Impact of Changes in Regular Use of Marijuana and/or Tobacco On Chronic Bronchitis

Abstract

We sought to evaluate possible changes in the prevalence of chronic bronchitis in relation to continuing or changing smoking status for marijuana and/or tobacco. For this purpose we followed 299 participants in a longitudinal cohort study of the impact of heavy habitual use of marijuana alone or with tobacco on respiratory symptoms over a mean of 9.8 years during which subjects underwent repeated administration of a detailed drug use and respiratory questionnaire at intervals of ≥1 yr. Using logistic regression, we calculated odds ratios to assess the relationship between chronic bronchitic symptoms and smoking status for marijuana and tobacco at the first visit (current smoking versus never smoking) and at the last follow-up visit (continuing smoking versus, separately, never smoking and former smoking). We found that continuing smokers of either marijuana or tobacco had a significantly increased likelihood of having chronic bronchitis at follow-up compared to both never smokers and former smokers. On the other hand, former smokers of either substance were no more likely to have chronic respiratory symptoms at follow-up than never smokers. These findings demonstrate the benefit of marijuana smoking cessation in resolving pre-existing symptoms of chronic bronchitis.

Keywords :

• Tobacco
• Chronic bronchitis
• Respiratory symptoms
• Smoking cessation
• Longitudinal observational study

Introduction

In many societies, marijuana is the second most commonly smoked substance after tobacco (1). Among young adults 19–26 years of age participating in a U.S. nationwide survey, 15.3–20.6% admitted to smoking marijuana within the previous month, while 5.1–6.3% stated that they smoked marijuana daily (2). In a cross-sectional community-based study of the pulmonary consequences of regular marijuana use, habitual smoking of marijuana (average of ∼3–4 marijuana cigarettes [“joints”] per day for >10 years) was shown to be associated with an increased prevalence of chronic cough and sputum production (18–19%), wheeze on ≥21 days a year (24%) and ≥2 episodes of increased cough and sputum production over the preceding 3 years (13%), compared to a significantly reduced prevalence of similar respiratory symptoms in nonsmokers (2–5%, 7% and 2%, respectively) (p < 0.05) (3).

In the same study, the proportion of tobacco smokers reporting these same symptoms was comparable to that of the marijuana smokers and no additive effects of concomitant smoking of the two different substances on respiratory symptoms was noted. Similar findings have been reported by other investigators (4–6), except that an additive effect of marijuana and tobacco on respiratory symptoms was noted in one of the latter studies (4). In a more recent population-based study, the combined use of marijuana and tobacco was associated with a significantly higher prevalence of chronic bronchitic symptoms than tobacco alone, but smokers of marijuana alone did not report more symptoms than nonsmokers (7).

A number of studies have shown that continuing tobacco cigarette smoking is associated with persistent respiratory symptoms, while smoking cessation leads to an improvement in chronic cough, phlegm and wheeze (8–10). In contrast, little information is available concerning the effect on respiratory symptoms of cessation of marijuana or tobacco smoking by habitual smokers of marijuana alone or with tobacco. To address this question, we analyzed respiratory symptom data from a longitudinal study of respiratory status in a convenience cohort of heavy, habitual smokers of marijuana alone (MS), regular smokers of tobacco alone (TS), smokers of both marijuana and tobacco (MTS) and nonsmokers (3,11).

As already noted, a significantly higher prevalence of symptoms of acute and chronic bronchitis and wheeze was found among MS, TS and MTS, compared to NS, at the time of initial enrollment of these subjects in the study in 1983–85 (3). Subsequently, subjects were re-contacted and re-administered a detailed drug use and respiratory questionnaire at intervals of ≥1 year over a mean total span of 9.8 years (maximum duration of follow-up 13.7 yrs), in addition to undergoing repeated spirometry (11). The follow-up spirometry results have already been presented (11). For the present report, data from the repeated drug use and respiratory questionnaires were analyzed to determine the impact on chronic respiratory symptoms of continuing or changing marijuana and/or tobacco smoking status.

Methods

The initial sample consisted of healthy volunteers 25 to 49 yr of age, including 144 MS, 135 MTS, 70 TS and 97 NS (3). Subjects were initially recruited from the general Los Angeles area in 1983–85 (Time 1) through newspaper and radio announcements. Entry criteria have been reported previously (3). Exclusionary criteria included the following: [1] current or previous intravenous drug abuse; [2] smoking of other illicit substances (e.g., crack cocaine, PCP, methamphetamine, heroin or opium) >12 times per lifetime or within the previous 6 months; [3] history of chronic respiratory illness; or [4] significant occupational exposures to substances potentially hazardous to respiratory health.

Eligible subjects completed a detailed respiratory and drug use questionnaire adapted from the DLD/NHLBI questionnaire (12) and the NIDA nationwide survey on drug abuse (13). An extensive battery of pulmonary function tests was also performed, the cross-sectional and longitudinal results of which have been published previously (3,11). Since 1985, extensive efforts were made to re-contact at least annually, by mail and telephone, all 446 participants who had completed a questionnaire at the time of the first interview. Participants who moved were traced through work telephone numbers, contacts with individuals identified by the subject as likely to know his or her whereabouts, department of motor vehicle rosters, voter registration files, the U.S. Social Security forwarding system and commercial credit searches.

A field visit to the last known residence of a participant was utilized, if necessary. Re-contacted participants were asked to complete an interval respiratory questionnaire either face-to-face, in conjunction with the performance of interval pulmonary function tests, or over the telephone. The National Death Index (NDI) was used to identify deaths in the study group; for deaths identified by the NDI, death certificates and hospital and pathology records were requested to determine the cause of death.

This longitudinal observational study was approved by the Institutional Review Board (IRB) of the UCLA School of Medicine Human Subjects Protection Committee (HSPC) and all study participants signed an informed consent form approved by the IRB. In addition, the study was conducted under a Certificate of Confidentiality issued by the National Institute on Drug Abuse/NIH.

Data analysis

The 299 cohort members (97 female, 201 male) who had at least two visits (range of number of visits 2 to 9 for a maximum follow-up time [for 9 visits] of 13.8 years) were included in the longitudinal analysis of respiratory symptoms. Using data from the first visit only (n = 299), the SAS proc logistic regression procedure (14) was used to assess the cross-sectional relationship between chronic symptoms (cough, sputum, wheezing, increased cough and sputum and shortness of breath) and smoking status for both tobacco and marijuana. Chronic respiratory symptoms were considered to be present if subjects reported the following: cough or sputum production first thing in the morning in bad weather and at any time of the day on most days for as much as 3 months out of the year or first thing in the morning on more than 50 days a year for 2 or more consecutive years; more than one episode of increased cough and sputum lasting for 3 weeks or more within the last 3 years; wheezing on 20 or more days a year or if associated with attacks of shortness of breath; and shortness of breath if subjects became breathless when walking with others their own age on more than 20 days a year. The analysis was adjusted for age and gender and odds ratios were calculated from the logistic regression model.

Using data from the follow-up visits, in addition to the first visit, the Generalized Estimating Equation (GEE) approach was used to accommodate missing observations and both time-independent and time-dependent explanatory variables (15). Odds ratios were calculated form the GEE model. Data from 299 subjects who completed at least visit 2 through visit 9 were included in the analysis. Outcome variables were cough, sputum, wheezing, increased cough and sputum and shortness of breath. Explanatory variables included age (at each visit), gender, symptom at first visit, tobacco use status at each visit (never, former and current) and marijuana use status at each visit (never, former and current).

Because eligibility criteria for initial recruitment of participants required either no or negligible prior use of tobacco (for NS and MS) and no or negligible use of marijuana (for NS and TS), effectively no subject during the first visit was considered to be a “former” smoker of either tobacco or marijuana. Consequently, the designation of a subject as a former smoker of marijuana and/or tobacco at any time during the follow-up phase of the study implied that that subject had been smoking either or both substances at baseline and subsequently quit smoking marijuana and/or tobacco. An auto-regressive working correlation structure was used in the model. Analysis with independence and exchangeable working correlation structures was also carried out, and the results were similar.

For descriptive purposes, subjects were also classified both by their smoking status at their First Visit (MS, TS, MTS or NS) [total of four smoking categories] and by whether their smoking status at their Final Visit did not change (MS→MS; TS→TS; MTS→MTS; NS→NS) or did change as a result of quitting smoking marijuana and/or tobacco (MS→NS; MTS→MS, MTS→TS; MTS→NS; TS→NS) [total of nine smoking categories]. Since very few participants started smoking marijuana, tobacco and/or another substance for the first time over the course of the study, the latter individuals were not included in the descriptive analysis of change in symptoms.

Results

Follow-up questionnaires were completed at least once by 299 subjects (67%) from the original cohort of 446 subjects. One-third of the follow-up cohort were female. The numbers of follow-up participants by smoking category at the first and final visit, including age, gender, baseline intensity of smoking and duration of follow-up are shown in Table 1. The mean age of the follow-up subjects in all smoking categories was 33.4 ± 6.4 (S.D.) years. The mean duration of follow-up for subjects in each of the smoking status categories ranged from 8.3 to 11.8 years. The numbers (and percentages) of follow-up participants with 2–9 visits were as follows: 2–299 (100%); 3–236 (79%); 4–184 (63%); 5–138 (46%); 6–96 (32%); 7–70 (23%); 8–37 (12%); 9–14 (5%). Comparison of the subjects in the original cohort who provided only baseline data with the follow-up cohort according to the original smoking status showed no significant differences in baseline characteristics with the sole exception of a higher proportion of tobacco-only smokers in the follow-up group who reported acute bronchitic episodes at baseline compared to the baseline-only subjects (p = 0.026; Chi square).

TABLE 1.  Number of follow-up participants by smoking category, age, duration of follow-up and smoking history

Smoking Status	N	Age	Years first to last visit	Tobacco			Marijuana
				Baseline pack-years	Cigarettes / Day		Baseline joint-years	Joints / Day
					First Interview	Last Interview		First Interview	Last Interview
UNCHANGED
NS ≡ NS	77	32.2 (0.7)*	10.8 (0.4)	0.0 (0.0)	0.0 (0.0)	0.0 (0.0)	0.0 (0.0)	0.0 (0.0)	0.0 (0.0)
MS ≡ MS	72	31.8 (0.6)	9.4 (0.5)	3.7 (1.2)	0.0 (0.0)	0.0 (0.0)	62.1 (20.4)	4.5 (1.1)	3.0 (0.4)
TS ≡ TS	40	36.9 (1.1)	8.6 (0.6)	21.6 (2.3)	28.7 (2.4)	22.4 (2.7)	0.0 (0.0)	0.0 (0.0)	0.0 (0.0)
MTS ≡ MTS	40	33.4 (1.0)	8.6 (0.7)	16.2 (2.0)	19.9 (2.0)	22.5 (2.4)	41.0 (6.0)	3.5 (0.5)	2.7 (0.8)
CHANGED
MS ≡ NS	23	32.7 (1.2)	10.5 (0.6)	2.8 (1.2)	0.0 (0.0)	0.0 (0.0)	53.2 (14.6)	4.2 (0.8)	0.0 (0.0)
TS ≡ NS	9	37.4 (2.0)	11.8 (0.4)	21.6 (5.0)	29.4 (3.6)	0.0 (0.0)	0.1 (0.0)	0.0 (0.0)	0.0 (0.0)
MTS ≡ NS	9	33.2 (2.0)	8.3 (3.6)	11.2 (2.4)	5.3 (2.8)	0.1 (0.1)	52.5 (19.1)	3.2 (1.2)	0.0 (0.0)
MTS ≡ MS	10	37.0 (2.2)	9.1 (1.4)	15.2 (4.9)	15.6 (5.7)	0.0 (0.0)	42.8 (9.7)	7.2 (4.8)	2.3 (0.5)
MTS ≡ TS	12	33.3 (2.1)	10.4 (0.9)	13.7 (2.1)	17.0 (4.2)	18.3 (5.2)	58.1 (15.2)	2.5 (1.1)	0.0 (0.0)

*Mean (SEM)

Figure 1 shows the prevalence of each respiratory symptom in the follow-up cohort (N = 299) at visit 1 by baseline smoking status. These results are comparable to those that have previously been presented for all 446 subjects studied at baseline (3). Specifically, in the follow-up cohort, for all respiratory symptoms, except shortness of breath, the frequency of reported symptoms was significantly higher in the marijuana-only smokers, as well as the tobacco-only smokers and combined smokers of marijuana and tobacco, compared to nonsmokers (p < 0.05, chi square). No significant differences in symptom frequency were noted between individuals within the different active smoking categories (MS, TS and MTS).

Figure 1.  Prevalence of chronic respiratory symptoms in the entire longitudinal cohort (N = 299) by baseline smoking status.

Similarly, at the baseline visit, odds ratios for each respiratory symptom, except shortness of breath, were significantly increased for marijuana smokers compared to non-marijuana smokers (controlling for tobacco smoking): cough (OR 2.5; 95% C.I. 1.3–5.1, p = 0.009); sputum (OR 7.7; 95% C.I. 2.2–26.3, p < 0.001); increased cough and sputum (OR 12.2; 95% C.I. 1.6–95.4, p < 0.02); and wheeze (OR 13.0; 95% C.I. 4.4–43.6, p < 0.0001). Similar findings were noted for tobacco smokers (data not shown), except that use of tobacco, unlike marijuana, was also associated with increased breathlessness (OR 9.1 95% C.I. 1.9–44.1, p < 0.01).

Figure 2 illustrates the prevalence of each respiratory symptom both at the baseline visit (2A) and the final visit (2B) in follow-up participants whose smoking status did not change between the first and last visits. As was noted in the entire follow-up cohort at baseline (Figure 1), each respiratory symptom remained significantly more common among MS, TS and MTS than among NS at both the baseline visit and the final visit, with the exception of shortness of breath, which was not more frequently reported by MS at either visit.

Figure 2.  Prevalence of respiratory symptoms at the baseline visit (A) and the final visit (B) in only those follow-up participants whose smoking status did not change.

The impact of changes in smoking status for marijuana or tobacco are shown in Table 2, based on the results of the GEE analysis. Here, for each symptom, the odds ratios indicate the likelihood of the presence of that particular symptom at follow-up in the continuing smokers of tobacco and, separately, the continuing smokers of marijuana compared with both never smokers of that substance and, separately, with former smokers of the same substance; the analysis takes into account the smoking status for the alternate substance in those who smoked both marijuana and tobacco at baseline.

TABLE 2.  Odds ratios (OR) and 95% C.I. for having a chronic respiratory symptom (cough, sputum, wheeze, acute bronchitic episodes) at follow-up 
1) for continuing smokers of marijuana or tobacco compared to never smokers, 2) continuing smokers of marijuana or tobacco compared to former smokers (those who quit since baseline) and 3) former smokers of marijuana or tobacco compared to never smokers

		Cough	Sputum	Increased bronchitis episodes	Wheeze
Tobacco
Current vs. Never	OR	3.7	2	3.3	4
	95% C.I.	2.1, 6.6	1.1, 3.6	1.8, 6.2	2.3, 7.1
	p	<.0001	0.0185	0.0002	<.0001
Current vs. Former	OR	2.8	2.8		4.9
	CL	1.2, 6.6	1.05, 7.6		2.4, 10.x
	p	0.0162	0.0405	NS	<.0001
Former vs. Never	OR
	95% C.I.
	p	NS	NS	NS	NS
Marijuana
Current vs. Never	OR	1.7	2.1	2.3	3.4
	95% C.I.	1.01, 2.9	1.1, 4.0	1.2, 4.4	1.7, 6.6
	p	0.0447	0.0201	0.0117	0.0003
Current vs. Former	OR	3.3	4.2		2.1
	95% C.I.	1.3, 8.6	1.6, 11.2		1.04, 4.07
	p	0.0155	0.0039	NS	0.0369
Former vs. Never	OR
	95% C.I.
	p	NS	NS	NS	NS

In addition, for former smokers of each substance the odds of having each symptom is compared with never smokers of the same substance. Data are not shown for shortness of breath since relatively few subjects in all smoking categories reported this symptom at follow-up. Findings indicate that for chronic cough, chronic sputum production and wheeze, continuing smokers of either marijuana or tobacco have an increased likelihood of having that symptom at follow-up compared to both never smokers and former smokers of the same substance.

The confidence intervals are wide, probably due to the small numbers, but nonetheless exclude 1.0, thus indicating statistically significant findings. In addition, comparison of former smokers with never smokers for either substance revealed no significant odds ratios for these symptoms. Taken together, these results indicate that former smokers are less likely than continuing smokers and no more likely than never smokers to have symptoms of chronic bronchitis. For episodes of increased cough and sputum (acute bronchitis), continuing smokers of tobacco and, separately, of marijuana had a greater likelihood of having these episodes compared to never smokers but not to former smokers; on the other hand, former smokers did not have a greater likelihood of having episodes of increased cough and sputum than never smokers.

For descriptive purposes, Table 3 shows the percentage of subjects in each follow-up smoking category who acquired a symptom (chronic cough, chronic sputum, acute bronchitic epidsodes, wheezing or shortness of breath) that was not present at baseline, as well as the percentage of subjects who lost a symptom that was present at baseline. Among those subjects who did not change their smoking status and did not have a particular symptom at baseline, a small proportion of never smokers acquired a symptom in each symptom category, but a larger proportion of the continuing smokers of either substance or of both substances acquired a symptom that had not been present previously.

Table 3.  Changes in chronic respiratory symptoms (lost or gained) in relation to persistent or changing smoking status from baseline to the last follow-up visit

		Symptoms
		Cough		Sputum		Acute Bronchitis		Wheeze
Smoking Status		Lost*	Gained†	Lost*	Gained†	Lost*	Gained†	Lost*	Gained†
Unchanged
NS (N = 77)		0/0 (0)§	5/77 (6.5)	2/3 (66.7)	3/72 (4.2)	1/1 (100)	1/76 (14.3)	3/4 (75)	5/73 (6.8)
MS (N = 72)		10/16 (62.5)	6/54 (11.1)	11/18 (61.1)	7/53 (13.2)	9/11 (81.8)	7/60 (11.7)	14/18 (77.8)	10/53 (19.2)
TS (N = 40)		7/11 (63.6)	5/29 (17.2)	8/11 (72.7)	2/29 (6.9)	3/4 (75)_	2/36 (5.5)	7/13 (53.8)	3/27 (11.1)
MTS (N = 40)		5/11 (45.4)	6/27 (22.2)	6/11 (54.5)	5/27 (18.5)	3/5 (60)	4/34 (11.8)	7/12 (58.3)	8/26 (30.8)
Changed
Baseline	Follow-up
MS	NS (N = 23)	3/4 (75)	0/18 (0)	3/4 (75)	0/18 (0)	2/2 (100)	2/21 (9.5)	3/4 (75)	2/19 (10.5)
TS	NS (N = 9)	4/4 (100)	0/5 (0)	3/3 (100)	0/6 (0)	4/4 (100)	0/5 (0)	2/3 (66.7)	1/6 (16.7)
MTS	NS (N = 9)	3/3 (100)	0/6 (0)	3/3 (100)	0/6 (0)	0/0 (0)	0/9 (0)	1/2 (50)	0/6 (0)
MTS	MS (N = 10)	0/1 (0)	4/9 (44.4)	1/3 (33.3)	2/7 (28.6)	0/1 (0)	0/9 (0)	2/4 (50)	1/5 (20)
MTS	TS (N = 12)	1/1 (100)	3/11 (27.2)	3/3 (100)	2/9 (22.2)	1/1 (100)	3/11 (27.2)	3/6 (50)	1/6 (16.7)

*Lost = the number of subjects who lost the designated symptom at follow-up (numerator) as a proportion of the number who had that symptom at baseline (denominator).

^†Gained = the number of subjects who gained the designated symptom at follow-up (numerator) as as a proportion of the number who had not had that symptom at baseline (denominator)

^§Numbers in parentheses represent percentages.

Conversely, among those subjects who did not change their smoking status, except for the never smokers (few of whom were symptomatic at baseline), generally half or more of those who had a particular symptom at baseline lost that symptom at follow-up, although the number of subjects who lost a symptom was roughly equivalent to the number who acquired the same symptom, so that the prevalence of each symptom remained approximately the same at the final visit compared to the first visit (Figures 2A and 2B).

Among those smokers of marijuana (MS) or tobacco (TS) who changed their smoking status to that of a nonsmoker (NS), as well as among those smokers of both marijuana and tobacco (MTS) who quit smoking both substances, none to very few acquired a symptom that they did not have at baseline, while generally half to all of those who did have a symptom at baseline lost that symptom at follow-up. Conversely, among the combined smokers of both marijuana and tobacco who gave up only one of these substances, 9–44% acquired a symptom that was not present at baseline (except for acute bronchitic episodes among MTS who became MS), while a variable proportion of these subjects lost a symptom that was present at baseline.

Discussion

Although several investigators have shown, in cross-sectional studies, that regular smokers of marijuana with or without concomitant tobacco have significantly more chronic respiratory symptoms than non-smokers (3–7), little follow-up data are available concerning the subsequent course of these symptoms in those who continue smoking marijuana with or without tobacco or quit smoking one or both substances. In the present longitudinal study, follow-up of 299 subjects from a community-based cohort of 446 nonsmokers or habitual smokers of marijuana and/or tobacco over a mean interval of 8.3 to 11.8 years demonstrated that continuing smoking of one or both of these substances was associated with a persistently increased frequency of symptoms of chronic bronchitis compared with both never smokers and former smokers.

Therefore, those who quit smoking and became nonsmokers of any substance were likely to have fewer symptoms than those who continued to smoke; moreover, these quitters were no more likely to have chronic bronchitic symptoms than never smokers. Similar findings were noted for both former marijuana and former tobacco smokers. Moreover, the results in those individuals who quit smoking marijuana are similar to previously reported findings in former tobacco smokers in whom symptoms of chronic bronchitis have been shown to resolve after smoking cessation (8).

Bronchoscopic studies have previously demonstrated that habitual smokers of marijuana with or without tobacco have a greater frequency than nonsmokers of histopathologic abnormalities in endobronchial biopsies, including goblet cell hyperplasia, loss of ciliated epithelial cells and intra- and sub-epithelial inflammation (16). As this abnormal bronchial mucosal histopathology in regular smokers of marijuana is consistent with an increased production of mucus in the face of a diminished capacity to clear the mucous secretions from the airway, it is likely that it contributes to the increased prevalence of symptoms of chronic bronchitis among habitual marijuana users.

Previous bronchoscopic studies in tobacco smokers who subsequently quit smoking tobacco have shown a significant reduction in goblet cell hyperplasia (17), which probably contributes to the reduction in bronchitic symptoms following tobacco smoking cessation. On the other hand, changes in the degree of lung inflammation after smoking cessation have only been variably demonstrated (17–20). Although bronchoscopic studies have not been carried out both before and after cessation of habitual marijuana smoking, it is reasonable to expect that comparable improvement in goblet cell hyperplasia would accrue in the quitters of marijuana smoking as in the quitters of tobacco smoking that would correspond to a similar reduction in bronchitic symptoms.

Cessation of tobacco smoking in chronic obstructive bronchitis is well known to be associated with improvements in lung function over the year following smoking cessation (20), as well as a reduction in the accelerated rate of decline characteristic of patients with chronic obstructive pulmonary disease (20–22). However, habitual smoking of marijuana in the absence of tobacco has not consistently been found to be associated with evidence of airflow obstruction, as reflected by the forced expired volume in 1 sec (FEV₁) (3,5,7) or with an accelerated loss of lung function over time (3,6,23).

Therefore, one would not expect to find any changes in lung function in habitual marijuana smokers who quit smoking marijuana. Indeed, in the same follow-cohort for whom respiratory symptoms are herein presented, it was previously reported that the annualized rate of change in FEV₁ in the habitual marijuana smokers was identical to that in the nonsmokers. On the other hand, an elevated airway resistance (in the absence of evidence of obstruction in the small airways), has been reported in regular marijuana smokers (3,6,24), consistent with the inflammatory edema of the walls of the central airways that has been demonstrated at videobronchoscopy and in endobronchial biopsies in regular marijuana smokers (25).

It is possible that the inflammatory edema associated with marijuana smoking may also contribute to the increased frequency of cough and wheezing observed in these individuals. Whether quitting marijuana smoking reduces this inflammatory swelling of the bronchial walls and thereby reduces the elevated airway resistance, possibly contributing to the reduction in chronic respiratory symptoms in former smokers of marijuana, has not been studied.

The strength of this study is that it is the first to systematically assess the impact of cessation of habitual marijuana smoking on smoking-related symptoms of chronic bronchitis, adjusted for concomitant tobacco smoking, in a well-characterized cohort of smokers of marijuana and/or tobacco. Whether chronic bronchitic symptoms are reversible following cessation of marijuana smoking is a matter of concern for some habitual marijuana smokers.

Although the data presented herein had been collected as part of a longitudinal cohort study that was completed approximately 15 years ago, no other data addressing this issue have been published during this interval from other prospective studies in regular marijuana smokers. However, our study has several limitations. First, smoking status for marijuana and tobacco was assessed only by self-report and was not verified by objective measurements of serum, salivary or urinary cotinine or urinary cannabinoids. However, this study was an observational study of self-reported smokers and nonsmokers with no intervention to assist smokers in stopping smoking; therefore, smoking subjects had no obvious motivation to deceive the investigators regarding changes in their smoking status.

Second, a large percentage of the initial cohort failed to return to the laboratory for follow-up visits with varying lengths of follow-up in those who did return. On the other hand, the baseline demographic and clinical characteristics of the subgroup without follow-up data were, for most features, similar to those of the follow-up subgroup. This similarity in baseline characteristics, coupled with the observational nature of the study and the absence of any smoking-cessation intervention, should minimize any bias toward experiencing or reporting (or not reporting) respiratory symptoms in relation to persistence or cessation of smoking in the follow-up cohort.

Conclusions

Findings from this longitudinal cohort study indicate that continuing smoking of marijuana, either alone or with tobacco, is associated with a net persistence of symptoms of chronic bronchitis, similar to the findings in continuing smokers of tobacco alone. In contrast, complete cessation of smoking by former habitual smokers of either marijuana or tobacco alone or of marijuana plus tobacco was accompanied by substantial declines in chronic respiratory symptoms. On the other hand, combined smokers of both marijuana and tobacco who quit smoking only one of these two substances only variably lost symptoms that had been present at baseline and occasionally acquired symptoms that had not been present previously. These findings provide, for the first time, documented evidence of the benefit of marijuana smoking cessation with respect to the resolution of pre-existing symptoms of chronic bronchitis that may reflect repair of the airway injury that has previously been demonstrated in endobronchial biopsies obtained from continuing smokers of marijuana (23). Communicating these findings to habitual smokers of marijuana may assist in motivating those smokers with chronic bronchitic symptoms to quit smoking.

Declaration of Interest

None of the authors has any financial, consulting or personal relationships with other people or organizations that could influence or bias the authors’ work. The authors did not obtain any scientific writing assistance. The study was funded by a grant from the National Institute on Drug Abuse (R37 DA003018; D. P. Tashkin, P.I.).

Acknowledgments

The study was funded by a grant from the National Institute on Drug Abuse (R37 DA003018; D. P. Tashkin, P.I.). We would like to acknowledge the contributions of Enoch Lee and John Dermand who were primarily responsible for recruitment and follow-up of the participants and the administration of the questionnaires.