Recent estimates of the prevalence of smoking among adults with chronic pain range from 17% to 31%.1–4 This is comparable to the 20.9% prevalence of smoking in the general United States population, as reported in 2005.5 Although effective treatments are available for treatment of nicotine dependence,6 concomitant use of opioids may be an important barrier to successful smoking cessation.7 This is particularly relevant to smokers with chronic pain, in that opioid analgesic medications have been increasingly prescribed for treatment of chronic pain during the past decade.8 For some patients, opioid use may itself be deleterious to the treatment of chronic pain, and pain rehabilitation programs often attempt to taper opioid analgesic medications as a part of treatment.9
The interaction between smoking and chronic opioid use among patients with chronic pain has not been rigorously studied, and requires drawing parallels from other disparate populations of smokers receiving long-term opioids. A threefold or more increase in the prevalence of smoking has been reported among patients receiving methadone maintenance therapy for opiate addiction,10,11 which may, in part, be influenced by high rates of psychiatric and behavioral comorbidities not generally found in patients with chronic pain.12–15 However, despite differences between these two patient populations, knowledge of the associations between smoking and opioid use in patients receiving methadone maintenance therapy could provide a general framework for understanding the potential associations among smokers receiving opioids for chronic pain. Early investigators observed that higher doses of methadone are associated with increased levels of smoking.16,17 Although the impact of opioid use on smoking cessation among patients with chronic pain has not been specifically studied, an evidence-based smoking cessation intervention for adults receiving stable methadone maintenance therapy (i.e., no changes in methadone dosages) yielded 6-mo abstinence rates of only 5%.7 The authors speculated that unabated opioid use during attempts to reduce tobacco consumption may have, in part, enhanced continuing tobacco use in these patients.7
Collectively, these experimental and clinical observations suggest that there is an association between chronic opioid use and nicotine dependence. This raises the possibility that smoking status may affect the ability of opioid-dependent chronic pain patients to reduce or eliminate their opioid use. The primary aim of this study was to determine if medically directed opioid tapering during a 3-wk pain rehabilitation program would be less successful among smokers compared to nonsmokers with chronic pain. Successful tapering was defined as completely discontinuing opioids during the 3-wk treatment period. Secondary aims included determining the associations between smoking status and baseline opioid use, pain severity, and successful program completion.
This study was approved by the Mayo Foundation IRB. All adults consecutively admitted to the Mayo Comprehensive Pain Rehabilitation Center from September 2003 through February 2007 were eligible for this study. During this time period, 1257 individuals with chronic pain were admitted. Data were collected prospectively as an integral part of the pain treatment program. Before admission, these patients provided authorization for the use of their medical records for medical research. Patients who used forms of tobacco other than cigarettes were excluded from the study. Seven patients used chewing tobacco, five smoked cigars or a pipe and one patient used multiple forms of tobacco. The form of tobacco use could not be determined for three patients. The study cohort thus included 1241 participants (women 928, men 313). The four most common admitting diagnoses or primary sites of pain included low back pain, fibromyalgia, headache, and generalized pain. These 4 groups accounted for 67% of all patients (Table 1).
The intensive outpatient pain rehabilitation program is 3 wk long and incorporates opioid tapering. The primary goal of treatment is to restore physical and emotional functioning. Admissions to the rehabilitation program occur on a revolving basis and patients attend 8 h daily for 15 consecutive working days. A cognitive-behavioral model serves as the basis for treatment.18 Patients referred for this type of intensive treatment have generally received medical care for chronic pain and experienced incomplete symptomatic relief from multiple pharmacologic trials, repeated courses of physical therapy, interventional pain procedures or surgery. Multidisciplinary pain treatment differs from independently prescribed outpatient physical therapy, occupational therapy and cognitive-behavioral therapy, in that patients are involved in all these treatment modalities concurrently on a daily basis throughout the treatment period.18
Before admission, patients undergo an evaluation by the medical and psychology staff members of the rehabilitation program. The evaluation serves two important purposes. First, patients are assessed to ensure that they do not have pain due to cancer. Otherwise, there are no specific diagnostic criteria for admission except for the presence of chronic pain and the patient’s decision to pursue multidisciplinary pain treatment. However, patients are excluded from the program if they have acute or subacute pain of <3-mo duration. Patients are also excluded if they have had major surgery within 6 mo. Examples of major surgery involving the spine include discectomy, laminectomy or fusion. Other examples of major surgery includes intraabdominal or intrapelvic surgery requiring inpatient postoperative care; total hip, knee or shoulder replacement; thoracic surgery or trauma surgery. These patients are excluded from program participation in that intensive physical reconditioning may worsen the patient’s medical or surgical status and lead to a deterioration in functionality. The 3-wk program is also not recommended for patients with active substance abuse or dependence disorders, severe mood disorders, psychotic disorders or dementia. The second purpose of the evaluation is to ensure that the patient understands and approves the recommendation that all opioid analgesic medications will be tapered during the 3-wk program. This is a well established practice of our rehabilitation program.9,18–20 Patients who do not desire to taper opioids are referred for less intensive pain rehabilitation services at our institution.
The goal of improving physical functioning is achieved by involving patients in daily physical therapy with a focus on aerobic and strengthening exercises. The intensity of exercise activity is individually advanced over the course of the 3-wk program under the direct, daily supervision of a physical therapist. Patients are also involved in daily occupational therapy where impairments in activities of daily living are addressed. Emotional functioning is optimized by involving patients in daily cognitive-behavioral group educational sessions under the direction of behavioral health psychologists with postdoctoral specialty training in pain psychology. Specific areas of mental health that are addressed include stress management, relaxation training (which is facilitated by individualized biofeedback training sessions), chemical health education, cognitive-behavioral-based techniques for management of depression and pain-related negative affect, and elimination of pain behaviors. Immediate treatment outcomes have been reported and are assessed using validated measures of physical and emotional functioning.9,19,20
Baseline demographic and clinical characteristics were collected at admission including age, pain duration, primary pain site, marital status, years of education, and medication use.
Measures and Procedures
Smoking Status and Daily Cigarette Use
Smoking status was assessed upon admission to the pain treatment center using a self-report questionnaire. All patients completed the following question: “Describe your history of smoking cigarettes.” Patients could identify themselves as never smokers, former smokers (at least 1 mo of abstinence) or current smokers. The number of cigarettes smoked daily by current smokers was determined by the following question: “If you are a current smoker, how many cigarettes do you smoke per day?” This self-report of smoking status was not biochemically confirmed.
Opioid and Non-opioid Analgesic Medication Tapering
At admission, the daily opioid dose of each subject is determined by self-report and review of pharmacy records. The opioid medication the patient is taking at the time of admission is used in the taper schedule. Given the large range of opioid dosages and diversity of patient clinical characteristics, each opioid taper is individually tailored and predetermined tapering schedules are not used. Benzodiazepines and non-opioid analgesic medications, including nonsteroidal antiinflammatory drugs (NSAIDs) and muscle relaxants, are reduced or discontinued. These non-opioid medications are withdrawn after the opioid taper schedule is underway or completed. Examples of muscle relaxants that are tapered include cyclobenzaprine, baclofen, carisoprodol, chlorzoxazone, and tizanidine. These various classes of drugs are eliminated or reduced on an individual basis if the patient reports limited or no symptomatic relief. Medication elimination or reduction is also undertaken on an individual basis to ameliorate adverse side effects, including daytime sedation, and to reduce long-term risks of medication-related complications. Medication discontinuation is not recommended for patients who use these drugs to treat comorbid medical or psychiatric illnesses. All medication tapers are initiated and coordinated by a physician after admission to the rehabilitation program.
Pain severity was assessed upon admission and dismissal using the pain severity subscale of the Multidimensional Pain Inventory (MPI). This self-report questionnaire has been validated and is used to measure the psychosocial impact of chronic pain.21 The pain severity subscale score is a composite score calculated from responses to the following 3 questions; 1) “Rate the level of your pain at the present moment;” 2) “On the average, how severe has your pain been during the last week;” 3) “How much suffering do you experience because of your pain.” Responses to each question are scored on a 7-point Likert scale where 0 indicates no pain or suffering and 6 indicates extreme pain or suffering. Raw scores were converted to standardized T-scores with a normative value of 50 in patients with chronic pain (range, 0–100) and a standard deviation of 10.22
Demographic and clinical characteristics of smokers, former smokers, and never smokers were compared across groups using analysis of variance (ANOVA) for continuous variables and the χ2 test for categorical variables. Post hoc contrast analyses were conducted using Tukey HSD to assess pairwise comparisons among smoking groups. Analyses of pain severity pre-post opioid tapering were completed using repeated measures ANOVA including main effects of time, smoking status and the interaction of time and smoking status. For analysis, daily opioid intake was converted to oral morphine equivalents using previously established methods.9,19,20 A post hoc analysis of covariance (ANCOVA) was conducted for mean morphine equivalent dose at admission using demographic and clinical characteristics that were significantly different among the three groups as covariates. These covariates included age, sex, marital status, years of education, and duration of pain. Likewise, a repeated measures ANCOVA was performed for pain severity using a similar set of covariates.
Logistic regression analyses were completed to identify independent correlates of unsuccessful opioid tapering and program noncompletion. In the analysis, where unsuccessful opioid tapering was the dependent variable, admission morphine equivalent use, smoking status, and pain severity were the independent variables. In the regression model, where unsuccessful program completion was the dependent variable, admission opioid use, smoking status, admission pain severity, and marital status were the independent variables. The findings were summarized using odds ratios and corresponding 95% confidence intervals. Two-sided tests were used in all analyses and the level of significance for all statistical tests was set at P < 0.05. All analyses were completed using Statistical Package for the Social Sciences, version 15.0.
Baseline Demographics and Smoking Characteristics
Of the 1241 patients, 313 (25%) were current smokers, 294 (24%) were former smokers, and 634 (51%) were never smokers. The mean number of cigarettes smoked daily at admission was 14.6 ± 9.4 [M ± sd]. Current smokers were younger, had fewer years of education, were less likely to be married and had experienced a briefer duration of pain (Table 1). The mean number of cigarettes smoked daily did not differ between smokers taking opioids and those who did not (14.5 ± 9.5 and 14.8 ± 9.3, respectively).
Opioid Use Characteristics
Upon admission, the proportion of current and former smokers using opioid analgesic medications was larger compared to never smokers (Table 2). The mean daily morphine equivalent opioid dose upon admission depended upon smoking status, with the mean opioid dose of current smokers being larger than that of never smokers (Table 2). Post hoc ANCOVA was conducted using the duration of opioid use and other clinical characteristics that were significantly different among the three groups at admission as covariates (age, sex, marital status, years of education, and pain duration). The effects of smoking status on opioid dose at admission remained significant in this multivariable model (P = 0.032).
The proportion of patients using opioids was significantly reduced by the time of program completion, as almost every patient remaining in the program was able to discontinue opioid use (P < 0.001). The frequency of opioid use at dismissal did not depend upon smoking status (Table 2). Logistic regression analysis showed higher morphine equivalent use at admission (P = 0.001), but not smoking status or pain severity, predicted unsuccessful opioid tapering at program completion (Table 3). For the small number of patients (n = 26) still using opioids at program completion, opioid dose did not depend on smoking status (Table 2).
Non-opioid Analgesic Medication Use
Upon admission, the proportion of smokers using muscle relaxants was larger compared to former and never smokers (Table 2). Smoking status did not affect the proportion of patients using NSAIDs or benzodiazepines at admission. The proportion of patients using muscle relaxants, NSAIDs, and benzodiazepines at admission was significantly reduced at the time of program completion. The frequency of non-opioid analgesic medication use at program completion did not depend upon smoking status.
The mean MPI pain severity score at admission depended upon smoking status (Table 2); the mean score of current smokers was higher compared to former and never smokers. The mean pain severity score of smokers at admission did not differ between those using opioids and those not using opioids (50.7 ± 8.7 and 48.8 ± 10.8, respectively, P = 0.9). At dismissal, no significant group differences in mean MPI pain severity scores were detected (P > 0.1). The mean pain severity score of patients who continued to use opioids at dismissal (n = 26) was 43.4 ± 11.0 and did not depend upon smoking status. This mean pain score was significantly higher than the mean pain severity score of all patients (37.0 ± 12.8) who completed opioid tapering (P = 0.012).
From admission to program dismissal, repeated measures ANOVA for pain severity showed a main effect of time (P < 0.001) and smoking status (P = 0.006) (Fig. 1). No significant interaction between time and smoking status was found (P > 0.1), indicating that smokers and nonsmokers experienced similar improvements in pain with treatment. Post hoc repeated measures ANCOVA was conducted using demographic and clinical characteristics that were significantly different among the groups at admission as covariates. The main effect of time (P = 0.001) and smoking status (P = 0.034) remained significant in this multivariable model.
The program was not completed by 15% (n = 184/1241) of patients. Twenty-four percent (n = 75) of smokers did not complete the program compared to 10% (n = 29) of former smokers and 13% (n = 80) of never smokers (χ2 = 28.9, P < 0.001). These patients left the program after a mean of 6.2 ± 3.7 days. Reasons for not completing the program included discrepant expectations of treatment related to either or all of the following including the cognitive behavioral basis of treatment, medication tapering or intensive physical reconditioning (n = 78); acute medical or psychiatric illness (n = 54); and acute nontreatment-related psychosocial stressors (n = 20). These 3 categories accounted for 83% (n = 152/184) of patients who did not complete the program. Other reasons for not completing the program included intention to pursue treatment closer to home (n = 9), lack of adequate finances (n = 3), conflicting religious activities (n = 1) and no reason was provided by 19 patients. These miscellaneous categories accounted for 17% (n = 32/184) of patients who did not complete the program. Although a significantly larger proportion of smokers did not complete the program, the frequency of smokers, former smokers, and never smokers among these various categories was not significantly different (P > 0.1). The mean number of cigarettes smoked daily was not different in smokers who did and did not complete the program (14.6 ± 9.2 and 14.7 ± 10.1, respectively).
In univariate analysis, both the admission morphine equivalent dose (166 ± 169 mg/d in patients not completing the program versus 107 ± 143 mg/d in those completing the program, P < 0.001) and the MPI pain severity score (50.4 ± 9.8 vs 47.5 ± 9.4, respectively, P < 0.001) were higher in those not completing the program. Logistic regression analysis revealed that current smoking (P = 0.002), opioid use at admission (P = 0.007) and pain severity at admission (P = 0.009) were predictive of program noncompletion (Table 4). Marital status was not predictive of program noncompletion (Table 4).
The main finding of this study was that the frequency of successful opioid tapering among current smokers with chronic pain who completed the 3-wk program was similar to that of former and never smokers, despite opioid use by a significantly larger proportion of smokers at admission, and a higher mean opioid dose in these smokers compared to never smokers. However, a significantly larger proportion of smokers did not complete treatment. Admission pain severity scores of smokers were larger compared to former and never smokers, but smoking status did not affect the improvement in pain severity scores with treatment in those completing the program.
At admission, the opioid dose of current smokers was significantly higher compared to never smokers. Whereas an explanation for this observation remains unclear, greater opioid consumption in smokers with chronic pain could be due, in part, to higher levels of pain. Indeed, among our patients, current smokers had significantly higher levels of pain severity at admission compared to nonsmokers. Alternatively, higher opioid dosages among our smokers with chronic pain may have been related to increased metabolism of opioids due to the induction of hepatic cytochrome P-450 isoenzymes by tobacco smoke.23,24
Prior studies have shown that long-term pain-related clinical outcomes are generally worse for smokers compared to nonsmokers.3,25–28 In the current study, smoking status was a predictor of program noncompletion, as a larger proportion of smokers did not complete the 3-wk program. Our findings suggest that, among patients with chronic pain, smoking, in addition to opioid use and increased pain severity, may represent an important risk factor for poor treatment adherence. Alternatively, smoking has been associated with other substance use among patients with chronic pain which could have been an unrecognized factor that contributed to our rate of program noncompletion.29 Further research will be needed to determine if poor treatment adherence explains some of the long-term smoking-related differences in the clinical outcomes of smokers with chronic pain.
In population-based studies, smokers generally report more severe, widespread pain compared to nonsmokers.4,30–33 All patients in our study, regardless of smoking status and differences in baseline clinical characteristics, experienced a significant improvement in pain severity during the 3-wk program. This improvement occurred in smokers despite higher levels of pain severity at program admission. We have previously reported similar reductions in pain severity among patients undergoing opioid tapering during our 3-wk program.9,19,20 Significant reductions in the frequency of non-opioid analgesic medication use shows reduced levels of pain severity at program dismissal were not a result of escalated use of non-opioid pain medications. Although increased pain intensity among smokers with chronic pain has been associated with worse clinical outcomes,27 our immediate posttreatment findings suggest that clinically small pretreatment increases in pain severity among smokers are amenable to treatment.
Our study has several important limitations. The predominantly Caucasian female study participants were specifically referred for pain rehabilitation and had health care resources and motivation to participate in the 3-wk outpatient program that incorporated opioid tapering. Thus, the study results may not be applicable to all smokers with chronic pain. A larger proportion of smokers did not complete the 3-wk program and this difference in program completion could have biased our results. If all opioid-using patients who did not complete the program were assumed to have failed opioid tapering, significantly more smokers (13%) would have been using opioids 3 wk after program initiation compared to former (6%) and never (9%) smokers. However, even with this conservative assumption, the absolute difference in the frequency of opioid use between current (13%) and never (9%) smokers at the end of 3 wk would be 4%, a difference of uncertain clinical significance. We did not assess the severity of nicotine dependence upon admission. Whereas the number of cigarettes smoked daily at admission was not associated with higher opioid dosages, successful opioid tapering or program completion status, individual differences in the severity of nicotine dependence could have affected our results. Similarly, we did not assess the severity of opioid withdrawal symptoms during opioid tapering. The severity of opioid withdrawal symptoms could have influenced the program completion rate and success of opioid tapering. However, the overall completion rate reported herein was similar to that previously reported for the pain treatment program.9,19 Finally, we did not assess whether patients relapsed to opioid use after treatment, which would require long-term follow-up data.
The findings from this large, consecutive series of patients with chronic pain undergoing multidisciplinary pain rehabilitation provide novel insights into the clinical associations between smoking, opioid use, pain severity, and treatment adherence. Knowledge of the interaction effects of these important health care indices could affect clinical practice in that, among patients with chronic pain who choose to discontinue chronic opioid therapy in the context of a pain rehabilitation program, smoking is a significant determinant of opioid use and program completion. Although, for those patients who completed the pain rehabilitation program, the success of opioid tapering did not depend on smoking status, a significantly larger proportion of smokers did not complete treatment for reasons that are not known.
1.Ryall C, Coggon D, Peveler R, Poole J, Palmer KT. A prospective cohort study of arm pain in primary care and physiotherapy–prognostic determinants. Rheumatology (Oxford) 2007;46:508–15
2.Wijnhoven HA, de Vet HC, Picavet HS. Explaining sex differences in chronic musculoskeletal pain in a general population. Pain 2006;124:158–66
3.Vogt MT, Hanscom B, Lauerman WC, Kang JD. Influence of smoking on the health status of spinal patients: the National Spine Network database. Spine 2002;27:313–19
4.Palmer KT, Syddall H, Cooper C, Coggon D. Smoking and musculoskeletal disorders: findings from a British national survey. Ann Rheum Dis 2003;62:33–6
5.Centers for Disease Control and Prevention. Tobacco use among adults-United States, 2005, 2006
6.Fiore M, Bailey W, Cohen S. Treating tobacco use and dependence: clinical practice guideline. Rockville, MD: U.S. Department of Health and Human Services, Public Health Service, 2000
7.Stein MD, Weinstock MC, Herman DS, Anderson BJ, Anthony JL, Niaura R. A smoking cessation intervention for the methadone-maintained. Addiction 2006;101:599–607
8.Olsen Y, Daumit GL, Ford DE. Opioid prescriptions by U.S. primary care physicians from 1992 to 2001. J Pain 2006;7:225–35
9.Rome JD, Townsend CO, Bruce BK, Sletten CD, Luedtke CA, Hodgson JE. Chronic noncancer pain rehabilitation with opioid withdrawal: comparison of treatment outcomes based on opioid use status at admission. Mayo Clin Proc 2004;79:759–68
10.Richter KP, Gibson CA, Ahluwalia JS, Schmelzle KH. Tobacco use and quit attempts among methadone maintenance clients. Am J Public Health 2001;91:296–9
11.Clemmey P, Brooner R, Chutuape MA, Kidorf M, Stitzer M. Smoking habits and attitudes in a methadone maintenance treatment population. Drug Alcohol Depend 1997;44:123–32
12.Brooner RK, King VL, Kidorf M, Schmidt CW Jr, Bigelow GE. Psychiatric and substance use comorbidity among treatment-seeking opioid abusers. Arch Gen Psychiatry 1997;54:71–80
13.King VL, Kidorf MS, Stoller KB, Carter JA, Brooner RK. Influence of antisocial personality subtypes on drug abuse treatment response. J Nerv Ment Dis 2001;189:593–601
14.Neufeld KJ, Kidorf MS, Kolodner K, King VL, Clark M, Brooner RK. A behavioral treatment for opioid-dependent patients with antisocial personality. J Subst Abuse Treat 2008;34:101–11
15.Schubiner H, Tzelepis A, Milberger S, Lockhart N, Kruger M, Kelley BJ, Schoener EP. Prevalence of attention-deficit/hyperactivity disorder and conduct disorder among substance abusers. J Clin Psychiatry 2000;61:244–51
16.Mello NK, Mendelson JH, Sellers ML, Kuehnle JC. Effects of heroin self-administration on cigarette smoking. Psychopharmacology (Berl) 1980;67:45–52
17.Chait LD, Griffiths RR. Effects of methadone on human cigarette smoking and subjective ratings. J Pharmacol Exp Ther 1984;229:636–40
18.Townsend CO, Bruce BK, Hooten WM, Rome JD. The role of mental health professionals in multidisciplinary pain rehabilitation programs. J Clin Psychol 2006;62:1433–43
19.Hooten WM, Townsend CO, Sletten CD, Bruce BK, Rome JD. Treatment outcomes after multidisciplinary pain rehabilitation with analgesic medication withdrawal for patients with fibromyalgia. Pain Med 2007;8:8–16
20.Hooten WM, Townsend CO, Decker PA. Gender differences among patients with fibromyalgia undergoing multidisciplinary pain rehabilitation. Pain Med 2007;8:624–32
21.Kerns RD, Turk DC, Rudy TE. The West Haven-Yale Multidimensional Pain Inventory (WHYMPI). Pain 1985;23:345–56
22.Multiaxial Assessment of Multidimensional Pain Inventory: Computer Program User’s Manual [computer program]. Version 2.0. Pittsburgh, PA: University of Pittsburgh, 1989
23.Kroon LA. Drug interactions with smoking. Am J Health Syst Pharm 2007;64:1917–21
24.Ackerman WE III, Ahmad M. Effect of cigarette smoking on serum hydrocodone levels in chronic pain patients. J Ark Med Soc 2007;104:19–21
25.Kaila-Kangas L, Leino-Arjas P, Riihimaki H, Luukkonen R, Kirjonen J. Smoking and overweight as predictors of hospitalization for back disorders. Spine 2003;28:1860–8
26.Hellsing AL, Bryngelsson IL. Predictors of musculoskeletal pain in men: a twenty-year follow-up from examination at enlistment. Spine 2000;25:3080–6
27.Fishbain DA, Lewis JE, Cutler RB, Cole B, Rosomoff RS, Rosomoff HL. Does smoking status affect multidisciplinary pain facility treatment outcome? Pain Med (in press)
28.Fishbain DA, Lewis JE, Cole B, Cutler RB, Rosomoff HL, Rosomoff RS. Variables associated with current smoking status in chronic pain patients. Pain Med 2007;8:301–11
29.Cicero TJ, Lynskey M, Todorov A, Inciardi JA, Surratt HL. Co-morbid pain and psychopathology in males and females admitted to treatment for opioid analgesic abuse. Pain 2008;139:127–35
30.Edwards R, McElduff P, Harrison RA, Watson K, Butler G, Elton P. Pleasure or pain? A profile of smokers in Northern England. Public Health 2006;120:760–8
31.Strine TW, Okoro CA, Chapman DP, Balluz LS, Ford ES, Ajani UA, Mokdad AH. Health-related quality of life and health risk behaviors among smokers. Am J Prev Med 2005;28:182–7
32.John U, Hanke M, Meyer C, Volzke H, Baumeister SE, Alte D. Tobacco smoking in relation to pain in a national general population survey. Prev Med 2006;43:477–81
© 2009 International Anesthesia Research Society
33.John U, Alte D, Hanke M, Meyer C, Volzke H, Schumann A. Tobacco smoking in relation to analgesic drug use in a national adult population sample. Drug Alcohol Depend 2006;85:49–55