj.2001.42.5.480 [DOI] [PubMed] [Google Scholar]
  • 29.Kaufman A, Augustson EM, Patrick H. Unraveling the relationship between smoking and weight: the role of sedentary behavior. J Obes. 2012;2012:735465. doi: 10.1155/2012/735465 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Dvorak RD, Del Gaizo AL, Engdahl RM, Eliason CJ. Tobacco use and body mass index: mediated effects through physical inactivity. J Health Psychol. 2009;14(7):919–23. doi: 10.1177/1359105309341005 [DOI] [PubMed] [Google Scholar]
  • 31.Carter BD, Abnet CC, Feskanich D, Freedman ND, Hartge P, Lewis CE. Smoking and mortality — beyond established causes. N Engl J Med. 2015;372(7):631–40. [DOI] [PubMed] [Google Scholar]
  • 32.Inoue-Choi M, Christensen CH, Rostron BL, Cosgrove CM, Reyes-Guzman C, Apelberg B, et al. Dose-response association of low-intensity and nondaily smoking with mortality in the United States. JAMA Netw Open. 2020;3(6):e206436. doi: 10.1001/jamanetworkopen.2020.6436 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Inoue-Choi M, Liao LM, Reyes-Guzman C, Hartge P, Caporaso N, Freedman ND. Association of long-term, low-intensity smoking with all-cause and cause-specific mortality in the national institutes of health-AARP diet and health study. JAMA Intern Med. 2017;177(1):87–95. doi: 10.1001/jamainternmed.2016.7511 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Peto R. Smoking and death: the past 40 years and the next 40. BMJ. 1994;309(6959):937–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Thun MJ, Day-Lally CA, Calle EE, Flanders WD, Heath CW Jr. Excess mortality among cigarette smokers: changes in a 20-year interval. Am J Public Health. 1995;85(9):1223–30. doi: 10.2105/ajph.85.9.1223 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Lariscy J, Hummer R, Rogers R. Cigarette smoking and all-cause and cause-specific adult mortality in the United States. Demography. 2018;55(5):1855–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Thomson B, Emberson J, Lacey B, Lewington S, Peto R, Jemal A, et al. Association between smoking, smoking cessation, and mortality by race, ethnicity, and sex among US adults. JAMA Netw Open. 2022;5(10):e2231480. doi: 10.1001/jamanetworkopen.2022.31480 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Jha P, Ramasundarahettige C, Landsman V, Rostron B, Thun M, Anderson R. 21st-century hazards of smoking and benefits of cessation in the United States. N Engl J Med. 2013;368(4):341–50. [DOI] [PubMed] [Google Scholar]
  • 39.Cohen SS, Signorello LB, Cope EL, McLaughlin JK, Hargreaves MK, Zheng W, et al. Obesity and all-cause mortality among black adults and white adults. Am J Epidemiol. 2012;176(5):431–42. doi: 10.1093/aje/kws032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Camargo CJ, Weiss S, Zhang S, Willett W, Speizer F. Prospective study of body mass index, weight change, and risk of adult-onset asthma in women. Arch Intern Med. 1999;159(21):2582–8. [DOI] [PubMed] [Google Scholar]
  • 41.Abramowitz MK, Hall CB, Amodu A, Sharma D, Androga L, Hawkins M. Muscle mass, BMI, and mortality among adults in the United States: a population-based cohort study. PLoS One. 2018;13(4):e0194697. doi: 10.1371/journal.pone.0194697 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Stokes A, Preston SH. How smoking affects the proportion of deaths attributable to obesity: assessing the role of relative risks and weight distributions. BMJ Open. 2016;6(2):e009232. doi: 10.1136/bmjopen-2015-009232 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Flegal K, Kit B, Orpana H, Graubard B. Association of all-cause mortality with overweight and obesity using standard body mass index categories: a systematic review and meta-analysis. JAMA. 2013;309(1):71–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Flegal KM, Graubard BI, Williamson DF, Gail MH. Excess deaths associated with underweight, overweight, and obesity. JAMA. 2005;293(15):1861–7. doi: 10.1001/jama.293.15.1861 [DOI] [PubMed] [Google Scholar]
  • 45.Zhang WB, Pincus Z. Predicting all-cause mortality from basic physiology in the Framingham heart study. Aging Cell. 2016;15(1):39–48. doi: 10.1111/acel.12408 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Aune D, Sen A, Prasad M, Norat T, Janszky I, Tonstad S. BMI and all cause mortality: systematic review and non-linear dose-response meta-analysis of 230 cohort studies with 3.74 million deaths among 30.3 million participants. BMJ. 2016:i2156. doi: 10.1136/bmj.i2156 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Angelantonio E, Bhupathiraju S, Wormser D, Gao P, Kaptoge S, Gonzalez AB de. Body-mass index and all-cause mortality: individual-participant-data meta-analysis of 239 prospective studies in four continents. Lancet. 2016;388(10046):776–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Huai P, Liu J, Ye X, Li W-Q. Association of central obesity with all cause and cause-specific mortality in US adults: a prospective cohort study. Front Cardiovasc Med. 2022;9:816144. doi: 10.3389/fcvm.2022.816144 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Visaria A, Setoguchi S. Body mass index and all-cause mortality in a 21st century U.S. population: a national health interview survey analysis. PLoS One. 2023;18(7):e0287218. doi: 10.1371/journal.pone.0287218 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Collaboration P. Body-mass index and cause-specific mortality in 900 000 adults: collaborative analyses of 57 prospective studies. Lancet. 2009;373(9669):1083–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Pischon T, Boeing H, Hoffmann K, Bergmann M, Schulze MB, Overvad K, et al. General and abdominal adiposity and risk of death in Europe. N Engl J Med. 2008;359(20):2105–20. doi: 10.1056/NEJMoa0801891 [DOI] [PubMed] [Google Scholar]
  • 52.Carmienke S, Freitag MH, Pischon T, Schlattmann P, Fankhaenel T, Goebel H, et al. General and abdominal obesity parameters and their combination in relation to mortality: a systematic review and meta-regression analysis. Eur J Clin Nutr. 2013;67(6):573–85. doi: 10.1038/ejcn.2013.61 [DOI] [PubMed] [Google Scholar]
  • 53.Jacobs E, Newton C, Wang Y, Patel A, McCullough M, Campbell P. Waist circumference and all-cause mortality in a large US cohort. Arch Intern Med. 2010;170(15):1293–301. [DOI] [PubMed] [Google Scholar]
  • 54.Mestre L, Lartey S, Ejima K, Mehta T, Keith S, Maki K. Body mass index, obesity, and mortality– part I: associations and limitations. Nutr Today. 2023;58(3):92. [Google Scholar]
  • 55.Jayedi A, Soltani S, Zargar M, Khan T, Shab-Bidar S. Central fatness and risk of all cause mortality: systematic review and dose-response meta-analysis of 72 prospective cohort studies. BMJ. 2020;370:m3324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Klein S, Allison D, Heymsfield S, Kelley D, Leibel R, Nonas C, et al. Waist circumference and cardiometabolic risk: a consensus statement from shaping America’s health: association for weight management and obesity prevention; NAASO, the obesity society; the American Society for Nutrition; and the American Diabetes Association. Diabetes Care. 2007;30(6):1647–52. [DOI] [PubMed] [Google Scholar]
  • 57.Garrison RJ, Feinleib M, Castelli WP, McNamara PM. Cigarette smoking as a confounder of the relationship between relative weight and long-term mortality. The Framingham heart study. JAMA. 1983;249(16):2199–203. doi: 10.1001/jama.1983.03330400045023 [DOI] [PubMed] [Google Scholar]
  • 58.Sempos C, Durazo-Arvizu R, McGee D, Cooper R, Prewitt T. The influence of cigarette smoking on the association between body weight and mortality. the Framingham heart study revisited. Ann Epidemiol. 1998;8(5):289–300. [DOI] [PubMed] [Google Scholar]
  • 59.Luijckx E, Lohse T, Faeh D, Rohrmann S. Joint effects of BMI and smoking on mortality of all-causes, CVD, and cancer. Cancer Causes Control. 2019;30(5):549–57. [DOI] [PubMed] [Google Scholar]
  • 60.National Center for Health Statistics. The linkage of national center for health statistics survey data to the national death index – 2019 Linked Mortality File (LMF): methodology overview and analytic considerations [Internet]. Hyattsville, Maryland; 2022. [cited 2023 Apr 1]. Available from: https://www.cdc.gov/nchs/data-linkage/mortality-methods.htm [Google Scholar]
  • 61.CDC. 2022 Interviewer procedures manual – October 2022; 2022. [Google Scholar]
  • 62.Flegal K, Shepherd J, Looker A, Graubard B, Borrud L, Ogden C. Comparisons of percentage body fat, body mass index, waist circumference, and waist-stature ratio in adults. Am J Clin Nutr. 2009;89(2):500–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Choi S, Stommel M, Ling J, Noonan D, Chung J. The impact of smoking and multiple health behaviors on all-cause mortality. Behav Med. 2022;48(1):10–7. [DOI] [PubMed] [Google Scholar]
  • 64.Yun M, Li S, Sun D, Ge S, Lai C-C, Fernandez C, et al. Tobacco smoking strengthens the association of elevated blood pressure with arterial stiffness: the Bogalusa Heart Study. J Hypertens. 2015;33(2):266–74. doi: 10.1097/HJH.0000000000000410 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Chi D, Masterson E, Carle A, Mancl L, Coldwell S. Socioeconomic status, food security, and dental caries in US children: mediation analyses of data from the National Health and Nutrition Examination Survey, 2007–2008. Am J Public Health. 2014;104(5):860–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Abramson JL, Vaccarino V. Relationship between physical activity and inflammation among apparently healthy middle-aged and older US adults. Arch Intern Med. 2002;162(11):1286–92. doi: 10.1001/archinte.162.11.1286 [DOI] [PubMed] [Google Scholar]
  • 67.CDC. NHANES survey methods and analytic guidelines [Internet]; 2017. [cited 2023 Jul 17]. Available from:https://wwwn.cdc.gov/nchs/nhanes/analyticguidelines.aspx#analytic-guidelines [Google Scholar]
  • 68.Bigaard J, Tjønneland A, Thomsen BL, Overvad K, Heitmann BL, Sørensen TIA. Waist circumference, BMI, smoking, and mortality in middle-aged men and women. Obes Res. 2003;11(7):895–903. doi: 10.1038/oby.2003.123 [DOI] [PubMed] [Google Scholar]
  • 69.Cerhan J, Moore S, Jacobs E, Kitahara C, Rosenberg P, Adami H. A pooled analysis of waist circumference and mortality in 650,000 adults. Mayo Clin Proc. 2014;89(3):335–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70.Christensen CH, Rostron B, Cosgrove C, Altekruse SF, Hartman AM, Gibson JT, et al. Association of cigarette, cigar, and pipe use with mortality risk in the US population. JAMA Intern Med. 2018;178(4):469–76. doi: 10.1001/jamainternmed.2017.8625 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 71.Hummer RA, Nam CB, Rogers RG. Adult mortality differentials associated with cigarette smoking in the USA. Popul Res Policy Rev. 1998;17(3):285–304. [Google Scholar]
  • 72.Preston S, Glei D, Wilmoth J. A new method for estimating smoking-attributable mortality in high-income countries. Int J Epidemiol. 2010;39(2):430–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73.Weir JM, Dunn JE Jr. Smoking and mortality: a prospective study. Cancer. 1970;25(1):105–12. doi: [DOI] [PubMed] [Google Scholar]
  • 74.Burgos Ochoa L, Rijnhart JJM, Penninx BW, Wardenaar KJ, Twisk JWR, Heymans MW. Performance of methods to conduct mediation analysis with time‐to‐event outcomes. Stat Neerl. 2019;74(1):72–91. doi: 10.1111/stan.12191 [DOI] [Google Scholar]
  • 75.García-Laencina PJ, Sancho-Gómez JL, Figueiras-Vidal AR. Pattern classification with missing data: a review. Neural Comput Appl. 2010;19(2):263–82. [Google Scholar]
  • 76.Shrout PE. Commentary: mediation analysis, causal process, and cross-sectional data. Multivar Behav Res. 2011;46(5):852–60. [DOI] [PubMed] [Google Scholar]
  • 77.Smith DL Jr, Yang Y, Mestre LM, Henschel B, Parker E, Dickinson S, et al. Impact of sustained calorie restriction and weight cycling on body composition in high-fat diet-fed male and female C57BL/6J mice. Obesity (Silver Spring). 2024;32(5):959–68. doi: 10.1002/oby.24015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 78.Lange T, Hansen JV. Direct and indirect effects in a survival context. Epidemiology. 2011;22(4):575–81. doi: 10.1097/EDE.0b013e31821c680c [DOI] [PubMed] [Google Scholar]
  • 89.Maxwell S, Cole D, Mitchell M. Bias in cross-sectional analyses of longitudinal mediation: Partial and complete mediation under an autoregressive model. Multivar Behav Res. 2011;46(5):816–41. [DOI] [PubMed] [Google Scholar]
  • 80.Cain M, Zhang Z, Bergeman C. Time and other considerations in mediation design. Educ Psychol Meas. 2018;78(6):952–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 81.Oh EJ, Shepherd BE, Lumley T, Shaw PA. Considerations for analysis of time-to-event outcomes measured with error: Bias and correction with SIMEX. Stat Med. 2018;37(8):1276–89. doi: 10.1002/sim.7554 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 82.Dare S, Mackay DF, Pell JP. Relationship between smoking and obesity: a cross-sectional study of 499,504 middle-aged adults in the UK general population. PLoS One. 2015;10(4):e0123579. doi: 10.1371/journal.pone.0123579 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 83.Chow A, Neville A, Kolozsvari N. Smoking in bariatric surgery: a systematic review. Surg Endosc. 2021;35(6):3047–66. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 File. Correlation between body mass index and waist circumference, NHANES 2003-2018.

(TIFF)

pone.0319560.s001.tiff (9.3MB, tiff)

Data Availability Statement

The data is publicly available through the CDC webpage (URL: https://wwwn.cdc.gov/nchs/nhanes/default.aspxt). All the used datasets were from NHANES 2003 until 2018.

Articles from PLOS One are provided here courtesy of PLOS

RESOURCES