 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
a Department of Dermatology, Boston University School of Medicine, Boston, Massachusetts
b Department of Epidemiology, Boston University School of Public Health, Boston, Massachusetts
c Channing Laboratory, Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts
d Division of Public Health Practice, Harvard School of Public Health, Boston, Massachusetts
e Department of Pediatric Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
METHODS. A retrospective study was conducted of the sun protection behaviors of the adolescent participants in the Growing Up Today Study (GUTS), who were offspring of mothers from the Nurses Health Study II. Adolescents' surveys were matched with their mothers' reports of a personal or family history of skin cancer and compared with adolescents whose mothers did not report a personal or family history of skin cancer. The outcome measures were (1) occurrence of frequent sunburns during the past summer, (2) use of a tanning bed during the past year, and (3) routine use of sunscreen. Frequent sunburns were defined as the report of 
3 sunburns during the past summer. We compared those who reported having used a tanning bed in the past year at least once with those who reported no tanning bed use in the past year. Routine use of sunscreen was defined as a respondent who replied that he or she "always" or "often" used sunscreen with sun protection factor of 15 or more when he or she was outside for >15 minutes on a sunny day during the past summer. General estimating equations were used to calculate odds ratios and 95% confidence intervals adjusted for gender, age, color of untanned skin, and number of friends who were tanned. We also conducted an additional analysis restricted to children whose mothers had received a diagnosis of skin cancer in which we assessed sun protection behaviors according to the child's age and mother's age at the time of the mother's diagnosis and the number of years that had passed since the diagnosis of the mother's skin cancer.
RESULTS. In 1999, 9943 children reported their sun protection behaviors; 8697 of their mothers had not received a diagnosis of skin cancer or reported a family history of melanoma, 463 participants' mothers had received a diagnosis of skin cancer, and 783 participants' mothers reported a family history of melanoma. Between 1989 and 1999, 371 mothers of GUTS participants received a diagnosis of skin cancer: melanoma (n = 44), squamous cell (n = 39), and basal cell cancer(n = 311); 23 mothers received a diagnosis of >1 type of skin cancer. Because GUTS includes siblings from the same family, the 371 mothers with skin cancer had 463 offspring in GUTS. Offspring of mothers with skin cancer were slightly more likely to report frequent sunburns in the past year compared with those with neither maternal diagnosis nor family history (39% vs 36%). Tanning bed use was not significantly different among those with either a maternal diagnosis of skin cancer or family history of melanoma as compared with nonaffected adolescents (8% vs 9% vs 10%). Sunscreen use among offspring of mothers with skin cancer was higher than among those whose mothers had a family history of melanoma or mothers with no personal history of skin cancer (42% vs 33% vs 34%). Tan-promoting attitudes were also similar across all groups. Only 25% thought that a natural skin color was most attractive, and on average, 25% in each group agreed that it was worth burning to get a tan. Children of mothers who had received a diagnosis >2 years in the past were less likely to use sunscreen, more likely to sunburn, and more likely to use tanning beds than children of mothers with a more recent diagnosis, although the results did not reach statistical significance.
CONCLUSION. Frequent sunburns, suboptimal sunscreen use, and high rates of tanning bed use are commonplace even among the children of health professionals who are at risk for developing skin cancer themselves as a result of personal or family history. With new information on family risk, pediatricians can use the potential of a teachable moment to ensure optimal sun protection for children who are at risk.
Key Words: melanoma • skin cancer • family history • children • parents • sun protection • sunscreen
Abbreviations: SCC—squamous cell carcinoma • BCC—basal cell carcinoma • GUTS—Growing Up Today Study • NHSII—Nurses' Health Study II • OR—odds ratio • CI—confidence interval
Risk for skin cancer is increased among the first-degree relatives of skin cancer patients.1,2 Family history of melanoma increases risk for melanoma 2- to 8-fold.1 Family history of squamous cell (SCC) and basal cell cancer (BCC) increases risk at least 2-fold, depending on the histology, number of lesions, and degree of invasiveness.3–5 Therefore, family members of people who receive a diagnosis of skin cancer represent a group that should take additional precautions to limit sun exposure.6,7 However, all previous interventional trials have focused only on the adult relatives of skin cancer patients. No previous studies, to our knowledge, have examined the effect that a skin cancer diagnosis, in either a parent or a first-degree relative, has on the sun protection behavior of the patient's progeny. Changing the behavior of the children of skin cancer patients is critical, because there is ample epidemiologic evidence that sun exposure during childhood and adolescence most affects future risk for skin cancer.3,6
Using data from a national cohort study of adolescents who participate in the Growing Up Today Study (GUTS), we previously reported that sun protection behaviors are less than optimal8 but approximate the rates observed in national surveys of children of similar age. Because the mothers of the adolescents in GUTS are participants in the Nurses Health Study II (NHSII), we can explore intergenerational influences on these behaviors. Using the mother's report of her own diagnosis of skin cancer and of family history of melanoma in either a parent or a sibling, we compared the use of sunscreen, frequency of multiple sunburns, and use of tanning beds among children with a family history of skin cancer versus those without such a history. We hypothesized that children with family history of skin cancer, either in the mother herself or 1 of the mother's first-degree relatives, would have more consistent sunscreen use, less frequent sunburns, and lower utilization of tanning beds than children of unaffected mothers. We also hypothesized that children whose mothers received a diagnosis when the children were at a more impressionable age (ie, before adolescence) and children whose mothers had received a diagnosis more recently, so that the information was more salient, would be more likely to report sun protective behaviors.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
In 1999, a section on sun protection practices and attitudes was added to the GUTS questionnaire. Overall, 12414 participants responded to the 1999 GUTS survey; however, 1447 completed the short version of the survey that is sent to those who do not respond to the long version. The short version did not include the questions on sun protection; hence, these participants were not included in this analysis. Only 9 nonwhite mothers (black, Hispanic, or Asian) received a diagnosis of skin cancer between 1989 and 1999. Therefore, only white participants were included in this analysis, and 876 nonwhite participants were excluded. An additional 148 participants were excluded from the analysis because they had failed to respond to key confounding variables, resulting in a final data set of 9943 white children.
Diagnosis of Skin Cancer
NHSII, a cohort study of >116000 female registered nurses, was established in 1989; details of the study are reported elsewhere.9 As part of a biennial questionnaire, NHSII participants report major medical diagnoses, including skin cancer (BCC, SCC, or melanoma). The diagnosis of melanoma and SCC is confirmed by review of medical records by a physician. In addition to report of their own diagnoses, the participants of NHSII report family history of specific cancers, including melanoma in either a parent or a sibling.
Outcome Measures
The outcome measures were (1) routine use of sunscreen, (2) occurrence of frequent sunburns during the past summer, and (3) use of a tanning bed during the past year. Routine use of sunscreen was defined as a respondent who replied that he or she "always" or "often" used sunscreen with sun protection factor 15 or more when he or she was outside for >15 minutes on a sunny day during the past summer. Frequent sunburns were defined as the report of 3 sunburns during the past summer. We compared those who reported having used a tanning bed in the past year at least once with those who reported no tanning bed use in the past year.
Covariates and Effect Modifiers
Respondents were asked to characterize the color of their untanned skin. Response categories were dichotomized as very fair/fair versus olive/dark. Fitzpatrick skin type was not assessed because there had been no previous studies using this self-reported measure among children as young as 11 years. As a measure of peer influence, participants reported the number of friends who had a tan at the end of the past summer. Responses were dichotomized to all/most/some versus none/few friends. Attitudes toward tanning were ascertained by 2 questions. Participants stated the extent to which they agreed or disagreed with the statement, "It is worth getting a little burned to get a good tan," which was dichotomized as strongly agree/agree versus strongly disagree/disagree/neither agree nor disagree. Participants were also asked, "What kind of tan do you find most attractive?" Responses were dichotomized into very dark brown/moderate brown/light brown versus little color/natural color.
Statistical Analyses
Our primary analyses compared routine sunscreen use, frequent sunburn occurrence, and ever use of tanning bed among children whose mothers reported that they had received a diagnosis of skin cancer or had a family history of melanoma versus children whose mothers reported no personal or family history. We also assessed whether the diagnosis of skin cancer in the mother affected attitudes toward tanning. We repeated these analyses restricting the group of offspring of mothers with cancer to those whose mothers had reported a diagnosis of the more consequential skin cancers (melanoma or SCC), ie, excluding BCC. We also conducted an additional analysis restricted to children whose mothers had received a diagnosis of skin cancer in which we assessed sun protection behaviors according to the child's age and mother's age at the time of the mother's diagnosis and the number of years that had passed since the diagnosis of the mother.
2 tests for categorical data were used to calculate P values for the null hypothesis of no association. General estimating equations, which account for sibling clustering among the cohort, were used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) adjusted for gender, age, color of untanned skin, and number of friends who were tanned. Statistical analyses were performed using SAS Version 8.2 (SAS Institute, Cary, NC).10
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Between 1989 and 1999, 371 mothers of GUTS participants received a diagnosis of skin cancer: melanoma (n = 44), SCC (n = 39), and BCC (n = 311); 23 mothers received a diagnosis of >1 type of skin cancer. Because GUTS includes siblings from the same family, the 371 mothers with skin cancer had 463 offspring in GUTS.
The distribution of gender, age, and untanned skin color was similar for the offspring of mothers with skin cancer, mothers with the family history of melanoma, and mothers with neither a personal nor a family history (Table 1). Tan-promoting attitudes were also similar across all groups. Overall, only 25% thought that a natural skin color was most attractive, and on average, 25% in each group agreed that it was worth burning to get a tan. The proportion of offspring's friends who were tanned also did not differ according to whether the mother had received a diagnosis of skin cancer or had a family history of melanoma. Offspring of mothers with family history of melanoma had a very similar profile on all outcome measures as the offspring of mothers with neither a personal nor a family history (Table 2).
|
|
Use of Tanning Beds
Tanning bed use did not vary by maternal diagnosis of skin cancer or by family history of melanoma. No differences in the use of tanning beds was demonstrable in analyses stratified by gender, age, color of untanned skin, and number of friends who were tanned.
Routine Use of Sunscreen
Offspring of mothers with skin cancer reported more routine use of sunscreen than the offspring of mothers without skin cancer (42% vs 34%; OR: 1.50; 95% CI: 1.23–1.82). This effect was limited to girls. Routine sunscreen use decreased with age for all groups, and by age 16 to 18 years, there was no difference in use of sunscreen among the 3 groups. Fair children were more likely to report sunscreen use than children with darker skin; this difference was accentuated further by the presence of the diagnosis of skin cancer in the mother. Forty-five percent of the children whose mothers had received a diagnosis of skin cancer reported routine use of sunscreen compared with only 27% of olive/dark-skinned children whose mothers had not received a diagnosis of skin cancer.
Effect Modification by Type of Skin Cancer Diagnosed in the Mother, Maternal Age at Diagnosis, Child Age at Maternal Diagnosis, and Remoteness of Cancer Diagnosis
Skin cancer diagnoses with more severe health implications (melanoma or SCC) did not change the results (data not shown). Maternal age at diagnosis ranged from 22 to 52 years (median: 40 years); no association was shown between mother's age at diagnosis and any of the outcome measures (data not shown). The age of the child at maternal diagnosis ranged from <1 to 17 years (median: 9 years); child's age at maternal diagnosis was not associated with any of the outcome measures. Children of mothers who had received a diagnosis >2 years in the past were less likely to use sunscreen (OR: 0.73; 95% CI: 0.45–1.21), more likely to sunburn (OR: 1.64; 95% CI: 1.00–2.77), and more likely to use tanning beds (OR: 2.03; 95% CI: 0.66–6.29) than children of mothers with a more recent diagnosis, although the results did not reach statistical significance (Table 3).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Suboptimal prevention practices found in this study are consistent with recent national studies of average-risk children and adolescents.11–14 A national population-based telephone survey of 1192 11- to 18-year-olds found low prevalence of applying sunscreen (31%) and staying in the shade (22%), and approximately 80% of children between 11 and 18 years of age reported at least 1 sunburn in the previous year, with >30% having received 3 or more sunburns.11 Other studies have also demonstrated that sun protection practices diminish with age.12,13 Our estimate of tanning bed use among teenage girls concurs with estimates of 10% in another population-based study.14
Family history is a risk factor for many chronic diseases of public health significance to pediatricians, including asthma,15 diabetes,16 breast cancer,17 and skin cancer.1 There is an emerging interest in the use of family medical history for identifying people who are at increased risk for common chronic diseases. A 2002 meeting entitled Family History for Public Health and Preventive Medicine: Developing a Research Agenda (www.cdc.gov/genomics/) can provide insights to pediatricians on ways to incorporate family history counseling into practice.18 Currently, no federal guidelines recommend pediatrician assessment of parental skin cancer history. Future surveys of pediatricians should determine the use of a question regarding family history of skin cancer on the parent assessment. Likewise, surveys of dermatologists might consider the frequency of family counseling for risk reduction after a skin cancer diagnosis. Subsequently, trials could consider testing various educational strategies to enhance physician counseling for sun protection in the family setting.
A number of studies have documented suboptimal pediatric counseling for sun protection among average-risk patients. Most recently, Gritz et al19 conducted a mailed survey of 202 Texas pediatricians and found that most pediatricians recommended sunscreen use, but only half recommended protective clothing, shade, or limiting midday exposure.
Slightly higher burning rates among the offspring of mothers with skin cancer, despite more routine sunscreen use, could occur for several reasons. First, offspring of mothers with skin cancer may burn more easily than their peers without a family history of skin cancer. Offspring of affected mothers were only slightly more likely than their counterparts to report having fair or very fair skin. In addition, we adjusted for color of the untanned skin, although this may not fully capture propensity to burn/inability to tan. Another reason that children of affected mothers may not alter their sun protection behaviors is that they may be totally unaware of their mother's diagnosis; we do not know whether mothers have discussed the need for better sun protection with their families. Another possibility is that even if mothers who have had skin cancer diagnoses more commonly recommend sunscreen use, the sunscreen is applied inadequately, as has been observed in other adolescents.20 In other words, increased sunscreen use may result in more burns as a result of a false sense of security. Another possible explanation is reporting bias: offspring of affected mothers may more accurately report the incidence of burning, as a result of having been sensitized to the issue.
These findings are subject to several potential limitations. First, the behaviors are reported after the diagnosis of the mother, and inferences about cause and effect need to be verified in longitudinal analyses that can assess whether a change in sun protection behaviors occurs after a diagnosis of skin cancer. This study does not represent a random sample of all US adolescents, although it does include participants from all 50 states. The mothers of the participants also hold nursing degrees, which may affect their children's health behaviors. Any lack of generalizability of results, however, does not negate the internal validity of the study, and the resultant data can be used to generate hypotheses to test in cohorts with different racial and socioeconomic profiles.
The greatest potential source of misclassification of the data is attributable to the self-report of all data in the study. Data from other major longitudinal studies of adolescent health, including the Youth Risk Behavior Surveillance System and the Longitudinal Survey of Adolescent Health, also are self-reported, and these data have reasonable validity and reliability.21,22 Understanding the potential impact of the family history of melanoma is limited by the fact that the date of diagnosis of the family member was not available, so we were not able to ascertain whether the diagnosis had occurred recently or in the distant past, perhaps even before the birth of the child. Sunscreen use was the only type of sun protection examined; therefore, overall rates of sun protection may be higher than reported, although sunscreen is the most common form of sun protection used.12 As noted earlier, there were no questions on the amount of solar exposure or propensity to burn, both of which may have differed among the offspring of affected children. Finally, we do not have any information about skin cancer diagnoses in the father.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
We are grateful to Lisa Li, MD, for statistical programming and analysis.
|
FOOTNOTES |
|---|
Address correspondence to Alan C. Geller, MPH, RN, Boston University School of Medicine, Department of Dermatology, 720 Harrison Ave, DOB801A, Boston, MA 02118. E-mail: ageller{at}bu.edu
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. F. Haas Teens and Tans: Implementing Behavioral Change Arch Dermatol, August 1, 2007; 143(8): 1058 - 1061. [Full Text] [PDF]
|
|
|
|
 |
|
 J. A Nerad All skin cancers are not created equal Br. J. Ophthalmol., March 1, 2007; 91(3): 276 - 277. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2006 American Academy of Pediatrics. All rights reserved. |
||
|
||
