An Exploratory Analysis of Patterns of Weight Change in a Post-Cessation Weight Management Trial

Abstract

Introduction

Previous research investigating trajectories of weight change after quitting smoking has been limited due to infrequent weight measurements; newer e-scales data facilitate more frequent weight measurement. We explored the rate of weight gain and when weight change stabilized via secondary data analyses from the post-cessation weight management-focused “Fit & Quit” trial.

Aims and Methods

After completion of one of three randomized weight intervention conditions, all participants located in the United States completed a behavioral smoking cessation intervention with varenicline pharmacotherapy in 2018–2022. Participants were asked to weigh daily on study-provided e-scales throughout the 12-month study. Mixed-effects models were performed to characterize weight change over time and to test whether the intervention condition or gender impacted the temporal rate of weight change (n = 132).

Results

Rate of weight change in the first month after quitting smoking was most rapid, at + 0.0111 kg/day. To identify when weight stabilized, we use a fifth-degree polynomial model and found that participants began gaining weight immediately after quitting, with the rate of gain peaking around the 21st day. Weight gain continued at a lower, but still statistically significant, rate until approximately 70 days after cessation, after which the weight began to plateau. All intervention condition groups had similar rates of weight gain, with no gender differences in the rate of weight gain.

Conclusions

The first month after quitting smoking was a particularly vulnerable time period for weight gain. Future programs combining weight management and smoking cessation should emphasize maintenance of weight management strategies within the first month after quitting smoking.

Implications

Future post-cessation weight management programs should prioritize maintenance of weight management strategies within the first month after quitting smoking, regardless of the type of weight management strategy or the gender of the participants. Studies examining post-cessation weight gain should ensure measurement of weight gain for at least 2–3 months after quitting smoking.

Trial Registration

The trial is registered on clinicaltrials.gov (NCT03156660).

Introduction

After quitting smoking, people gain 4–5 kg of weight on average over the course of the first year, with most of this gain occurring within the first 3 months.¹ Although quitting smoking has many positive effects on health,^1–6 weight gain after quitting smoking can also confer health risks.^7,8 Individuals who smoke also cite weight gain after quitting smoking as a concern,^9–11 and a greater concern about weight gain predicts less success in quitting smoking.¹⁰

It is clear that weight gain typically occurs after smoking cessation¹; however, less is known about the trajectory of that weight gain, largely due to methodological limitations of previous studies on post-cessation weight gain. Previous trials investigating weight change after smoking cessation assessed weight using mostly in-person weigh-ins,¹ which limited the frequency of the weight measurement due to participant and staff burden. Data from electronic scales (or “e-scales”), which send weights directly back to research servers via the cellular network or Bluetooth connections,¹² allow for more frequent capture of objective weight measurements and have brought new insights to the field of weight management. For example, by modeling the trajectory of weights using e-scale data, Ross et al. discovered a three-phase weight loss model of weight loss (phase 1) followed by immediate weight gain (phase 2) followed by slower weight gain (phase 3), which had not been captured before using previous methods.¹³ To our knowledge, fine-grain detail regarding trends and rates of weight change over time among participants taking part in a smoking cessation intervention has yet to be explored. Furthermore, these weight change patterns have not yet been explored among participants in a post-cessation weight management intervention. Thus, using daily e-scales data collected throughout a randomized controlled trial designed to evaluate promising interventions for addressing post-cessation weight gain (the “Fit & Quit” trial),¹⁴ we aim to explore trends and rates of weight change in individuals engaged in smoking cessation.

In the following secondary analyses, we aim to answer these subsequent questions. First, are the quit dates of smoking followed temporally by an increased rate of weight change? Second, if smoking cessation is followed by an increased rate of weight change, what is the rate of weight change, and when does the weight change potentially stop? In terms of the first question, we hypothesized that the quit dates of smoking would be followed temporally by an increased rate of weight change due to the various methods by which nicotine and cigarette smoking are hypothesized to reduce body weight.¹⁵ Additionally, a previous study found that during times of smoking abstinence, weight was gained at a higher rate than during times of relapse, supporting this hypothesis.¹⁶ Finally, we also proposed to explore whether the three different weight management intervention conditions included in the Fit and Quit trial (ie, “Stability,” “Loss,” “Bibliotherapy”)¹⁴ experienced different temporal rates of weight change. The answers to these questions can be used in future smoking cessation/weight management trial designs to determine at which time points and for which study participants additional weight management support may be most warranted.

Methods

“Fit and Quit” Trial Overview

The data used for this secondary analysis exploratory paper are from the “Fit & Quit” randomized controlled trial.^14,17 Overall, the trial was successful in decreasing or eliminating the 4–5 kg weight gain typically observed post-cessation,¹ depending on the randomized condition. Briefly, participants in the Stability condition had a mean weight loss of 0.24 kg (SD = .25), participants in the Loss condition had a mean weight loss of 1.11 kg (SD = .20), and participants in the Bibliotherapy group had a mean weight gain of 1.25 kg (SD = .27) from baseline to Month 12.¹⁴ Weight change for the Loss condition was significantly different from the Bibliotherapy (comparison) condition, but weight change for the Stability condition was not significantly different from the Bibliotherapy condition. However, important to these analyses, participants, on average, lost weight between baseline and Month 2, with the largest weight losses in the Loss condition; thus, some of the weight change observed was weight regain and some was weight gain above their baseline weight (particularly for participants in the Bibliotherapy condition, but for participants in any condition with less success in managing their weight). As previously reported in a secondary analysis, the participants in the Stability condition had a significantly greater self-weighing frequency compared to participants in the Loss and Bibliotherapy conditions across all 12 months.¹⁸ There were no significant differences between the conditions related to self-reported or biochemically verified smoking cessation at Month 12.¹⁴

Trial participants (n = 305 total) were randomized (blocked by BMI and gender) by the study statistician and allocated into one of three weight intervention conditions, as described below. All participants received the same smoking cessation intervention (ie, Chantix (varenicline) pharmacotherapy and a 7-session behavioral intervention). The Institutional Review Board of the University of Tennessee Health Science Center approved this study.

Inclusion and Exclusion Criteria

The full study protocol¹⁷ and main outcomes¹⁴ of the “Fit & Quit” trial have been published previously. Briefly, participants needed to be adults who smoked cigarettes who wanted to quit smoking, who could participate in the study (eg, able to exercise, able to use a phone and email), and who could safely lose weight (ie, a BMI of 22 kg/m² or greater). Potential participants were excluded if they were unable to take Chantix, weighed over 385 pounds (due to limitations of the study-provided e-scale), were pregnant or breastfeeding, had recently lost weight or participated in another smoking and/or weight loss intervention, or had another physical or mental health condition that would preclude participation in the trial.

Recruitment

Advertising for the study included electronic and traditional methods in addition to a “refer a friend” program. Prospective trial participants underwent an initial phone screening to determine eligibility for an in-person screening visit (and virtually once the COVID-19 pandemic began), where informed consent was obtained and the full determination of participant eligibility was determined. After the in-person screening, interested individuals received clearance from a physician and tracked their exercise and diet for three days. After filling each of the 10 waves, participants were randomized, provided with an e-scale along with other supplies specific to their intervention at the randomization visit, and met individually with their interventionists (in-person, then virtually once the COVID-19 pandemic began).

Interventions

Briefly, all participants were encouraged to use the study-provided e-scale to self-weigh each day. Based on previous evidence that indicated the superiority of sequential compared to concurrent weight management and smoking cessation inteventions,¹⁹ the trial was designed to focus on weight management in the first eight weeks and then transition to smoking cessation. Participants in the Stability intervention (based on the protocol from the Small Changes intervention)²⁰ had the goal of preventing an increase in weight by introducing small changes in steps and diet, with personalized interventionist feedback on their weight trajectory, their small dietary changes, and their number of steps. Participants in the Stability condition were also mailed a small green-colored prize weekly (eg, postage stamp, bookmark) if their weight remained in the “green zone” (defined as <1.4 kg above their baseline weight). Participants in the Loss intervention (which was based upon the “Look AHEAD” “Intensive Lifestyle Intervention”)²¹ had the goal of a ≥5% decrease in weight from baseline through exercise and diet modifications by week eight, facilitated by meal replacements and diet and exercise self-monitoring with personalized interventionist feedback related to their weight trajectory and calorie and exercise goal attainment. During the first eight weeks, Stability and Loss participants had weekly 60-minute phone sessions with their respective groups, which, once the smoking cessation intervention had been completed, transitioned into monthly booster sessions (five total). Participants in the Bibliotherapy intervention were given a book titled “The Eating Well Diet.”²² Participants in the Bibliotherapy intervention waited the same amount of time before starting the intervention for smoking cessation and did not receive weight management phone sessions or weight-related feedback during the trial.

The interventions for smoking cessation in this study followed the recommendations of the Clinical Practice Guidelines²³ and have been described previously.¹⁷ Briefly, at the beginning of week nine, the participants in all conditions began receiving group-based behavioral smoking intervention via phone sessions (60 minutes each for six weeks total) that, for the Stability and Loss groups, took place with the same cohort as their weight management group phone sessions. During the week that participants were recommended to quit smoking (ie, week eleven), all participants had an individual phone session and they started taking Chantix. All participants in this subsample received at least one box of Chantix at the start of the smoking cessation period and were encouraged to continue taking Chantix for 6 months, regardless of quit status. Research staff evaluated medication availability by phone each month and sent participants another box of Chantix each month, as needed. Medication disbursement was highly consistent with the protocol (mean±standard deviation = 4.687±1.359 of boxes received (out of a possible 6 boxes).

Measures

The weight data used was obtained from study-provided BodyTrace e-scales, which transmitted weights to the study database via the cellular network, in kilograms, with stated accuracy within 0.1 kg.²⁴ E-scale weights have been shown to be highly concordant with in-person measured weights.^25,26 Participants were asked to weigh daily at the same time of day, ideally first thing in the morning, without wearing clothing, before eating or drinking, and after using the toilet.¹² Among participants who reported point prevalence abstinence (ie, if they responded “no” to the question regarding whether they “had smoked a cigarette, even just a puff, in the past 7 days”), the quit date(s) (ie, the date of smoking cessation, which sometimes was attempted on multiple dates, with all attempts recorded) were obtained at assessment visits at Month 2, 4, 8, and 12. The quit date used for this study was the first quit date for each participant. Among participants who reported that they had quit smoking at follow-up visits, smoking status was verified by measuring salivary cotinine levels at follow-up visits (using NicAlert^TM prior to the COVID-19 pandemic and iScreen^TM after the pandemic started). Biochemically verified smoking abstinence rates had a high level of concordance with the self-reported results in previous analyses¹⁴; thus, we used self-reported abstinence in these analyses. Participants self-reported their race, ethnicity, education level, age, and gender at baseline.

Statistical Analysis

Weight Data Exclusion Protocol

Before conducting the analyses, all weights <20 kg and outliers (often indicating that another person or a pet stepped on the scale) of weight data were removed. If more than one weight was recorded in a single day, the first weight of the day was chosen.²⁵ Outliers of weights 2.27 kg above the expected weight were found via the use of a generalized additive model. In congruence with previous research,¹³ of these outliers were investigated by the research team and if they were found to be inconsistent with the previous 2 days of self-weighing and the next 2 days of self-weighing, they were excluded.

Analysis

A linear mixed effect model was performed to characterize weight change over time, in a similar fashion as in Prod’hom et al..¹⁶ This model was adjusted for age, gender, and intervention condition as follows:

$${\displaystyle \begin{array}{l}{\displaystyle \begin{array}{lllll}{W}_{it}& =& {\beta }_0+{\beta }_1\cdot time\mathrm{\_}since\mathrm{\_}quittin{g}_{it}+{\beta }_2\cdot Ag{e}_i+{\beta }_3\cdot Gende{r}_i& +& {\beta }_4\cdot Grou{p}_i+b_{0i}+b_{1i}\cdot time\mathrm{\_}sinc{e}_{\mathrm{\_}}quittin{g}_{it}+{ϵ}_{it}\end{array}}\end{array}}$$

Where W_it denotes the weight of i^th individual at time point t; time_since_quitting^it is the number of days after quitting smoking of i^th individual at time point t; Age_i, Gender_i, and Group_i denote the age, gender, and intervention condition of i^th individual, respectively; β_0,…,β₄ are fixed effects; $b_{0i}\sim N(0,{\sigma }_{b_0}^2)$ and $b_{1i}\sim N(0,{\sigma }_{b_1}^2)$ are random effects to capture variation in intercepts and slopes of time_since_quitting for different individuals; ${ϵ}_{it}\sim N(0,{\sigma }_e^2)$ is the random error that captures variation in the outcome. To test whether the intervention condition or gender impacted the temporal rate of weight change, we examined the linear mixed model with the related interaction term. Additionally, to detect the point where weight change increases, the interval where weight change is significantly positive, and the time at which weight change stabilized, we modeled the weight change as a polynomial function of time since quitting. Furthermore, for all analyses in this study, a linear interpolation approach was used to impute missing daily weight values.

Results

From the 305 participants in the trial overall, data for 291 participants remained after removing participants whose self-weighing data did not match the outcome data, indicating that most likely another person was regularly using the e-scale. Of those, 188 participants (64%) had a smoking quit date. However, 45 participants who quit smoking before the suggested quit date were excluded since their weight management period overlapped with their quit date, and 11 participants were removed due to having fewer than 5 weight observations. Thus, the sample size for these analyses was 132 participants. Baseline characteristics for these 132 participants can be seen in Table 1. Data from 49 participants in the Stability group, 45 from the Loss group, and 38 from the Bibliotherapy group were included in the study analyses. The number of weight observations throughout the study decreased over time (Figure S1).

Table 1.

Baseline Demographic Characteristics of Included Participants (n = 132)

	Stability condition (n = 49)	Loss condition (n = 45)	Bibliotherapy condition (n = 38)	Total (n = 132)	*p-value
Gender, N (%)					.03
Women	39 (29.5)	25 (18.9)	29 (22.0)	93 (70.5)
Men	10 (7.6)	20 (15.2)	9 (6.8)	39 (29.5)
Age, N (%)					.59
18–39 years	7 (5.3)	7 (5.3)	4 (3.0)	18 (13.6)
40–49 years	9 (6.8)	7 (5.3)	4 (3.0)	20 (15.1)
50–59 years	16 (12.1)	9 (6.8)	14 (10.6)	39 (29.5)
60 years and older	17 (12.9)	22 (16.7)	16 (12.2)	55 (41.8)
Race, N (%)					.16
Black/ African American	24 (18.2)	12 (9.1)	11 (8.3)	47 (35.6)
More Than One Race	-	-	1 (0.8)	1 (0.8)
Native Hawaiian/ Pacific Islander	1 (0.8)	-	-	1 (0.8)
White	23 (17.4)	33 (25.0)	25 (18.9)	81 (61.4)
Unknown	1 (0.8)	-	1 (0.8)	2 (1.5)
Education, N (%)					.93
High school diploma or less	24 (18.1)	21 (16)	19 (14.4)	64 (48.5)
Associate degree	9 (6.8)	6 (4.5)	8 (6.1)	23 (17.4)
4-year college degree or more	16 (12.1)	18 (13.7)	11 (8.3)	45 (34.1)
Relationship status, N (%)					.06
Married	17 (12.9)	25 (18.9)	16 (12.2)	58 (44)
(Single/divorced/widowed)	32 (24.2)	20 (15.1)	22 (16.6)	74 (56)
Income, N (%)					.28
<$30 000	21 (15.9)	10 (7.6)	12 (9.1)	43 (32.6)
$30 000–<$50 000	13 (9.8)	10 (7.6)	8 (6.1)	31 (23.5)
$50 000–<$100 000	9 (6.8)	20 (15.2)	12 (9.1)	41 (31.1)
≥$100 000	5 (3.8)	5 (3.8)	5 (3.8)	15 (11.4)
Unknown	1 (0.8)	-	1 (0.8)	2 (1.5)
Cigarettes per day, M (SD)					-
Average (SD)	16.6 (14.1)	17.8 (7.6)	16.2 (8.3)	16.9 (10.6)
Fagerstrom test of nicotine dependence, N (%)					.97
Low dependence	4 (3)	5 (3.8)	2 (1.5)	11 (8.3)
Low to moderate dependence	16 (12.2)	14 (10.6)	13 (9.9)	43 (32.7)
Moderate dependence	23 (17.4)	23 (17.4)	21 (15.9)	67 (50.7)
High dependence	6 (4.5)	3 (2.3)	2 (1.5)	11 (8.3)

^*The p-value is calculated by chi-square tests.

The results of the mixed-effects model indicated that the coefficient for time since quitting smoking was 0.0032 kg/day (SE = 0.002; around 0.0036% from baseline weight per day) over the entire study period and was not significant (p = .094), suggesting a positive, but nonsignificant rate of weight change over the entire duration of the study. However, when shorter durations immediately following the quit date were considered, the rate of weight gain was higher and statistically significant (Table 2). Results indicated that the weight gain rate was highest immediately after quitting smoking and gradually decreased over time. Specifically, the rate of weight change in the first month after smoking cessation was + 0.0111 kg/day (0.0123% from baseline weight per day). In the first 2 months, the rate was + 0.0107 kg/day (0.0119% per day). This pattern continued until 8 months after quitting smoking, when the rate of weight change was no longer significant. (To provide a comparison, we have included Table S1, which shows the slight weight loss trajectory for the participants who did not report quitting smoking at any of the data collection visits.) We have also provided Table S2, which describes the weight change trajectory in the subset of the sample (n = 32) with biochemically verified abstinence at all of the follow-up data collection points.

Table 2.

Weight Gain Rate (kg/day) for Periods Ranging From One to Eight Months After the Quit Date

Weight gain rate	Estimate (kg/day)	Standard error	p-value
1 month post-quit	0.0111	0.0037	.003
2 months post-quit	0.0107	0.0030	<.001
3 months post-quit	0.0088	0.0026	<.001
4 months post-quit	0.0066	0.0023	.004
5 months post-quit	0.0052	0.0020	.012
6 months post-quit	0.0042	0.0019	.028
7 months post-quit	0.0038	0.0018	.036
8 months post-quit	0.0032	0.0017	.058

To identify when weight stabilizes, we used a polynomial model in which weight change was modeled as a fifth-degree function of time since quitting. This model was chosen due to the observed nonlinear pattern of weight change over time, allowing us to capture key fluctuations and inflection points that lower-degree polynomials might overlook. Additionally, the fifth-degree term was the highest statistically significant term, confirming that higher-order effects enhance the model’s accuracy in describing the stabilization phase.^27,28  Figure 1 illustrates the slope of weight change over time following smoking cessation, with participants gaining weight immediately after quitting, and the rate of weight change peaking around the 21st day. The slope of weight gain continued at a lower, but still significant, rate until approximately 70 days after cessation, after which point weight change began to plateau.

Figure 1.

Trajectory of the slope of weight change following the quit date.

Table 3 presents the results of two mixed-effects models examining weight changes after smoking cessation. The first model investigated interactions between time since quitting and intervention conditions, while the second model explored interactions between time since quitting and gender. The results show that neither intervention condition nor gender significantly moderated the observed associations between time since quitting and weight change. Specifically, although the rate of weight change for both the Loss and Stability conditions were slightly higher than that of the Bibliotherapy condition, these differences were not statistically significant (Figure S2). This figure also shows that the rate of weight gain was steeper in the first weeks after quitting for all three intervention groups (particularly steeper for the Stability and Loss conditions compared to the Bibliotherapy condition). The overall weight gain stabilized at slightly above 1% for the Stability and Loss conditions, and slightly below 1% for the Bibliotherapy condition.

Table 3.

Mixed-Effects Models Examining Interactions Between Time Since Quitting and Intervention Condition (Left Panel) and Interactions Between Time Since Quitting and Gender (Right Panel) on Weight (kg)

Variables	Interactions between time since quitting and intervention condition			Interactions between time since quitting and gender
	Estimate (kg)	Standard error	p-value	Estimate (kg)	Standard error	p-value
Intercept	81.3090	6.0837	<.001	81.2781	4.2565	<.001
Time since quitting	0.0015	0.0028	.6014	0.0035	0.0018	.0501
Gender (men)	8.4563	2.6515	.0016	8.8166	1.8645	<.001
Age	0.0180	0.1015	.8594	0.0158	0.0711	.8238
Condition (loss)	−2.6566	3.0395	.3829	−2.5962	2.1108	.2192
Condition (stability)	1.7808	2.9443	.5458	1.8331	2.0443	.3702
Time since quitting* condition (loss)	0.0020	0.0038	.5984
Time since quitting* condition (stability)	0.0010	0.0038	.7902
Time since quitting* gender (men)				−0.0034	0.0033	.3104

Discussion

Principal Results and Comparison With Prior Work

Overall, we found a pattern of rapid initial weight gain in the first 21 days after quitting smoking, with weight change slowing somewhat but still occurring at a statistically significant rate until about day 70, after which point weight began to plateau. Specifically, the rate of weight change after quitting was approximately + 0.0111 kg/day (0.0123% from baseline weight per day) within the first month after quitting smoking, compared to + 0.0032 kg/day (0.0036% per day) over the entire 8-month follow-up period. These findings indicate that the first month after quitting smoking may represent the most vulnerable time period, which may warrant additional monitoring and support in post-cessation weight management programs.

Our estimated rates of weight gain after quitting smoking are slightly lower than rates found in previous research, which may partially be due to the effectiveness of the weight management interventions in the “Fit & Quit” trial. Specifically, Prod’hom et al. found an estimated rate of weight gain of 0.020 kg/day in women and 0.025 kg/day in men during periods of smoking abstinence.¹⁶ Also, in contrast to the results reported by Prod’hom et al.,¹⁶ we did not find evidence of a gender difference in weight gain after quitting smoking in our model. Concerns about weight gain after quitting smoking have traditionally been found to be more common among women^10,29,30 (and this was also found in a previous study using the current sample²⁷). However, research has also found that weight gain in men post-smoking cessation is associated with a greater risk of smoking relapse.²⁸ Thus, results of the current study indicate that both men and women may be similarly vulnerable to post-cessation weight gain and, taken together with results of previous studies,^27,28 our results suggest that both men and women should be included in future post-cessation weight gain programs.

Interestingly, there were no differences in the rate of weight change during the post-cessation time period among the three intervention conditions. This is unexpected as one may hypothesize that the Stability and Loss groups would experience weight change at a slower rate compared to the Bibliotherapy intervention condition, given that participants in the Stability and Loss groups received a more intensive exposure to weight management skills. Indeed, the Bibliotherapy intervention was found to be significantly less efficacious at preventing weight gain as compared to the Loss intervention.¹⁴ However, the participants in the Bibliotherapy condition had less weight gain at 12-month follow-up (ie, 1.25 kg),¹⁴ compared to previous trials that have had no weight management intervention (4–5 kg).¹ Alternatively, one could have hypothesized that the Loss participants would experience more weight change, given that they have just completed a weight loss program over the prior 2 months, and they are experiencing weight regain. Thus, these results indicate that participants in none of the three weight management programs were at a particular advantage or disadvantage in eliminating the small amounts of weight gain experienced immediately after quitting smoking.

Strengths

The usage of e-scales provided an opportunity to analyze weight trends at a granular level, allowing for greater insight into weight trends after quitting smoking. Additionally, because of the usage of e-scales, the study was able to continue uninterrupted even once the COVID-19 pandemic began. There was significant diversity of race and gender in this clinical trial,¹⁴ increasing external validity. Comparing trends of smokers in three different weight management intervention groups allowed us to see if there were differences in trends that may be due to the weight management intervention. Our utilization of different statistical models in our analysis provides multiple analytical perspectives and increases data validity. Overall, this exploratory analysis adds greater detail about weight trends after quitting smoking to the existing body of literature and provides new insights into weight trends in smokers who quit smoking while undergoing a weight management intervention.

Limitations

One limitation of this study is the exclusion of data from participants who either never reported quitting smoking at any of the data collection visits or who self-weighed fewer than five times. We did not conduct a subgroup analysis of excluded participants, so it is unknown whether these participants had certain characteristics making them less likely to quit smoking or self-weigh. Additionally, some participants were removed from this subsample for having e-scale data that included someone other than the participant using the e-scale, which decreased the sample size for the current analyses. Furthermore, most participants did not self-weigh daily as instructed, diminishing the resolution of the data available to analyze. In these analyses, we assumed that participants did not relapse after their quit date, since only nine participants who met the inclusion criteria for these secondary analyses relapsed at any point during the study. Thus, it will be important in future research to examine weight trajectories among individuals who relapse to smoking. Due to the nature of the trial design, we were unable to assess weight trends of individuals quitting smoking who did not receive a weight management intervention; however, the Bibliotherapy condition was a low-intensity intervention (ie, a book that would be available to anyone with post-cessation weight concern). In addition, since we did not measure the proportion of smoking menthol cigarettes, we are not able to report on this characteristic of the sample; the relationship between the type of cigarette and weight trajectories will be an important area of future study.³¹ Furthermore, we were not able to evaluate the post-cessation weight change trajectory in the absence of varenicline use, since all of the participants in this study received at least one box of varenicline; however, a recent meta-analysis indicates that varenicline has a small and short-term effect on weight.³² Also, due to the trial design, individuals who quit smoking in this study may be inherently more weight-conscious compared to individuals who quit smoking in the general public, resulting in weight trends that may be different from the general public. Finally, consistent with previous weight management-focused research,^33,34 self-weighing may have been a potent strategy for impacting the post-cessation weight trajectory of these participants, and the potential role of self-weighing alone as a stand-alone post-cessation weight intervention strategy should be examined.

Conclusions

Via secondary analysis of a combined smoking cessation and weight management trial, we found that weight is gained at a rate of 0.0111 kg/day (0.0123% from baseline weight per day) in the first month after quitting smoking with a slower rate of weight change continuing until about day 70, after which weight appears to plateau. Our findings indicate that future combined smoking cessation and weight management trials should emphasize maintenance of weight management strategies in the first month after quitting smoking.

Funding

The study was funded by a grant from the National Institutes of Diabetes and Digestive and Kidney Diseases, with the title of “Efficacy of Two Novel Behavioral Post Cessation Weight Gain Interventions” (R. Krukowski, Principal Investigator, R01 DK107747).

Declaration of interests

The authors have no relevant financial disclosures.

Author Contributions

Mackenzi Oswald (Conceptualization [equal], Writing - original draft [lead]), Arin Shahbazian (Data curation [lead], Formal analysis [lead], Visualization [lead], Writing - review & editing [supporting]), Kathryn Ross (Conceptualization [supporting], Writing - review & editing [supporting]), Wen You (Formal analysis [supporting], Supervision [supporting], Writing - review & editing [supporting]), and Rebecca Krukowski (Conceptualization [equal], Funding acquisition [lead], Investigation [lead], Project administration [lead], Supervision [lead], Writing - review & editing [lead])

Data Availability

The data are available upon request from the corresponding author.