Relationship Between Smoking and the Clinical Severity of Psoriasis FREE

Psoriasis is a chronic, persistent, lifelong disease with a reported point prevalence of 0.8 to 2.8 in the general population.¹ Psoriasis is characterized by scaly, red cutaneous plaques that contain inflammatory infiltrates and epidermal hyperproliferation.² Although psoriasis is not a life-threatening disease, cure is seldom achieved, and its impact on patient health-related quality of life is dramatic.^{3 - 6} The disability experienced by patients with psoriasis is comparable with that of patients with other chronic illnesses, such as heart disease, diabetes mellitus, cancer, and depression.⁷

Many factors have been associated with causing psoriasis, such as the genetic determinants⁸ and behavioral and environmental factors.^{9 - 14} Although risk factors for psoriasis onset are generally thought to also be linked to exacerbation or to increased clinical severity of psoriasis, to date this assumption has been scarcely tested. To our knowledge, no published epidemiologic study has evaluated the association between tobacco smoking and the clinical severity of psoriasis. The objective of this study is to evaluate the association between different components of smoking history and the clinical severity of psoriasis while controlling for possible confounding factors.

This study was performed between February 21, 2000, and February 19, 2002, within the framework of a large project on the clinical, epidemiologic, emotional, and quality-of-life aspects of psoriasis, the IMPROVE (Istituto Dermopatico dell'Immacolata Multipurpose Psoriasis Research on Vital Experiences) study.¹⁵ A total of 818 inpatients with psoriasis gave their written consent and were included in this analysis. Trained dermatologists from the IMPROVE group collected information on sociodemographic variables, smoking history, and clinical variables, including body mass index (BMI), the presence of chronic diseases (eg, lipid metabolic disorders, diabetes mellitus, renal diseases, hypertension, and asthma), and the clinical severity of psoriasis. Training in how to administrate the questionnaire and how to assess psoriasis severity was performed before the study by a senior staff dermatologist (F.M.).

The clinical severity of psoriasis was assessed using the Psoriasis Area and Severity Index (PASI),¹⁶ which is based on the quantitative assessment of erythema, desquamation, and induration of the plaques combined with the skin surface area involved. To minimize sources of variation, PASI readings were made during the first hours of hospitalization, before medication was administered. The PASI scores were computed using the original formulas.¹⁶ Levels of clinical severity were coded as quartiles of PASI scores to define the 4 categories of the outcome variable (0.0-4.9, 5.0-7.0, 7.1-9.7, and >9.7)

Data on smoking included the age at which cigarette smoking began and ceased and the average amount smoked daily. Smoking status was classified as never, current, and former. Former smokers were defined as those who had stopped smoking at least 1 year before being interview for this study. Intensity (number of cigarettes smoked per day), duration of smoking (years), and time since cessation (years) were categorized. Age at initiation of smoking was represented by a continuous variable. Cigarette-years was calculated as the product of the intensity and duration of smoking.¹⁷ In the analyses that included never smokers, an indicator of ever smoking (yes/no) was systematically included in the model, and all the continuous smoking–related variables (ie, age at initiation and cigarette-years) were centered by subtracting the mean value from the original value for all smokers (current and former) and setting the value to 0 for never smokers.¹⁷

Within 24 to 48 hours of hospital admission, patients completed a questionnaire on sociodemographic variables and several questionnaires yielding patient-centered measures, including the 12-item General Health Questionnaire (GHQ-12). The GHQ-12 is a self-administered questionnaire designed to measure psychological distress and to detect current minor psychiatric nonpsychotic disorders.^{18 - 19} Scores can range from 0 to 36, with higher scores reflecting greater psychological distress. The GHQ-12 was treated as a continuous variable.

Educational attainment was used as a proxy measure of social class and was classified in 3 categories (<6, 6-8, and >8 years). Alcohol consumption (eg, wine, beer, liqueurs, and spirits) was categorized as follows: nondrinkers, occasional drinkers (less than daily), and drinkers (1-2 drinks per day and >2 drinks per day). Family history of psoriasis and present comorbidities were categorized into 2 classes (yes/no). Both BMI and self-reported duration of disease were treated as continuous variables.

Univariate and multivariate odds ratios (ORs) and 95% confidence intervals (CIs) for potential risk factors for psoriasis severity were estimated using the cumulative logit model.²⁰ This model, which allows for the analysis of ordinal dependent variables, is a direct extension of the usual logistic regression model, but rather than applying the logit transformation to the binary response probabilities P (Y = 1 | x), it is applied to the cumulative response probabilities P (Y≤y_j | x). The estimate derived from this model (exp β) can be viewed as a summary OR that is independent of the cutoff points used to classify the outcome.²¹

The roles of smoking-related variables (ie, intensity, duration, time since cessation, age at initiation, and cigarette-years) were first investigated separately, without adjustment for each other or for other potential confounders. We then ran different multiple regression models, each including a different combination of smoking-related variables and potential confounders (ie, sex, age, BMI, GHQ-12, familial psoriasis, alcohol consumption, and self-reported duration of disease). In the first multiple regression model, the smoking variable considered was smoking status (never/current/former). In the second model, intensity and duration of smoking were considered as 2 separate variables, and the analysis was restricted to current smokers only to avoid confounding by time since cessation in former smokers. In the third model, intensity and duration of smoking were replaced by cigarette-years. Because we included never and former smokers in this analysis, we added an indicator for ever smoking, centered the cigarette-years variable, and adjusted for time since cessation. Because a statistical interaction was observed between cigarette-years and sex (P = .03), we then ran the third model in men and women separately, which formed models 4 and 5, respectively. No statistical interaction between alcohol consumption and smoking was detected. The Wald test for trend was performed for each ordinal variable. All analyses were performed using a statistical software package (Stata 6.0; StataCorp, College Station, Tex).

Of the 818 patients with psoriasis (mean ± SD age, 46.8 ± 16.0 years), 62% were men and 38% were women, and the mean ± SD PASI score was 8.6 ± 5.9. Table 1 reports the associations between psoriasis severity and sex, age, education, family history of psoriasis, presence of chronic diseases, alcohol consumption, BMI, psychological distress, and duration of disease. Alcohol intake (>2 glasses daily) was associated with twice the risk of more severe psoriasis. Family history of psoriasis was also associated with increased severity of psoriasis. Table 2 depicts the association between smoking history and severity of psoriasis. Patients with more severe psoriasis (PASI score >9.7) tended to smoke more cigarettes per day (mean ± SD, 24.1 ± 14.6) than patients with less severe psoriasis (mean ± SD, 15.1 ± 9.4). Intensity and duration of smoking were associated with psoriasis severity.

Table 3 reports the results of regression models 1 to 5 (see the “Statistical Analysis” subsection). After adjustment for sex, age, familial psoriasis, duration of disease, BMI, alcohol consumption, and psychological stress (GHQ-12), the OR for current and former smokers became largely nonsignificant compared with never smokers (model 1). However, intensity of smoking was associated with psoriasis severity among current smokers (Wald test for trend: P = .007). Specifically, patients who smoked more than a pack of cigarettes (>20 cigarettes) daily had twice the risk of more severe psoriasis compared with those who smoked 10 cigarettes or less per day (model 2) (Table 3). In all smokers (including current and former smokers), cigarette-years was associated with a 30% increased risk of more severe psoriasis for a 600-U increase (which corresponds, eg, to 20 cigarettes per day for 30 years). Models 4 and 5 indicate that smoking had a much stronger impact in women than in men. Although none of the smoking-related variables (ever smoking, cigarette-years, and time since quitting) had a significant effect in men (model 4), model 5 shows that in women, average current and recent former smokers had a 72% higher risk of more severe psoriasis than never smokers. As opposed to men, the effect of cigarette-years was significant in women (OR, 1.8; 95% CI, 1.2-2.6). In men and women, time since quitting was not associated with a reduced risk of more severe psoriasis.

Cigarette smoking is a risk factor for many chronic diseases, including psoriasis, and it has been widely considered as an important preventable cause of morbidity. However, little is known about the effect of smoking on psoriasis severity. This is the first epidemiologic study suggesting that cigarette smoking is associated with the clinical severity of psoriasis. In the present study, various aspects of smoking history were investigated in relation to the clinical severity of psoriasis after controlling for possible confounding factors, such as sex, age, BMI, psychological distress, family history of psoriasis, duration of disease, and alcohol consumption.

It has long been known that smoking induces functional and morphologic alterations in polymorphonuclear leukocytes, and it may also cause an exaggerated release of chemotactic factors.²¹ Some studies^{22 - 25} have shown that cigarette smoking induces an overproduction of interleukin 1β and increases the production of tumor necrosis factor α and transforming growth factor β, which have been associated with psoriasis severity.

In this study, when combining men and women, smoking status (never/current/former) after adjustment for potential confounders was no more statistically associated with an increased risk of psoriasis severity. Age at initiation of smoking did not seem to be related to psoriasis severity. Among current smokers, an increased risk was observed for the number of cigarettes smoked per day but not for the duration of smoking. This result is consistent with a previous etiologic case-control study²⁶ that suggested that the intensity, but not the duration, of smoking was associated with an increased risk of psoriasis. When including former smokers in the analysis and adjusting for time since cessation, the effect of cigarette-years was significant (a 600-U increase in cigarette-years was associated with an increase of 30% in psoriasis severity).

Separate analyses for men and women showed that the effect of cigarette-years on psoriasis severity was stronger for women than for men. This result is consistent with 2 etiologic studies^{12 ,27} that found that smoking was associated with an increased risk of psoriasis in women but not in men. Poikolainen and colleagues²⁷ conducted a study on 55 female patients with psoriasis and 108 controls and showed that the risk of psoriasis for those smoking 20 cigarettes daily compared with nonsmokers was 3.3 (95% CI, 1.4-7.9). Moreover, indirect evidence of a possible role of smoking on psoriasis severity in females was suggested by a large cohort study²⁸ (n = 17 032) that showed a 2-fold increase in the risk of hospital referral for psoriasis in females who smoke compared with nonsmokers. Because most skin disorders in England and Scotland are managed in the primary care setting, it seems that psoriasis among females who smoke is more severe.²⁸

In men, family history of psoriasis was a stronger prognostic factor for psoriasis severity than smoking. Our findings are in agreement with those of a case-control study²⁶ conducted on 215 patients with psoriasis and 267 controls that showed that family history of psoriasis was a major risk factor for psoriasis in men after controlling for alcohol intake and smoking (OR, 18.8; 95% CI, 6.4-54.8). It is possible that for men, family history of psoriasis may be the most important prognostic factor not only for psoriasis risk but also for psoriasis severity. In summary, although previous studies have shown that the impact of smoking on the risk of psoriasis is different in males and in females, our findings suggest that this difference persists when looking at the risk of having more severe psoriasis.

Some previous studies have suggested that psoriasis is associated with heavy drinking,^{29 - 30} and others²⁶ have found a dose-response relationship. In the present study, drinking alcohol was associated with an increased risk of psoriasis severity, but it was not statistically significant after controlling for confounding variables. Our results are in agreement with those of 2 previous studies^{12 ,27} that showed no relation between the extent of body area affected by psoriasis and alcohol intake. Psychological distress has also been suggested to interfere in a negative way with the outcome of psoriasis treatment.⁹ In the present study, psychological distress was not associated with the clinical severity of psoriasis.

This study has some strengths and limitations. This is the first study on psoriasis with such a large sample size and that carefully controlled for important potential confounding factors, such as sex, age, BMI, psychological distress, family history of psoriasis, duration of disease, and alcohol consumption, to show that cigarette smoking worsens psoriasis. A limitation of this study is that only inpatients were included. However, the hospitalization pattern at Istituto Dermopatico dell'Immacolata, Istituto di Ricovero e Cura a Carattere Scientifico, Rome, Italy, is such that our results could probably be generalized to dermatologic outpatients in other health systems where hospitalization for skin diseases is rare. The possible generalizability of our findings to outpatients is corroborated by the frequent occurrence of low PASI scores in our sample. Finally, the results of this study suggest a negative effect of cigarette smoking on the severity of psoriasis, in particular in women. These findings might be of clinical importance because they would support the dermatologist’s recommendation to patients with psoriasis to quit smoking to prevent worsening of their psoriasis. Also, in clinical trials and in observational studies of drug efficacy in psoriasis, smoking history should be taken into consideration as a possible modifier of treatment effects.

Correspondence: Cristina Fortes, PhD, Clinical Epidemiology Unit, Istituto Dermopatico dell'Immacolata, IDI-IRCCS, Via dei Monti di Creta 104, 00167 Roma, Italia (c.fortes@idi.it).

Accepted for Publication: July 27, 2005.

Author Contributions:Study concept and design: Fortes, Melchi, and Abeni. Analysis and interpretation of data: Fortes, Mastroeni, Leffondré. Drafting of the manuscript: Fortes. Critical revision of the manuscript for important intellectual content: Mastroeni, Leffondré, Sampogna, Melchi, Mazzotti, Pasquini, and Abeni. Statistical analysis: Fortes and Mastroeni. Obtained funding: Pasquini and Abeni. Study supervision: Fortes, Melchi, and Abeni.

Group Members: The IMPROVE investigators who contributed to this study are as follows: Massimo Alotto, MD,* Gianluca Antonelli, MD,* Simone Bolli, Rino Cavalieri, MD, Massimo Luca Chinni, MD, Marcello Fazio, MD, Giampiero Girolomoni, MD, Elisabetta Luchetti, Nidia Melo Salcedo, Paola Moscatelli, MD,* Paolo Piazza, MD, Angelo Picardi, MD, Orietta Picconi, MSc, Maria Antonietta Pilla, MD,* Grazia Primavera, MD,* Pietro Puddu, MD, Paolo Ruatti, MD, Giuseppe Ruggiero, MD, Valentina Salvatori, Francesco Sera, MSc, Romeo Simoni, MD, Donatella Sordi, MD,* Stefano Tabolli, MD, and Albertina Tiago, MD.* Asterisk indicates dermatologists who were trained and collected data in the study.

Financial Disclosure: None.

Funding/Support: This study was supported by the Istituto Dermopatico dell'Immacolata, Istituto di Ricovero e Cura a Carattere Scientifico, and by the Progetto Ricerca Corrente 2003, Italian Ministry of Health (data collection and management), Rome.

Previous Presentation: This study was presented in part at the European Congress of Epidemiology; September 10, 2004; Porto, Portugal.