This cohort study examines the association between changes in smoking status and the development of hidradenitis suppurativa.
Key Points
Question
Does smoking cessation reduce the risk of hidradenitis suppurativa (HS) development?
Findings
In this nationwide cohort study of 6 230 189 Korean participants, quitting smoking was associated with a statistically significant lowered risk of HS onset compared to continuous smoking; conversely, those who initially quit smoking but resumed and those who had no previous smoking history but started had similar risks of HS onset as sustained smokers. Time-smoking interaction analyses revealed a rapid reduction in HS risk after approximately 3 years of smoking cessation, resembling levels of those with a history of smoking but who maintained cessation and those who reported never smoking.
Meaning
Quitting smoking and maintaining a smoking-free status may reduce the risk of HS onset, suggesting the importance of smoking cessation and refraining from initiating or resuming smoking for HS prevention.
Abstract
Importance
Although tobacco smoking is established as a risk factor for hidradenitis suppurativa (HS), studies on the effects of smoking cessation on HS are limited, and evidence is lacking.
Objective
To examine the association between changes in smoking status and the development of HS.
Design, Setting, and Participants
This population-based cohort study enrolled participants from the Korean National Health Insurance Service database who had undergone 2 consecutive biennial health examinations (2004-2005 and 2006-2007) as the primary cohort. Within the primary cohort, the secondary cohort comprised individuals who underwent all biennial health examinations throughout the follow-up period and maintained the same smoking status from 2006 to 2007 to the end of the follow-up period. Data were analyzed from July to December 2023.
Exposures
Changes in smoking habit status.
Main Outcomes and Measures
Risk of HS development. The HS risk according to change in smoking status between the 2 consecutive health examinations was estimated using a Cox proportional hazards model.
Results
Of the 6 230 189 participants enrolled, the mean (SD) age was 47.2 (13.5) years, and 55.6% were male. During 84 457 025 person-years of follow-up, 3761 HS events occurred. In the primary cohort, compared to those who consistently reported active smoking at both checkups (ie, sustained smokers), lower HS risk was seen among those who were confirmed to smoke initially but quit by the second checkup (ie, smoking quitters) (adjusted hazard ratio [AHR], 0.68; 95% CI, 0.56-0.83), those who maintained cessation status throughout (AHR, 0.67; 95% CI, 0.57-0.77), and those who reported never smoking at either checkup (ie, never smokers) (AHR, 0.57; 95% CI, 0.52-0.63). Those who initially quit smoking but resumed by the second checkup and those who had no previous smoking history but started at the second checkup (ie, new smokers) exhibited similar HS risk as sustained smokers. The secondary cohort results aligned with those of the primary cohort, showing a more pronounced risk reduction with smoking cessation (AHR, 0.57; 95% CI, 0.39-0.83). Considering time-smoking interaction, the cumulative incidence and the risk of HS in smoking quitters were similar to those in sustained smokers in the early stages of observation. However, 3 to 4 years after smoking cessation, the rate decelerated, resembling that of never smokers, and there was a statistically significant decrease in the risk that persisted (between 3 and 6 years from the index date: AHR, 0.58; 95% CI, 0.36-0.92; and ≥12 years from the index date: AHR, 0.70; 95% CI, 0.50-0.97). New smokers initially paralleled never smokers but accelerated after 2 to 3 years, reaching sustained smokers’ levels.
Conclusions and Relevance
In this cohort study, quitting smoking and sustaining a smoke-free status were associated with a reduced risk of HS development compared to continuous smoking. In contrast, resuming or initiating smoking may have as detrimental an effect on HS development as continual smoking.
Introduction
Hidradenitis suppurativa (HS), a chronic inflammatory skin disorder characterized by painful nodules, abscesses, and tunnel formation in intertriginous areas, poses a substantial burden on affected individuals.1 Its cause involves multiple factors, including genetic predisposition, immune dysregulation, and environmental factors implicated in its development and progression.2,3,4 Among these environmental factors, cigarette smoking has gained interest due to its potential role in both developing and exacerbating the disease.5,6
Recently, there has been a growing recognition of the associations between smoking and dermatological disorders, including psoriasis, alopecia, lupus erythematosus, and skin cancer.7 Among them, the association of smoking with HS is receiving more attention, particularly due to the chronicity and detrimental natural course of the condition. Smoking has been consistently associated with an elevated risk of HS, exacerbation of the condition, and reduced treatment effectiveness.5,8,9 Most of the studies showing these results focused on classifying individuals into those who currently, formerly, or never smoked based solely on their smoking status at a single time point.
In daily clinical practice, it is crucial to understand how changes in smoking habits, such as starting or quitting smoking, affect the development of HS. However, research on the effect of changes in smoking habits over time on the development of HS is scarce. Moreover, a substantial portion of research investigating the association between smoking and HS has been conducted in Western populations, which has resulted in a relative scarcity of evidence regarding Asian populations.5,10 Therefore, there is a critical need for longitudinal studies that observe the consequences of starting or quitting smoking from a particular starting point and that investigate whether it truly has any affect on HS risk.11 Using a nationwide, large-scale database from the Korean National Health Insurance Service (NHIS), we conducted a 14-year longitudinal cohort study to explore the associations of changes in smoking habits with the development of HS.
Methods
Study Population
The NHIS provides an extensive range of medical services, including hospital admissions, outpatient visits, medical procedures, and prescription medications.12,13 As part of its health care services, the NHIS conducts biennial health screenings for insured individuals 20 years and older. These health screenings include anthropometric parameter measurements, laboratory test results, and information on health behaviors.
This study enrolled participants from the entire Korean population who underwent 2 consecutive biennial health examinations: the first checkup was between 2004 and 2005, and the second checkup was between 2006 and 2007. We excluded individuals with a history of HS, those who died, and those whose records related to smoking or other factors were incomplete prior to January 1, 2008. Subsequently, the remaining eligible participants were established as the primary cohort. Among them, to establish a more robust study population cohort, a secondary cohort was constructed by enrolling only participants who underwent consecutive biennial health examinations throughout the entire follow-up period (≤14 years) and maintained the same smoking status from the second checkup period (2006-2007) to the end of the follow-up period in the study. Figure 1 illustrates the selection process for the study population.
Figure 1. Flow Diagram for the Selection of the Study Participants.
This study was approved by the institutional review board of Seoul National University Hospital. The requirement for explicit patient consent was waived because the NHIS cohort database adhered to stringent confidentiality protocols, rendering the data anonymous. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.
Key Variables
The participants’ smoking habits were evaluated using a self-reported questionnaire. Subsequently, the study participants were divided into 6 groups based on their responses at both the initial and follow-up health checkups: (1) those who consistently reported active smoking at both checkups (ie, sustained smokers), (2) those who initially quit smoking but resumed by the second checkup (ie, relapsed smokers), (3) those who had no previous smoking history but started at the second checkup (ie, new smokers), (4) those who were confirmed to smoke initially but quit by the second checkup (ie, smoking quitters), (5) those who maintained cessation status throughout (ie, sustained ex-smokers), and (6) those who reported never smoking at either checkup (ie, never smokers) (eTable 1 in Supplement 1).
Data from hospitals and clinics gathered by the NHIS were used to detect cases of HS. Following the guidelines of the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10), coding system, we used ICD-10 codes to confirm HS events based on the medical records. To ensure a rigorous definition of HS, we identified patients who had received medical consultations from a physician on at least 3 separate documented visits with a confirmed diagnosis of HS (ICD-10 code L73.2) between January 1, 2008, and December 31, 2021.
Statistical Analysis
The continuous follow-up of all participants in this cohort began on January 1, 2008, and ended either on the occurrence of an HS event, death, or December 31, 2021, whichever came first. To evaluate the hazard ratios (HRs) and 95% CIs for HS according to changes in smoking habits, we performed a Cox proportional hazard regression analysis after adjusting for potential confounding factors, such as age, sex, household income, body mass index, level of physical activity, alcohol consumption, systolic blood pressure, fasting serum glucose, total cholesterol, and Charlson Comorbidity Index (CCI). Moreover, we conducted subgroup analyses by categorizing participants based on confounding factors.
To control rigorously for potential changes in smoking status after 2007 due to the time-varying characteristics of smoking, we conducted sensitivity analyses using data from the secondary cohort, which included participants who underwent all consecutive biennial health examinations throughout the follow-up period. Sensitivity analyses were performed specifically for individuals whose smoking habits remained consistent from the second health examination to the end of the follow-up period (≤14 years).
Furthermore, to investigate the acute or cumulative effect of smoking initiation or cessation on the risk of HS over time, we conducted an extended stratified Cox regression analysis, wherein the model accounted for the interaction between time and smoking status. The observation period was subdivided into 5 distinct periods (3-year intervals). The HRs for each period were estimated to comprehensively examine the changes in the risk of HS over time.
Statistical significance was defined as P < .05 in a 2-sided test. We used SAS, version 9.4 (SAS Institute), for the analyses, data collection, and data mining. Data were analyzed from July to December 2023.
Results
Of the 7 706 421 participants who underwent biennial health checkups consecutively from 2004 to 2005 and from 2006 to 2007 in South Korea, 6 230 189 who did not meet the exclusion criteria were enrolled in this study as the primary cohort (Figure 1). Among these, 1 905 499 participants who underwent all biennial health checkups throughout the follow-up period and maintained a consistent smoking status until the end of the follow-up period were enrolled in the secondary cohort. The detailed characteristics of the participants are summarized in Table 1 and eTable 2 in Supplement 1. In the primary cohort, smoking quitters tended to be older and had higher physical activity levels; higher body mass index; increased household income; reduced alcohol consumption; a higher prevalence of hypertension, diabetes, and dyslipidemia; and a higher CCI than sustained smokers. Similar tendencies were observed in the secondary cohort. The household income of patients with HS was comparable with those without HS (eTable 3 in Supplement 1), with no statistically significant differences in HS risk based on income levels (eTable 4 in Supplement 1).
Table 1. General Characteristics of the Study Population.
| Characteristic | No. (%) | ||||||
|---|---|---|---|---|---|---|---|
| Total | Smoking status | Smoking cessation | Never | ||||
| Sustained | Relapsed | New | Quit | Sustained former | |||
| No. of people | 6 230 189 | 1 431 380 | 121 679 | 209 822 | 163 451 | 321 052 | 3 982 805 |
| Age, mean (SD), y | 47.2 (13.5) | 43.2 (11.8) | 43.5 (11.7) | 44.1 (12.7) | 44.5 (11.9) | 47.0 (11.2) | 49.0 (14.0) |
| Sex | |||||||
| Female | 2 765 893 (44.4) | 35 261 (2.5) | 4253 (3.5) | 17 101 (8.1) | 5681 (3.5) | 9502 (3.0) | 2 694 095 (67.6) |
| Male | 3 464 296 (55.6) | 1 396 119 (97.5) | 117 426 (96.5) | 192 721 (91.9) | 157 770 (96.5) | 311 550 (97.0) | 1 288 710 (32.4) |
| BMI, mean (SD) | 23.6 (3.3) | 23.9 (3.1) | 24.0 (3.0) | 23.9 (3.1) | 24.4 (2.9) | 24.3 (2.7) | 23.4 (3.4) |
| Cigarettes smoked per d | |||||||
| <10 | NA | 25 660 (17.9) | 45 090 (38.0) | 58 888 (28.8) | NA | NA | NA |
| 10-19 | NA | 833 692 (59.1) | 58 605 (49.3) | 107 921 (52.7) | NA | NA | NA |
| 20-39 | NA | 310 773 (22.0) | 14 401 (12.1) | 35 737 (17.5) | NA | NA | NA |
| ≥40 | NA | 14 486 (1.0) | 715 (0.6) | 2038 (1.0) | NA | NA | NA |
| Smoking history, y | |||||||
| <10 | NA | 221 689 (16.1) | 28 690 (24.8) | NA | 35 785 (27.5) | 92 533 (35.6) | NA |
| 10-19 | NA | 585 241 (42.5) | 45 396 (39.3) | NA | 47 945 (36.8) | 97 263 (37.4) | NA |
| ≥20 | NA | 571 605 (41.4) | 41 444 (35.9) | NA | 46 415 (35.7) | 70 506 (27.0) | NA |
| Household income by quartile | |||||||
| 1st (highest) | 2 273 484 (36.5) | 487 291 (34.0) | 50 454 (41.5) | 72 830 (34.7) | 71 105 (43.5) | 172 431 (53.7) | 1 419 373 (35.6) |
| 2nd | 1 782 394 (28.6) | 483 449 (33.8) | 38 311 (31.5) | 66 863 (31.9) | 49 486 (30.3) | 84 290 (26.3) | 1 059 995 (26.6) |
| 3rd | 1 183 393 (19.0) | 272 946 (19.1) | 19 434 (16.0) | 40 656 (19.4) | 24 274 (14.9) | 36 559 (11.4) | 789 524 (19.8) |
| 4th (lowest) | 990 918 (15.9) | 187 694 (13.1) | 13 480 (11.0) | 29 473 (14.0) | 18 586 (11.3) | 27 772 (8.6) | 713 913 (17.8) |
| Physical activity, times/wk | |||||||
| 0 | 3 106 724 (49.9) | 620 901 (43.4) | 48 191 (39.6) | 93 723 (44.7) | 59 511 (36.4) | 98 665 (30.7) | 2 185 733 (54.9) |
| 1-2 | 1 878 898 (30.2) | 561 541 (39.2) | 48 074 (39.5) | 78 525 (37.4) | 63 530 (38.9) | 129 495 (40.3) | 997 733 (25.1) |
| 3-4 | 755 877 (12.1) | 162 148 (11.3) | 16 924 (13.9) | 23 784 (11.3) | 26 777 (16.4) | 61 559 (19.2) | 464 685 (11.7) |
| ≥5 | 488 690 (7.8) | 86 790 (6.1) | 8490 (7.0) | 13 790 (6.6) | 13 633 (8.3) | 31 333 (9.8) | 334 654 (8.3) |
| Alcohol intake, times/wk | |||||||
| 0 | 3 154 211 (50.6) | 274 347 (19.2) | 20 821 (17.1) | 49 250 (23.5) | 35 476 (21.7) | 61 310 (19.1) | 2 713 007 (68.1) |
| 1-2 | 2 549 870 (40.9) | 889 696 (62.2) | 81 637 (67.1) | 126 046 (60.1) | 104 428 (63.9) | 211 634 (65.9) | 1 136 429 (28.5) |
| 3-4 | 380 645 (6.1) | 195 564 (13.6) | 14 493 (11.9) | 24 869 (11.9) | 17 857 (10.9) | 36 197 (11.3) | 91 665 (2.3) |
| ≥5 | 145 463 (2.4) | 71 773 (5.0) | 4728 (3.9) | 9657 (4.5) | 5690 (3.5) | 11 911 (3.7) | 41 704 (1.1) |
| Systolic blood pressure, mean (SD), mm Hg | 122.8 (15.6) | 124.2 (14.3) | 123.4 (14.1) | 123.5 (14.5) | 124.7 (14.3) | 125.2 (14.3) | 121.9 (16.2) |
| <129 | 3 954 437 (63.5) | 859 848 (60.1) | 75 540 (62.1) | 128 442 (61.2) | 96 506 (59.0) | 183 784 (57.2) | 2 610 317 (65.5) |
| ≥130 | 2 275 752 (36.5) | 571 532 (39.9) | 46 139 (37.9) | 81 380 (38.8) | 66 945 (41.0) | 137 268 (42.8) | 1 372 488 (34.5) |
| Fasting serum glucose, mean (SD), mg/dL | 95.1 (23.0) | 96.6 (25.7) | 96.0 (23.2) | 96.0 (25.5) | 97.2 (24.3) | 97.3 (21.7) | 94.3 (21.7) |
| <99 | 4 526 962 (72.7) | 1 000 982 (69.9) | 85 792 (70.5) | 149 253 (71.1) | 111 081 (68.0) | 213 791 (66.6) | 2 966 063 (74.5) |
| ≥100 | 1 703 225 (27.3) | 430 398 (30.1) | 35 887 (29.5) | 60 567 (28.9) | 52 370 (32.0) | 107 261 (33.4) | 1 016 742 (25.5) |
| Total cholesterol, mean (SD), mg/dL | 194.0 (36.6) | 194.3 (36.5) | 192.7 (35.5) | 193.5 (36.3) | 196.8 (36.4) | 195.7 (35.1) | 193.7 (36.9) |
| <199 | 3 687 774 (59.2) | 840 688 (58.7) | 73 792 (60.6) | 125 062 (59.6) | 91 434 (55.9) | 182 989 (57.0) | 2 373 809 (59.6) |
| ≥200 | 2 542 415 (40.8) | 590 710 (41.3) | 47 886 (39.4) | 84 760 (40.4) | 72 016 (44.1) | 138 063 (43.0) | 1 608 980 (40.4) |
| Hypertension | 1 758 879 (28.2) | 329 025 (23.0) | 28 638 (23.5) | 50 013 (23.8) | 44 132 (27.0) | 98 174 (30.6) | 1 208 897 (30.4) |
| Diabetes | 882 219 (14.2) | 173 019 (12.1) | 14 908 (12.3) | 27 184 (13.0) | 21 997 13.5) | 45 940 (14.3) | 599 171 (15.0) |
| Dyslipidemia | 1 392 105 (22.3) | 271 314 (19.0) | 23 548 (19.4) | 40 982 (19.5) | 37 257 (22.8) | 76 433 (23.8) | 942 571 (23.7) |
| Charlson Comorbidity Index | |||||||
| 0 | 3 078 758 (49.4) | 850 900 (59.5) | 67 755 (55.7) | 117 946 (56.2) | 87 597 (53.6) | 166 716 (51.9) | 1 787 844 (44.9) |
| 1 | 1 846 242 (29.6) | 383 331 (26.8) | 34 909 (28.7) | 57 862 (27.6) | 46 548 (28.5) | 94 750 (29.5) | 1 228 842 (30.9) |
| ≥2 | 1 305 189 (21.0) | 197 149 (13.7) | 19 015 (15.6) | 34 014 (16.2) | 29 306 (17.9) | 59 586 (18.6) | 966 119 (24.2) |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); NA, not applicable.
SI conversion factor: To convert total cholesterol to millimoles per liter, multiply by 0.0259; to convert serum glucose to millimoles per liter, multiply by 0.0555.
A total of 3761 cases of HS were detected during the 84 457 025 person-years of observation. Table 2 presents the association between changes in smoking habits and the development of HS. In the primary cohort, relapsed smokers and new smokers did not show a statistically significant difference in the risk of HS development compared with sustained smokers. In contrast, smoking quitters (adjusted HR [AHR], 0.68; 95% CI, 0.56-0.83), sustained ex-smokers (AHR, 0.67; 95% CI, 0.57-0.77), and never smokers (AHR, 0.57; 95% CI, 0.52-0.63) had a reduced risk of developing HS compared to sustained smokers. While there was a remarkable trend of reduced HS risk associated with tobacco consumption decline from sustained smokers to never smokers, no statistically significant differences in HS risk were observed among sustained smokers, relapsed smokers, and new smokers, and no statistically significant difference in HS risk was observed between smoking quitters and sustained ex-smokers. The results from the secondary cohort were consistent with those obtained in the primary cohort. Notably, the association of smoking cessation with HS was even more pronounced (AHR, 0.57; 95% CI, 0.39-0.83).
Table 2. Adjusted Hazard Ratios (AHRs) for Hidradenitis Suppurativa (HS) According to the Change in Smoking Habit.
| Categorya,b | Total HS events, No. | Total person-years observed | AHR (95% CI)c |
|---|---|---|---|
| Primary cohort | |||
| Change in smoking status between health examinations | |||
| Sustained smokers | 1381 | 19 330 308 | 1 [Reference] |
| Relapsed smokers | 107 | 1 651 601 | 0.90 (0.74-1.09) |
| New smokers | 203 | 2823 344 | 1.02 (0.88-1.19) |
| Smoking cessation | |||
| Smoking quitters | 110 | 2 216 415 | 0.68 (0.56-0.83) |
| Sustained ex-smokers | 205 | 4 372 546 | 0.67 (0.57-0.77) |
| Never smokers | 1755 | 54 062 811 | 0.57 (0.52-0.63) |
| Secondary cohort | |||
| Change in smoking habits between health examinations | |||
| Sustained smokers | 631 | 3 132 025 | 1 [Reference] |
| Relapsed smokers | 31 | 130 102 | 1.17 (0.82-1.68) |
| New smokers | 74 | 337 216 | 1.09 (0.86-1.38) |
| Smoking cessation | |||
| Smoking quitters | 28 | 241 378 | 0.57 (0.39-0.83) |
| Sustained ex-smokers | 94 | 872 000 | 0.53 (0.43-0.66) |
| Never smokers | 1035 | 20 753 991 | 0.30 (0.26-0.34) |
The primary cohort comprised participants from the Korean National Health Insurance Service database who had undergone 2 consecutive biennial health examinations (2004-2005 and 2006-2007). Among participants in the primary cohort, those who consecutively underwent biennial health checkups throughout the entire follow-up period and maintained a consistent smoking status from the second health examination until the end of the follow-up were enrolled in the secondary cohort.
Those who consistently reported active smoking at both checkups were considered sustained smokers, those who initially quit smoking but resumed by the second checkup were considered relapsed smokers, those who had no previous smoking history but started at the second checkup were considered new smokers, those who were confirmed to smoke initially but quit by the second checkup were considered smoking quitters, those who maintained cessation status throughout were considered sustained ex-smokers, and those who reported never smoking at either checkup were considered never smokers.
Calculated by Cox hazards regression analysis after adjusting for age, sex, household income, body mass index, physical activity, alcohol intake, systolic blood pressure, fasting serum glucose, total cholesterol, and Charlson Comorbidity Index.
Kaplan-Meier curves depicting the cumulative incidence of HS exhibited distinct patterns based on smoking status (Figure 2). In the primary cohort, the increase in HS incidence was steeper in sustained smokers than in never smokers, depending on the follow-up duration. Remarkably, among smoking quitters, there was a similar increase in the cumulative incidence of HS as observed in sustained smokers for approximately 3 to 4 years after cessation. However, thereafter, the rate of increase decelerated, resembling the trend observed in never smokers. On the other hand, new smokers displayed a pattern where, after initiating smoking, the rate of increase in the cumulative incidence of HS was similar to that of never smokers for approximately 2 to 3 years. However, after this period, the rate accelerated, reaching levels comparable to those observed in sustained smokers. These trends were consistent in the secondary cohort. Notably, after a relatively short period of 2 to 4 years, the acute results of smoking initiation or cessation became evident.
Figure 2. Cumulative Incidences of Hidradenitis Suppurativa Based on Smoking Status.
The primary cohort comprised participants from the Korean National Health Insurance Service database who had undergone 2 consecutive biennial health examinations (2004-2005 and 2006-2007). Within the primary cohort, the secondary cohort comprised individuals who underwent all biennial health examinations throughout the follow-up period and maintained the same smoking status from 2006 to 2007 to the end of the follow-up period. Those who consistently reported active smoking at both checkups were considered sustained smokers, those who had no previous smoking history but started at the second checkup were considered new smokers, those who were confirmed to smoke initially but quit by the second checkup were considered smoking quitters, and those who reported never smoking at either checkup were considered never smokers.
The risk of developing HS exhibited a time-dependent change based on smoking status (Figure 3). In the extended Cox model, considering the interaction between time and smoking status, no statistically significant difference in the risk of developing HS was observed among smoking quitters compared to sustained smokers at the beginning of the study period (between 0 and 3 years from the index date). However, after 3 years, there was a statistically significant decrease in the risk of developing HS among smoking quitters compared to sustained smokers, and this trend persisted over time (between 3 and 6 years from the index date: AHR, 0.58; 95% CI, 0.36-0.92; between 6 and 9 years from the index date, AHR, 0.55; 95% CI, 0.36-0.84; and ≥12 years from the index date, AHR, 0.70; 95% CI, 0.50-0.97). The secondary cohort presented similar results with diminished statistical significance after 9 years, which may be due to the cohort’s inherent nature of smaller study population and HS events.
Figure 3. Forest Plots of the Risk of Hidradenitis Suppurativa According to the Observation Period Stratified by Smoking Status.
The primary cohort comprised participants from the Korean National Health Insurance Service database who had undergone 2 consecutive biennial health examinations (2004-2005 and 2006-2007). Within the primary cohort, the secondary cohort comprised individuals who underwent all biennial health examinations throughout the follow-up period and maintained the same smoking status from 2006 to 2007 to the end of the follow-up period. Those who initially quit smoking but resumed by the second checkup were considered relapsed smokers, those who had no previous smoking history but started at the second checkup were considered new smokers, those who were confirmed to smoke initially but quit by the second checkup were considered smoking quitters, those who maintained cessation status throughout were considered sustained ex-smokers, and those who reported never smoking at either checkup were considered never smokers. AHR indicates adjusted hazard ratio.
aP < .05.
bP < .01.
cP < .001.
Across various covariate subgroups, including age, weight status, alcohol intake, physical activity, CCI, systolic blood pressure, fasting serum glucose, and total cholesterol, the risks of HS were consistently reduced in smoking quitters, sustained ex-smokers, and never smokers compared with sustained smokers (eFigures 1 and 2 in Supplement 1). When stratified by sex, the risk of HS reduction was observed in smoking quitters, sustained ex-smokers, and never smokers among both male and female individuals, but statistical significance was attenuated among female smoking quitters, probably due to the small event size.
Discussion
Given the latest data revealing a surge in HS in Western countries,14,15,16 the importance of lifestyle modification for disease prevention is further emphasized. This extensive long-term cohort study provides empirical evidence establishing that smoking cessation was associated with a decreased risk of developing HS compared to continuous smoking. Notably, approximately 3 years after quitting smoking, the risk-reduction effect was comparable to that in individuals maintaining smoking cessation, described as sustained ex-smokers. Conversely, individuals resuming smoking after quitting or those starting smoking anew, even in the early phase of the observation period, showed no statistically significant difference in HS risk compared to sustained smokers. To our knowledge, this is the first study to investigate the associations of changes in smoking habits with the subsequent development of HS.
Although prior research has not investigated the correlation between changes in smoking habits and HS, cross-sectional studies have established an association between HS and smoking at a single time. A systematic review and meta-analysis of 25 studies revealed a notable association between current smoking status and HS development, even though most of these studies were conducted in Western countries.5 Unlike in Western countries, studies investigating the association between smoking and HS in an Asian population are relatively scarce.10 The scarcity of studies in Asian populations may have been influenced by the notion that the prevalence of smoking among patients with HS in Asia appears to be lower compared to that reported in Western countries. In Western populations, previous studies indicated high smoking rates among patients with HS, reaching up to 90%.17 However, in East Asia, the reported smoking rates were relatively lower: 29% in a Japanese study,18 34% to 38.3% in Korean studies,19,20 and 32.8% in Singapore.21 For these reasons, 1 study suggested that the underlying factors associated with HS and smoking in East Asia may differ from those in Western countries.10 The present results, however, indicate that individuals who never smoked had a lower risk of developing HS compared to those who continued smoking, aligning with prior Western studies. Simultaneously, this study provides valuable evidence that smoking can be a risk factor for the development of HS in Asian populations.
In a regular practice setting with long-term observation, we found that those who quit smoking had a reduced risk of HS development. This implies that quitting smoking may hinder the onset of HS and could have a substantial affect on HS prevention. An interesting point is that the HS risk-reduction effect observed in smoking quitters was as remarkable as the HS risk reduction observed in sustained ex-smokers. This suggests that smoking may have a fairly acute effect on HS pathogenesis, highlighting the importance of quitting smoking as quickly as possible. Indeed, these findings show that a brief period of about 3 years after smoking cessation can have a statistically significant contribution to lowering the incidence of HS over time and reducing the risk of HS as much as maintaining a smoke-free status. These results indicate the importance of encouraging smokers to quit smoking and maintain a smoke-free lifestyle as a preventive measure against HS development.
The findings also revealed that initiating smoking, whether by ex-smokers or never smokers, did not result in a statistically significant difference in HS risk compared to sustained smoking. This aligns with the hypothesis of smoking exerting a relatively immediate effect on HS pathogenesis, suggesting that simply initiating smoking may considerably contribute to increasing HS risk, comparable to sustained smoking. Indeed, these results indicate that initiating smoking, even from an early phase, can accelerate the development of HS as rapidly as sustained smoking. Furthermore, new smokers or relapsed smokers showed no statistically significant HS risk difference compared to sustained smokers, even within the initial 3 years. Therefore, for HS prevention, it is crucial to avoid restarting or initiating smoking anew, as both behaviors seem no less harmful than continued smoking. Indeed, never smokers demonstrated consistently stronger and more robust health benefits compared to sustained ex-smokers across both the main and subgroup analyses, implying that abstaining from smoking altogether may maximize the health benefits for HS. Ultimately, this study’s findings may assist physicians in educating patients about preventing skin diseases, particularly focusing on the benefits of never smoking and smoking cessation.
Despite the ongoing uncertainty regarding the exact pathophysiological role of smoking in HS, previous research has outlined the potential mechanisms that connect tobacco smoking with HS. Tobacco smoke, containing numerous chemicals including nicotine, polyaromatic hydrocarbons, and dioxinlike compounds, activates various cell types through nicotinic acetylcholine receptors and aryl hydrocarbon receptors.22,23,24 Mainly, nicotine may trigger the production of proinflammatory cytokines, such as tumor necrosis factor α and interleukin (IL)-1α, IL-1β, and IL-8, resulting in the migration of neutrophils, modulation of macrophage function, and the activation of Th17 cells.25,26,27,28 Additionally, dioxins and dioxinlike chemicals in cigarette smoke may activate the aryl hydrocarbon receptor, promoting a proinflammatory state, potentially leading to Th17 cell expansion, thus linking tobacco smoking to HS pathophysiology.24,29 In addition, nicotine, through nicotinic acetylcholine receptors found in various cell types implicated in HS,30 induces infundibular epithelial hyperplasia, upregulates acetylcholine receptors within the follicular infundibulum, and may alter the skin’s microbiome.23,31 Collectively, the evidence strengthens the hypothesis that smoking plays a meaningful role in the development of HS, underscoring the importance of quitting smoking.
According to a previous study,32 the difference in the onset of smoking and HS was relatively short at 4.5 years, which supports the findings in this study. In the case of psoriasis, a skin disorder well known for its statistically significant association with smoking, individuals who smoked for fewer than 10 years showed a psoriasis risk comparable to those who did not smoke, whereas those with a smoking history of 30 years or more had an almost 2-fold increase in psoriasis risk compared to those who did not smoke,33 indicating a cumulative effect of smoking on psoriasis pathogenesis, unlike HS. These findings suggest that the underlying mechanisms through which smoking influences the development of these diseases may differ. To elucidate the biological mechanisms underlying the relationship between smoking and HS, integrated basic and translational research should be conducted to determine the extent to which smoking exposure or cessation over time contributes to the cellular aspects of HS pathogenesis.
Strengths and Limitations
The limitations of the study include unexplored potential confounding factors, such as hereditary background, due to the retrospective nature of the study. Future research should investigate the relationship between smoking cessation and HS, incorporating hereditary background data. Additionally, the reliance on a questionnaire to determine participants’ smoking status may not accurately reflect the exact smoking habits of each participant. Due to data limitations, this study did not consider electronic cigarettes or nicotine replacement therapy when defining those who smoked and those who quit, highlighting the importance for future research to consider these factors. Additionally, this study features a dominant proportion of male smokers compared to female smokers, which is characteristic of the Korean population wherein the smoking prevalence among Korean women is among the lowest globally (6.0% in 1995 to 7.7% in 2018) and may be underreported due to societal stigma.34 Future research targeting more female smokers will be necessary. In contrast to prior US studies,35 a lack of difference in HS risk by income level was found in Korean individuals, potentially limiting generalizability. Finally, given the prolonged diagnostic delay associated with HS, the actual onset of the disease may not align with diagnosis timing. While this study focused on the general population, future research on the effect of smoking or quitting smoking on individuals diagnosed with HS will be crucial. Such studies may bolster recommendations for smoking cessation and help predict how behavioral changes affect HS severity. Additionally, while this study did not examine weight changes in those who quit smoking, it would also be worthwhile to investigate this in future research.
Despite these limitations, this study has several strengths. This is, to our knowledge, the first empirical evidence revealing a markedly and rapidly reduced HS risk through smoking cessation while also demonstrating that resuming or starting new smoking presents risk of HS development similar to smoking continuation. Furthermore, we performed comprehensive analyses to account for potential confounding factors that could affect the development of HS, including adjustments for multiple covariates and the use of stratification. Particularly, due to rigorous control over the time-varying characteristics of smoking habits, these results are highly reliable, as evidenced by the conducted sensitivity analyses using the secondary cohort.
Conclusions
In this cohort study, smoking cessation was associated with a lower risk of developing HS, and this risk-reduction effect was comparable to that observed in individuals who maintained smoking cessation for an extended period. In contrast, those who formerly smoked and then restarted smoking or individuals who had never smoked and then initiated smoking did not exhibit a notable difference in their risk of developing HS compared to individuals who continued to smoke. Therefore, smoking cessation and maintaining a smoke-free lifestyle may be important preventive measures against the development of HS.
eTable 1. Definition of the smoking status group
eTable 2. General characteristics of the study population in the secondary cohort
eTable 3. Income level of individuals with and without HS
eTable 4. Hazard ratios for hidradenitis suppurativa according to income level
eFigure 1. Association between changes in smoking habits and hidradenitis suppurativa according to primary cohort subgroups
eFigure 2. Association between changes in smoking habits and hidradenitis suppurativa according to secondary cohort subgroups
Data Sharing Statement
References
- 1.Nguyen TV, Damiani G, Orenstein LAV, Hamzavi I, Jemec GB. Hidradenitis suppurativa: an update on epidemiology, phenotypes, diagnosis, pathogenesis, comorbidities and quality of life. J Eur Acad Dermatol Venereol. 2021;35(1):50-61. doi: 10.1111/jdv.16677 [DOI] [PubMed] [Google Scholar]
- 2.Prens E, Deckers I. Pathophysiology of hidradenitis suppurativa: an update. J Am Acad Dermatol. 2015;73(5)(suppl 1):S8-S11. doi: 10.1016/j.jaad.2015.07.045 [DOI] [PubMed] [Google Scholar]
- 3.Goldburg SR, Strober BE, Payette MJ. Hidradenitis suppurativa: epidemiology, clinical presentation, and pathogenesis. J Am Acad Dermatol. 2020;82(5):1045-1058. doi: 10.1016/j.jaad.2019.08.090 [DOI] [PubMed] [Google Scholar]
- 4.Seyed Jafari SM, Hunger RE, Schlapbach C. Hidradenitis suppurativa: current understanding of pathogenic mechanisms and suggestion for treatment algorithm. Front Med (Lausanne). 2020;7:68. doi: 10.3389/fmed.2020.00068 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Acharya P, Mathur M. Hidradenitis suppurativa and smoking: a systematic review and meta-analysis. J Am Acad Dermatol. 2020;82(4):1006-1011. doi: 10.1016/j.jaad.2019.10.044 [DOI] [PubMed] [Google Scholar]
- 6.Bukvić Mokos Z, Miše J, Balić A, Marinović B. Understanding the relationship between smoking and hidradenitis suppurativa. Acta Dermatovenerol Croat. 2020;28(1):9-13. [PubMed] [Google Scholar]
- 7.Lipa K, Zając N, Owczarek W, Ciechanowicz P, Szymaǹska E, Walecka I. Does smoking affect your skin? Postepy Dermatol Alergol. 2021;38(3):371-376. doi: 10.5114/ada.2021.103000 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Denny G, Anadkat MJ. The effect of smoking and age on the response to first-line therapy of hidradenitis suppurativa: an institutional retrospective cohort study. J Am Acad Dermatol. 2017;76(1):54-59. doi: 10.1016/j.jaad.2016.07.041 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Dessinioti C, Zisimou C, Tzanetakou V, Ntritsos G, Kontochristopoulos G, Antoniou C. A retrospective institutional study of the association of smoking with the severity of hidradenitis suppurativa. J Dermatol Sci. 2017;87(2):206-207. doi: 10.1016/j.jdermsci.2017.04.006 [DOI] [PubMed] [Google Scholar]
- 10.Chandran NS, Lee JH, Kurokawa I. Hidradenitis suppurativa in South-East Asia and East Asia. Exp Dermatol. 2021;30(suppl 1):23-26. doi: 10.1111/exd.14340 [DOI] [PubMed] [Google Scholar]
- 11.Chu CB, Yang CC, Tsai SJ. Global data analysis supports smoking as the fundamental element associated with geographical sex disparities in hidradenitis suppurativa. Br J Dermatol. 2021;185(5):1054-1056. doi: 10.1111/bjd.20581 [DOI] [PubMed] [Google Scholar]
- 12.Cheol Seong S, Kim YY, Khang YH, et al. Data resource profile: the National Health Information Database of the National Health Insurance Service in South Korea. Int J Epidemiol. 2017;46(3):799-800. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Seong SC, Kim YY, Park SK, et al. Cohort profile: the National Health Insurance Service-National Health Screening Cohort (NHIS-HEALS) in Korea. BMJ Open. 2017;7(9):e016640. doi: 10.1136/bmjopen-2017-016640 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Cosmatos I, Matcho A, Weinstein R, Montgomery MO, Stang P. Analysis of patient claims data to determine the prevalence of hidradenitis suppurativa in the United States. J Am Acad Dermatol. 2013;68(3):412-419. doi: 10.1016/j.jaad.2012.07.027 [DOI] [PubMed] [Google Scholar]
- 15.Vazquez BG, Alikhan A, Weaver AL, Wetter DA, Davis MD. Incidence of hidradenitis suppurativa and associated factors: a population-based study of Olmsted County, Minnesota. J Invest Dermatol. 2013;133(1):97-103. doi: 10.1038/jid.2012.255 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ingvarsson G. Regional variation of hidradenitis suppurativa in the Norwegian Patient Registry during a 5-year period may describe professional awareness of the disease, not changes in prevalence. Br J Dermatol. 2017;176(1):274-275. doi: 10.1111/bjd.14990 [DOI] [PubMed] [Google Scholar]
- 17.Alikhan A, Lynch PJ, Eisen DB. Hidradenitis suppurativa: a comprehensive review. J Am Acad Dermatol. 2009;60(4):539-561. doi: 10.1016/j.jaad.2008.11.911 [DOI] [PubMed] [Google Scholar]
- 18.Kurokawa I, Hayashi N; Japan Acne Research Society . Questionnaire surveillance of hidradenitis suppurativa in Japan. J Dermatol. 2015;42(7):747-749. doi: 10.1111/1346-8138.12881 [DOI] [PubMed] [Google Scholar]
- 19.You HR, Yun SJ, Lee SC, Won YH, Lee JB. Clinical characteristics and epidemiology of hidradenitis suppurativa in Korea: a single-center study. Korean J Dermatol. 2016;54(9):723-727. [Google Scholar]
- 20.Yang JH, Moon J, Kye YC, et al. ; Korean Society for Acne Research . Demographic and clinical features of hidradenitis suppurativa in Korea. J Dermatol. 2018;45(12):1389-1395. doi: 10.1111/1346-8138.14656 [DOI] [PubMed] [Google Scholar]
- 21.Choi E, Cook AR, Chandran NS. Hidradenitis suppurativa: an Asian perspective from a Singaporean institute. Skin Appendage Disord. 2018;4(4):281-285. doi: 10.1159/000481836 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Smith CJ, Hansch C. The relative toxicity of compounds in mainstream cigarette smoke condensate. Food Chem Toxicol. 2000;38(7):637-646. doi: 10.1016/S0278-6915(00)00051-X [DOI] [PubMed] [Google Scholar]
- 23.Hana A, Booken D, Henrich C, et al. Functional significance of non-neuronal acetylcholine in skin epithelia. Life Sci. 2007;80(24-25):2214-2220. doi: 10.1016/j.lfs.2007.02.007 [DOI] [PubMed] [Google Scholar]
- 24.Esser C, Rannug A, Stockinger B. The aryl hydrocarbon receptor in immunity. Trends Immunol. 2009;30(9):447-454. doi: 10.1016/j.it.2009.06.005 [DOI] [PubMed] [Google Scholar]
- 25.Jeong SH, Park JH, Kim JN, et al. Up-regulation of TNF-alpha secretion by cigarette smoke is mediated by Egr-1 in HaCaT human keratinocytes. Exp Dermatol. 2010;19(8):e206-e212. doi: 10.1111/j.1600-0625.2009.01050.x [DOI] [PubMed] [Google Scholar]
- 26.Torii K, Saito C, Furuhashi T, et al. Tobacco smoke is related to Th17 generation with clinical implications for psoriasis patients. Exp Dermatol. 2011;20(4):371-373. doi: 10.1111/j.1600-0625.2010.01224.x [DOI] [PubMed] [Google Scholar]
- 27.van der Zee HH, Laman JD, Boer J, Prens EP. Hidradenitis suppurativa: viewpoint on clinical phenotyping, pathogenesis and novel treatments. Exp Dermatol. 2012;21(10):735-739. doi: 10.1111/j.1600-0625.2012.01552.x [DOI] [PubMed] [Google Scholar]
- 28.Nikota JK, Shen P, Morissette MC, et al. Cigarette smoke primes the pulmonary environment to IL-1α/CXCR-2-dependent nontypeable Haemophilus influenzae-exacerbated neutrophilia in mice. J Immunol. 2014;193(6):3134-3145. doi: 10.4049/jimmunol.1302412 [DOI] [PubMed] [Google Scholar]
- 29.Tsuji G, Takahara M, Uchi H, et al. Identification of ketoconazole as an AhR-Nrf2 activator in cultured human keratinocytes: the basis of its anti-inflammatory effect. J Invest Dermatol. 2012;132(1):59-68. doi: 10.1038/jid.2011.194 [DOI] [PubMed] [Google Scholar]
- 30.Kurzen H, Wessler I, Kirkpatrick CJ, Kawashima K, Grando SA. The non-neuronal cholinergic system of human skin. Horm Metab Res. 2007;39(2):125-135. doi: 10.1055/s-2007-961816 [DOI] [PubMed] [Google Scholar]
- 31.Pavia CS, Pierre A, Nowakowski J. Antimicrobial activity of nicotine against a spectrum of bacterial and fungal pathogens. J Med Microbiol. 2000;49(7):675-676. doi: 10.1099/0022-1317-49-7-675 [DOI] [PubMed] [Google Scholar]
- 32.O’Keeffe C, Kirby B. Cigarette smoking precedes the onset of hidradenitis suppurativa. Int J Dermatol. 2022;61(9):e320-e321. doi: 10.1111/ijd.16117 [DOI] [PubMed] [Google Scholar]
- 33.Naldi L. Psoriasis and smoking: links and risks. Psoriasis (Auckl). 2016;6:65-71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Kim S, Byun G, Jo G, et al. Gender and tobacco epidemic in South Korea: implications from age-period-cohort analysis and the DPSEEA framework. BMJ Open. 2022;12(4):e058903. doi: 10.1136/bmjopen-2021-058903 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Wertenteil S, Strunk A, Garg A. Association of low socioeconomic status with hidradenitis suppurativa in the United States. JAMA Dermatol. 2018;154(9):1086-1088. doi: 10.1001/jamadermatol.2018.2117 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Definition of the smoking status group
eTable 2. General characteristics of the study population in the secondary cohort
eTable 3. Income level of individuals with and without HS
eTable 4. Hazard ratios for hidradenitis suppurativa according to income level
eFigure 1. Association between changes in smoking habits and hidradenitis suppurativa according to primary cohort subgroups
eFigure 2. Association between changes in smoking habits and hidradenitis suppurativa according to secondary cohort subgroups
Data Sharing Statement