Multiple Skin Cancer Risk Behaviors in the U.S. Population

Procedure 

The data for this study were drawn from the 2005 NHIS, which is an annual representative U.S. health survey conducted using a multistage, clustered, cross-sectional design, with state-level stratification and oversampling of black and Hispanic populations. The response rate for the data used in this study was 69.0%.16 Additional details regarding the 2005 NHIS are available elsewhere.16 The data were collected in 2005 and analyzed in 2007.

Participants 

Participants were drawn from the 31,428 individuals selected as sample adults for the 2005 NHIS. Individuals were excluded from the current study if they reported a personal history of CMM or NMSC, were unaware of the type of skin cancer they had, or reported not knowing if they had ever had skin cancer or were missing data on the skin cancer history variables (n=767). An additional 2426 individuals were excluded from the analyses due to missing data on one or more of the variables that made up the multiple skin cancer risk behavior measure that was used as the primary outcome for this study (for details, see below). The resulting sample consisted of 28,235 individuals.

Measures 

The exact wording of all measures listed below is available elsewhere.17

Data Weighting and Statistical Analyses 

All statistical analyses were conducted using SUDAAN, version 9.0.1, and were weighted based on design, ratio, and nonresponse adjustments, with poststratification adjustments for 2000 U.S. Census-based estimates of age, gender, and race/ethnicity. All percentages reported in the Results section are weighted and all sample sizes are unweighted.

Given the multiple associations examined and the large sample size, a cutoff of p<0.001 was used to determine statistical significance for all analyses. A series of chi-square tests examined whether individuals who were excluded from analyses due to missing data for the multiple skin cancer risk behavior variable differed on demographic factors compared to individuals who were not missing data for that variable. Next, the demographic characteristics for the full sample were examined, followed by the age-stratified prevalence of each of the five skin cancer risk behaviors, and the percentage of individuals within each age group who reported having each number (from 0 to 5) of the skin cancer risk behaviors. A series of age-stratified ordinal logistic regression analyses using SUDAAN’s PROC MULTILOG procedure were conducted to examine the association between each potential correlate and the multiple skin cancer risk behavior score. In each analysis, the covariate was included as a single independent variable with the skin cancer risk behavior score as the ordinal dependent variable.

Missing Data Analyses and Sample Demographic Characteristics 

Individuals missing data for the multiple skin cancer risk behavior variable were less likely to be non-Hispanic white individuals (65.8%) than those who were not missing data for that variable (71.1%) (χ2=19.39; p=0.0003). There were no differences in missing data for the multiple skin cancer risk behavior variable according to gender, age, education, or marital status (χ2≤13.54; p≥0.004). The demographic characteristics of the sample are shown in Table 1.

Table 1. Demographic characteristics of sample (N=28,235), 2005 National Health Interview Survey

		Sample (%)
	Gender
	Male	48.0
	Female	52.0
	Missing (n)	0
	Age (years)
	18–29	22.3
	30–39	18.9
	40–49	20.7
	50–64	22.8
	≥ 65	15.3
	Missing (n)	0
	Race/ethnicity
	Non-Hispanic white	71.1
	Non-Hispanic black	11.4
	Non-Hispanic other	4.6
	Hispanic	12.9
	Missing (n)	0
	Education level
	College graduate	25.7
	Some college	28.5
	High school graduate	29.4
	Some high school or less	16.4
	Missing (n)	219
	Marital status
	Married/partnered	63.2
	Not married/partnered	36.8
	Missing (n)	88

Note: All percentages are weighted. Data source: National Center for Health Statistics, 2005.23

Age-Stratified Prevalence of Single and Multiple Skin Cancer Risk Behaviors 

The age-stratified prevalence of each of the four skin cancer risk behaviors is shown in Table 2. Infrequent use of sun-protective clothing was high across all age groups. Around half of the participants in each group did not regularly use a sunscreen with an SPF of ≥15. Fewer individuals reported having had a sunburn in the past year, and the sunburn rate was particularly low (11.2%) among older adults (aged ≥65 years). Across all age groups, there was a low prevalence of having used an indoor tanning device in the past year, although the rate was highest (20.2%) among those aged 18–29 years.

Table 2. Age-stratified prevalence of skin cancer risk behaviors, 2005 National Health Interview Survey

		Sample % ± 95% CI
		Age 18–29 (n=5370)	Age 30–39 (n=5473)	Age 40–49 (n=5589)	Age 50–64 (n=6583)	Age 65+ (n=5220)
	Infrequent use of sun-protective clothinga	84.7±1.2	81.5±1.2	76.7±1.3	67.8±1.4	49.2±1.7
	Stay in the sun when outside on a sunny day	35.4±1.8	31.8±1.6	30.2±1.4	24.4±1.4	19.8±1.3
	Infrequent use of SPF 15+ sunscreen	56.7±1.7	49.6±1.7	48.0±1.5	51.3±1.4	55.3±1.7
	Indoor tanning device use in the past year	20.2±1.3	16.7±1.2	13.6±1.1	9.9±0.9	7.6±0.9
	Sunburn in the past year	45.6±1.7	43.6±1.7	40.0±1.6	26.6±1.3	11.2±1.0

Note: All percentages are weighted. Data source: National Center for Health Statistics, 2005.23 SPF, sun protection factor.

a Composite measure based on a low reported frequency of doing the following when out in the sun: wearing a wide-brimmed hat, wearing a long-sleeved shirt, and wearing long pants (or other clothing reaching the ankles).

As shown in Table 3, across each age group, around one third of individuals had two of the five skin cancer risk behaviors. Additionally, in all except the oldest age group (those aged ≥65 years), the proportion of participants having three or more of the five skin cancer risk behaviors varied from just over a quarter (among those aged 50–64 years) to just under a half (among 18–29 year olds). Among those aged ≥65 years, around one quarter (24.4%) had none of the skin cancer risk behaviors and a further one quarter (28.0%) had one risk behavior. The mean number of skin cancer risk behaviors was highest among individuals aged 18–29 years (M=2.43) and was lowest among those aged ≥65 years (M=1.43).

Table 3. Age-stratified prevalence of multiple skin cancer risk behaviors, 2005 National Health Interview Survey

	Number of skin cancer risk behaviorsa	Sample % ± 95% CI
		Age 18–29 M=2.43 (n=5370)	Age 30–39 M=2.23 (n=5473)	Age 40–49 M=2.09 (n=5589)	Age 50–64 M=1.80 (n=6583)	Age 65+ M=1.43 (n=5220)
	0	4.2±0.7	6.1±0.8	8.6±0.8	12.7±1.0	24.4±1.4
	1	14.3±1.3	17.5±1.2	21.1±1.2	26.9±1.3	28.0±1.4
	2	35.2±1.7	37.4±1.4	35.0±1.4	35.2±1.3	30.4±1.5
	3	30.0±1.6	27.2±1.3	25.0±1.3	19.1±1.1	14.9±1.2
	4	13.5±1.2	9.7±0.9	8.9±0.9	5.4±0.6	2.2±0.5
	5	2.8±0.6	2.1±0.4	1.4±0.4	0.7±0.3	0.2±0.1

Note: All percentages are weighted. Data source: National Center for Health Statistics, 2005.23

a Skin cancer risk behaviors: infrequent use of sun-protective clothing; stay in the sun when outside on a sunny day; infrequent use of SPF 15+ sunscreen; indoor tanning device use in the past year; sunburn in the past year.

Age-Stratified Correlates of Multiple Skin Cancer Risk Behavior Prevalence 

The results of a series of ordinal logistic regression analyses examining correlates of multiple skin cancer risk behaviors are shown in Table 4. The results are summarized here. With regard to demographic factors, more skin cancer risk behaviors were found among individuals in the Midwest (except among those aged 65 years and over); men (except among individuals aged 18–29 years); non-Hispanic white individuals (aged 18–49 years); and individuals with lower levels of education (except among those aged ≥65 years). There was little evidence that individuals’ number of skin cancer risk behaviors varied according to whether they visited a physician in the last year and the type of healthcare coverage they had. Among several age groups, more skin cancer risk behaviors were found among individuals who were overweight or obese, physically active individuals, and risky drinkers. Across all age groups, current smokers had more skin cancer risk behaviors than nonsmokers. Individuals aged 18–39 years who perceived themselves as being at higher risk for cancer had more skin cancer risk behaviors. Individuals who reported having skin that was less sensitive to the sun had more skin cancer risk behaviors. Among individuals aged 40–64 years, those who reported never having had a total skin exam had more skin cancer risk behaviors.

Table 4. Age-stratified ordinal logistic regression analyses examining correlates of multiple skin cancer risk behaviors,a 2005 National Health Interview Survey

		Age 18–29		Age 30–39		Age 40–49		Age 50–64		Age 65+
		OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
	Region
	Northeast	Refb		Refb		Refb		Refb		Refb
	Midwest	1.55	1.25–1.93	1.39	1.13–1.71	1.42	1.23–1.64	1.19	1.02–1.39	1.04	0.85–1.28
	South	0.91	0.74–1.11	0.87	0.72–1.05	0.87	0.76–0.99	0.76	0.65–0.90	0.76	0.63–0.91
	West	0.68	0.54–0.85	0.68	0.56–0.83	0.77	0.66–0.89	0.65	0.54–0.77	0.71	0.58–0.87
	Gender
	Male	Ref		Refb		Refb		Refb		Refb
	Female	0.83	0.73–0.94	0.71	0.64–0.79	0.59	0.53–0.67	0.58	0.53–0.64	0.52	0.47–0.59
	Race/ethnicity
	Non-Hispanic white	Refb		Refb		Refb		Ref		Ref
	Non-Hispanic black	0.39	0.33–0.46	0.56	0.47–0.65	0.54	0.47–0.62	0.86	0.74–1.00	1.13	0.95–1.33
	Non-Hispanic other	0.28	0.22–0.36	0.47	0.36–0.60	0.60	0.44–0.82	0.80	0.62–1.03	0.76	0.59–0.99
	Hispanic	0.46	0.40–0.52	0.54	0.47–0.62	0.64	0.54–0.75	0.78	0.66–0.92	0.89	0.70–1.15
	Education level
	College graduate	Refb		Refb		Refb		Refb		Ref
	Some college	1.46	1.23–1.74	1.37	1.19–1.59	1.23	1.07–1.41	1.31	1.16–1.48	1.04	0.87–1.24
	High school graduate	1.21	1.02–1.43	1.34	1.15–1.56	1.61	1.40–1.86	1.43	1.26–1.63	1.02	0.87–1.21
	Some high school or less	1.26	1.02–1.54	1.10	0.93–1.29	1.20	1.00–1.44	1.28	1.10–1.49	1.09	0.92–1.29
	Marital status
	Married/partnered	Ref		Ref		Ref		Ref		Refb
	Not married/partnered	0.95	0.84–1.08	0.89	0.79–0.99	0.96	0.86–1.08	0.92	0.83–1.02	0.68	0.61–0.75
	Visited a physician in the last year
	Yes	Ref		Ref		Ref		Ref		Ref
	No	0.93	0.81–1.06	1.01	0.90–1.15	1.21	1.05–1.40	1.21	1.05–1.40	1.43	1.13–1.80
	Healthcare coverage
	Private	Ref		Ref		Ref		Refb		Ref
	Public	0.77	0.64–0.93	0.78	0.65–0.94	0.86	0.70–1.04	0.71	0.61–0.83	0.94	0.83–1.06
	None	0.90	0.79–1.04	0.93	0.82–1.06	1.09	0.95–1.25	0.96	0.82–1.11	0.96	0.42–2.18
	Body mass index
	Normal	Ref		Ref		Refb		Ref		Refb
	Overweight	1.12	0.96–1.30	1.22	1.07–1.38	1.30	1.15–1.46	1.19	1.05–1.34	1.24	1.11–1.40
	Obese	0.96	0.81–1.13	1.03	0.90–1.17	1.18	1.03–1.35	1.11	0.97–1.27	1.49	1.27–1.74
	Physical activity
	None	Refb		Ref		Ref		Ref		Refb
	Some	1.15	0.97–1.37	1.19	1.01–1.40	1.13	0.98–1.31	1.04	0.91–1.20	1.14	1.00–1.30
	Meet recommendations	1.38	1.20–1.59	1.25	1.10–1.41	1.18	1.04–1.34	1.12	1.00–1.26	1.40	1.23–1.60
	Smoking status
	Current smoker	Refb		Refb		Refb		Refb		Refb
	Former smoker	0.74	0.58–0.93	0.75	0.63–0.90	0.74	0.63–0.86	0.77	0.66–0.89	0.83	0.69–1.00
	Never smoker	0.51	0.44–0.59	0.54	0.48–0.62	0.55	0.49–0.63	0.63	0.54–0.72	0.57	0.47–0.69
	Alcohol use
	Risky drinker	Refb		Refb		Refb		Ref		Ref
	Not risky drinker	0.44	0.34–0.56	0.55	0.40–0.76	0.60	0.46–0.77	0.77	0.61–0.97	0.63	0.46–0.87
	Perceived cancer risk compared to others
	Less likely	Refb		Refb		Ref		Ref		Ref
	As likely	1.26	1.09–1.46	1.21	1.05–1.38	0.96	0.85–1.09	1.04	0.94–1.16	0.89	0.79–1.01
	More likely	2.06	1.67–2.54	1.61	1.33–1.94	1.09	0.91–1.31	1.07	0.91–1.26	0.77	0.64–0.93
	Don’t know	0.90	0.66–1.23	0.96	0.70–1.31	0.68	0.51–0.92	0.98	0.77–1.26	0.88	0.72–1.08
	Family history of melanoma
	Yes	Ref		Ref		Ref		Ref		Ref
	No	1.14	0.78–1.64	1.04	0.61–1.78	1.08	0.74–1.58	1.50	1.16–1.95	1.20	0.80–1.81
	Skin reaction after 1 hour in the sun
	Moderate/severe sunburn	Refb		Refb		Refb		Refb		Refb
	Mild sunburn	1.38	1.15–1.65	1.30	1.11–1.53	1.31	1.12–1.53	1.35	1.18–1.55	1.31	1.09–1.57
	No sunburn	0.85	0.73–0.98	0.95	0.83–1.08	1.17	1.01–1.34	1.57	1.39–1.77	2.16	1.85–2.52
	Do not go out in the sun	0.05	0.03–0.07	0.04	0.03–0.07	0.06	0.04–0.09	0.10	0.08–0.13	0.15	0.12–0.19
	Skin reaction after 2 weeks in the sun
	Sunburn repeatedly/freckle	Refb		Refb		Refb		Refb		Refb
	Mild tan	1.22	1.01–1.47	1.32	1.12–1.54	1.50	1.26–1.80	1.67	1.44–1.94	2.11	1.77–2.50
	Moderate tan	1.42	1.19–1.70	1.82	1.55–2.13	1.85	1.57–2.19	2.29	1.96–2.68	2.66	2.26–3.14
	Very dark tan	1.75	1.40–2.17	1.60	1.34–1.90	2.08	1.71–2.54	2.90	2.37–3.55	2.72	2.14–3.45
	Do not go out in the sun	0.13	0.09–0.20	0.11	0.08–0.17	0.14	0.10–0.19	0.18	0.14–0.23	0.24	0.20–0.30
	Ever had a total skin exam
	Yes	Ref		Ref		Refb		Refb		Ref
	No	1.05	0.85–1.31	1.14	0.99–1.32	1.31	1.12–1.54	1.37	1.19–1.56	1.09	0.95–1.25

Note: Data source: National Center for Health Statistics, 2005.23

a Skin cancer risk behaviors: infrequent use of sun-protective clothing; stay in the sun when outside on a sunny day; infrequent use of SPF 15+ sunscreen; indoor tanning device use in the past year; sunburn in the past year. b Denotes a significant association (p<0.001) between the variable and having multiple skin cancer risk behaviors.

Study Limitations 

There are several limitations to the current study. The study design was cross-sectional, which limits conclusions made regarding age differences (e.g., differences may represent cohort effects) as well as the causal associations between perceived risk and risk behaviors. Skin cancer risk behaviors were assessed by self-report, which may not correspond with actual engagement in these behaviors. Skin cancer risk is cumulative, and some of the risk measures (sunburn, indoor tanning) evaluated only past-year behaviors. Finally, the NHIS database contains very few psychological variables that are known correlates of sun protection behaviors (e.g., appearance motivations), and did not assess skin self-examination.