A Population-Based Survey of Eczema Prevalence in the United States

Jon M. Hanifin; Michael L. Reed; Eczema Prevalence and Impact Working Group


Dermatitis. 2007;18(2):82-91. 

In This Article



This analysis used a population-based stratified random sample of 60,000 US households identified by the marketing research firm TNS (formerly known as National Family Opinion, or NFO). The TNS household panel is constructed to match the US census for marketing, opinion, and other types of survey research. Potential TNS panel households are initially selected as a stratified probability sample to be representative of the US population with regard to census region, size of the population center (urban vs rural residence), age of the head of the household, household income, and household size. Households are recruited by volunteer response to an initial mailing. A follow-up mailing is conducted to obtain a detailed household census and demographic information. Recruited households are randomly assigned to a sampling of 5,000 households each. Each sampling block is constructed to be a stand-alone representative sample of the US population. The random assignment ensures good geographic and demographic distribution for each block. Every 2 years, updated household census and demographic information is obtained and 20 to 30% of each sampling block is replaced. Households that are persistent nonresponders to periodic surveys are removed from the sampling frame. The voluntary mechanism of the TNS sampling frame is representative of the US population, except that extreme high- and low-income households are underrepresented.[19,20]


In October 1998, a self-administered 10-item screening questionnaire was mailed to a stratified random sample of the TNS panel households. A designated member from each household, most often the female household head, provided age and gender information (Items 1 and 2) for herself (or himself) and up to seven other family members and noted the presence or absence of "itchy, red, dry, or irritated skin, or eczema" in the past 12 months. More detailed symptom and severity information was then provided for up to four family members. If more than four family members had experienced symptoms, those with the most serious skin condition were selected for reporting. For each household member with symptoms, the questionnaire obtained information regarding the following issues: (Item 3) skin conditions or symptoms experienced in the last 12 months, including itching/scratching, red or inflamed rash, excessive dryness/scaling, blisters or bumps, thickening of the skin, and open sores or oozing; (Item 4) the location of the condition, including head, face or scalp, arms or hands, legs or feet, skin folds of arms or legs, chest, abdomen or back, groin or underarms, and elbows or knees; (Item 5) the age of the individual at the onset of skin symptoms; (Item 6) the number of times per year that symptoms flare up or occur (1-2 times, 3-5 times, 6-10 times, more than 10 times, or every day); (Item 7) when the symptoms occur and how many days they last; (Item 8) the severity of skin symptoms (mild, moderate, or severe); (Item 9) how often symptoms disturb sleep (never, rarely, sometimes, or frequently); and (Item 10) if a physician had ever diagnosed seborrheic dermatitis, atopic dermatitis, contact dermatitis, eczema, ichthyosis, psoriasis, hives, herpes, acne, asthma, hay fever, or allergic rhinitis. Patients were considered eczematous if they reported any one of the following physician diagnoses: seborrheic dermatitis, atopic dermatitis, contact dermatitis, or eczema.


Sociodemographic characteristics of the sample were defined; these included geographic region, size of the population center, household income, household size, age of the household head, and race and Hispanic origin. Nine geographic regions were defined: New England, mid-Atlantic, east north central, west north central, south Atlantic, east south central, west south central, mountain, and Pacific. Four sizes of population center were defined: less than 100,000; 100,000 to 499,999; 500,000 to 1,999,999; and ≥ 2,000,000. The five household income groups were as follows: less than $15,000; $15,000 to $29,999; $30,000 to $49,999; $50,000 to $74,999; and ≥ $75,000 (US).

A three-step weighting scheme was applied to the data to ensure that the study sample more closely matched US census demographics and to refine the prevalence estimates. Step 1 adjusted the ethnicity of the sample to compensate for the underrepresentation of nonwhites in the TNS sample (a result of the absence of ethnicity as a control). Step 2 adjusted the sample on the basis of response bias, compensating specifically for the lower rate of questionnaire return from younger households (household head less than 40 years of age) and larger households. Step 3 adjusted the sample, based on any apparent discrepancies in reported data, as a comparison of the reporting levels in married households suggested that the TNS member respondent (usually the female head of household) underrepresented the health status (diagnosis) of the spouse. Taking the self-reported levels of members as accurate, researchers adjusted the spousal levels of diagnosis upward to correct for the underreporting.

Three subpopulations of individuals were defined according to reported symptoms and other characteristics: (1) symptomatic (defined by one or more of the following four symptoms or lesions: itching/scratching, red or inflamed rash, excessive dryness/scaling, or thickening of skin); (2) empirical eczema (defined by itching/scratching and one or both of the following: [a] red or inflamed rash and [b] excessive dryness/scaling); and (3) empirical AD (defined by itching/scratching and at least three of the following: red or inflamed rash, excessive dryness/scaling, condition located in skinfolds of arms or legs, an age of onset between 0 and 5 years, symptoms lasting ≥ 14 days, and a physician diagnosis of asthma or allergic rhinitis/hay fever). The empirical AD group was a subset of the empirical eczema group. The criteria for empirical diagnoses of eczema and AD were based on diagnostic criteria[2,21] by consensus of the Eczema Prevalence & Impact Working Group.

For the total weighted population sample, the percentage of symptomatic individuals during the previous 12 months was calculated. Also calculated was the percentage of physician diagnoses reported by the patients (Item 10). The subpopulation of those reporting a physician diagnosis of eczema was characterized according to the size of the population center, income, and race and was compared in terms of symptoms and comorbidities to the subpopulation of those not reporting a diagnosis of eczema.

The prevalence of empirically defined eczema and empirically defined AD was calculated as a percentage of the total weighted population and as a percentage of the symptomatic subpopulation. The prevalence as a percentage of the total weighted population was further defined in terms of population center size, income, and race. The subpopulation with empirical eczema and the subpopulation with empirical AD were compared in terms of mean values for (1) number, type, frequency, severity of symptoms, and age of the subject at the onset of symptoms; (2) location and number of locations of symptoms; (3) duration of symptom episodes; (4) total symptom days (calculated as frequency multiplied by duration); (5) frequency of sleep disturbance; and (6) self-reporting of physician-diagnosed asthma, hay fever/allergic rhinitis, and skin-related comorbidities (psoriasis, acne, hives, herpes, and ichthyosis).


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.