Epidemiology and Comorbidities of Severe Acne in Adolescence

In This Article

Abstract and Introduction

Abstract

Background. Little is known about the epidemiology of severe acne in the U.S.

Objectives. We sought to study the U.S. prevalence, determinants and comorbidities of severe acne in adolescence.

Methods. We analysed data from the 2007 National Health Interview Survey, a cross-sectional questionnaire-based study of 9417 children ages 0–17 years. Prevalence of severe acne, demographics and comorbid disorders were determined.

Results. The U.S. prevalence of severe acne was virtually nil in the first decade of life, but increased in a linear fashion from 11 years [1·7% (95% confidence interval (CI) 0·4–3·0%)] to 17 years of age [12·1% (95% CI 7·8–16·5%)] (Rao-Scott Chi-square, P < 0·0001). Severe acne was more common in Whites compared with other racial groups at age 14–15 years (P = 0·0004) and girls at age 11–13 (P = 0·02). Severe acne was associated with a number of comorbid disorders. Sinopulmonary disease included sinus infection (P = 0·0003), sore throat other than strep infection (P = 0·0003), asthma (P = 0·03) and nonasthmatic lung disease (P = 0·03). Upper gastrointestinal comorbidities included reflux/heartburn (P = 0·0003), abdominal pain (P = 0·03), nausea/vomiting (P = 0·0001) and food/digestive allergy (P = 0·01). Psychological comorbidities included depression (P = 0·02), anxiety (P < 0·0001), attention deficit disorder/attention deficit hyperactivity disorder (P = 0·01) and insomnia (P = 0·02).

Conclusions. In conclusion, severe acne was more prevalent in older age, Whites, female sex and higher socioeconomic status. Future studies are needed to confirm the associations with sinopulmonary, upper gastrointestinal and psychological disorders in adolescents.

Introduction

Acne vulgaris is a common dermatological disorder. Little is known about the recent disease epidemiology.^[1] Firstly, U.S. population-based estimates are lacking for the prevalence and determinants of severe acne. Secondly, while African American/Black patients are more prone to post-inflammatory hyperpigmentation, few studies have examined racial/ethnic differences with respect to severe acne. We hypothesized that there are racial/ethnic differences with respect to acne severity.

Acne is often regarded as a purely cosmetic disorder. Some studies found increased psychological comorbidity associated with acne.^[2–5] Despite the well-known principle that skin findings are often a clue of systemic disease, few studies have examined the systemic disorders associated with acne. We aim to study the population prevalence, determinant and systemic disease and psychological comorbidities of severe acne in the U.S. population.

We hypothesized that Propionibacterium acnes may play a role in the pathogenesis of other disorders, particularly those involving common areas where P. acnes is found such as the head and neck and gastrointestinal system. A recent study of 13,215 Chinese adolescents found an association between sebaceous gland disease, including acne, and gastric reflux, abdominal bloating and constipation.^[6] A case–control study exploring the potential association between isotretinoin and inflammatory bowel disease (IBD) found a near significant association between acne per se and IBD.^[7,8] Together, these studies suggest that there is an association between acne and gastrointestinal disease. We aimed to study the gastrointestinal and sinopulmonary comorbidities of severe acne in the U.S. population.

Table 1. Statistical interactions between severe acne and chronic medication use and its effect on sinopulmonary and gastrointestinal disease ( n = 3799).
Severe acne	Prescription medication use for 3 months	OR (95% CI)	P-value	aOR (95% CI)	P-value
Sore throat other than strep
No	No	1·00 (reference)	–	1·00 (reference)	–
No	Yes	1·23 (1·02–1·49)	0·03	1·23 (1·02–1·49)	0·03
Yes	No	1·24 (1·03–1·50)	0·02	1·24 (1·03–1·50)	0·03
Yes	Yes	1·53 (1·16–2·03)	0·003	1·52 (1·15–2·02)	0·003
Abdominal pain
No	No	1·00 (reference)	–	1·00 (reference)	–
No	Yes	1·42 (1·05–1·93)	0·02	1·41 (1·04–1·93)	0·03
Yes	No	1·28 (0·94–1·74)	0·11	1·21 (0·90–1·63)	0·21
Yes	Yes	1·82 (1·26–2·62)	0·001	1·71 (1·17–2·50)	0·006
Nausea/vomiting
No	No	1·00 (reference)	–	1·00 (reference)	–
No	Yes	1·35 (1·09–1·65)	0·005	1·34 (1·08–1·65)	0·007
Yes	No	1·34 (1·09–1·65)	0·006	1·32 (1·07–1·62)	0·01
Yes	Yes	1·80 (1·35–2·41)	< 0·0001	1·76 (1·31–2·37)	0·0002

CI, confidence interval; OR, odds ratio; aOR, adjusted OR. Logistic regression models were constructed with 1-year history of various aetiologies of sinopulmonary and gastrointestinal disease as the dependent variables. The independent variables were a history of severe acne in the past year, use of prescription medication in the past 3 months and a linear interaction term between them (binary). Other covariates included in the models were age, race and sex. aOR and 95% CI were calculated for each of the four combinations of severe acne and prescription medication use. Linear interactions were included in the final models (and presented in the table above) if the interaction term was significant (P < 0·05) and/or there was an effect modification on estimates by > 10%. For sinopulmonary disease, there was a significant interaction for sore throat other than strep. For gastrointestinal disease, there was a significant interaction for abdominal pain and nausea/vomiting. Parental refusal to answer a particular question or response of 'don't know' occurred for the questions pertaining to severe acne in three (0·03%), sinus infection in three (0·04%), strep throat in 13 (0·2%), other aetiologies of sore throat in 22 (0·2%), asthma in 23 (0·2%), other pulmonary disease in six (0·1%), hay fever in 23 (0·2%), respiratory allergy in 15 (0·1%), reflux/heartburn in 16 (0·1%), abdominal pain in 11 (0·1%), nausea/vomiting in five (0·1%), diarrhoea/colitis in four (0·04%), constipation in 10 (0·1%), food/digestive allergy in six (0·07%), age in zero (0·0%), race in zero (0·0%), number of children in the household in zero (0·0%), and highest level of education in the household in 65 (0·7%), respectively. Percentages are of the whole cohort (N = 9417), but there are some missing values.