A, A largely linear trend in total body nevus counts is shown among non-Hispanic white participants, with boys accumulating nevi at a faster rate than girls beginning at approximately 6 years of age and continuing through 16 years of age. B, A largely linear trend in total body nevus counts is shown among Hispanic white participants, with boys accumulating nevi at a faster rate than girls beginning at approximately 3 years of age and continuing through 10 years of age. The lighter gray band is the 95% CI around the regression line for boys, the tan band is the 95% CI around the regression line for girls, and the darker gray band is the overlap in the 95% CIs for boys and girls.
A, A leveling off of nevus counts on body sites chronically exposed to the sun by 16 years of age is shown among non-Hispanic white participants, with boys accumulating nevi at a faster rate than girls at all ages. B, A leveling off of nevus counts on body sites chronically exposed to the sun by 16 years of age is shown among Hispanic white participants, with boys accumulating nevi at a faster rate than girls at all ages. The lighter gray band is the 95% CI around the regression line for boys, the tan band is the 95% CI around the regression line for girls, and the darker gray band is the overlap in the 95% CIs for boys and girls.
A, A largely linear trend in nevus counts on body sites intermittently exposed to the sun and an upward curvature at the oldest ages are shown among non-Hispanic white participants, with girls accumulating nevi at a faster rate than boys. B, A largely linear trend in nevus counts on body sites intermittently exposed to the sun and an upward curvature at the oldest ages are shown among Hispanic white participants, with steeper increases for girls compared with boys from 11 to 16 years of age. The lighter gray band is the 95% CI around the regression line for boys, the tan band is the 95% CI around the regression line for girls, and the darker gray band is the overlap in the 95% CIs for boys and girls.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Asdigian NL, Barón AE, Morelli JG, et al. Trajectories of Nevus Development From Age 3 to 16 Years in the Colorado Kids Sun Care Program Cohort. JAMA Dermatol. 2018;154(11):1272–1280. doi:10.1001/jamadermatol.2018.3027
How does nevus development from age 3 to 16 years vary by sex, Hispanic ethnicity, and on body sites that are chronically vs intermittently exposed to the sun?
In this longitudinal cohort study, overall, nevus accumulation increased linearly over time. Boys accumulated nevi at a faster rate than girls in both ethnicities, girls accumulated more nevi on body sites intermittently exposed to the sun, and nevus acquisition leveled off on sites chronically exposed to the sun for both sexes by the age of 16 years.
Although sun protection should be emphasized for all children and adolescents, increased attention appears to be merited for boys in general and for girls specifically during the adolescent years, and for body sites intermittently exposed to the sun.
Nevi are a risk factor for melanoma and other forms of skin cancer, and many of the same factors confer risk for both. Understanding childhood nevus development may provide clues to possible causes and prevention of melanoma.
To describe nevus acquisition from the ages of 3 to 16 years among white youths and evaluate variation by sex, Hispanic ethnicity, and body sites that are chronically vs intermittently exposed to the sun.
Design, Setting, and Participants
This annual longitudinal observational cohort study of nevus development was conducted between June 1, 2001, and October 31, 2014, among 1085 Colorado youths. Data analysis was conducted between February 1, 2015, and August 31, 2017.
Main Outcomes and Measures
Total nevus counts on all body sites and on sites chronically and intermittently exposed to the sun separately.
A total of 557 girls and 528 boys (150 [13.8%] Hispanic participants) born in 1998 were included in this study. Median total body nevus counts increased linearly among non-Hispanic white boys and girls between the age of 3 years (boys, 6.31; 95% CI, 5.66-7.03; and girls, 6.61; 95% CI, 5.96-7.33) and the age of 16 years (boys, 81.30; 95% CI, 75.95-87.03; and girls, 77.58; 95% CI, 72.68-82.81). Median total body nevus counts were lower among Hispanic white children (boys aged 16 years, 51.45; 95% CI, 44.01-60.15; and girls aged 16 years, 53.75; 95% CI, 45.40-63.62) compared with non-Hispanic white children, but they followed a largely linear trend that varied by sex. Nevus counts on body sites chronically exposed to the sun increased over time but leveled off by the age of 16 years. Nevus counts on sites intermittently exposed to the sun followed a strong linear pattern through the age of 16 years. Hispanic white boys and girls had similar nevus counts on sites intermittently exposed to the sun through the age of 10 years, but increases thereafter were steeper for girls, with nevus counts surpassing those of boys aged 11 to 16 years.
Conclusions and Relevance
Youths are at risk for nevus development beginning in early childhood and continuing through midadolescence. Patterns of nevus acquisition differ between boys and girls, Hispanic and non-Hispanic white youths, and body sites that are chronically exposed to the sun and body sites that are intermittently exposed to the sun. Exposure to UV light during this period should be reduced, particularly on body sites intermittently exposed to the sun, where nevi accumulate through midadolescence in all children. Increased attention to sun protection appears to be merited for boys, in general, because they accumulated more nevi overall, and for girls, specifically, during the adolescent years.
The incidence of melanoma increased dramatically in the United States during the last 40 years.1 The lifetime risk of developing in situ or invasive melanoma is 1 in 28,2 representing the fifth leading cancer among men and the sixth leading cancer among women.3 Following current trajectories, new cases of melanoma are expected to increase to 112 000 annually in the United States by 2030, and treatment expenditures are projected to increase by 252%, reaching $1.6 billion.4 The incidence of squamous and basal cell carcinomas has likewise increased in the United States.5,6 Although mortality is not high for these cancers, morbidity and costs are.6
Insight into the possible causes and prevention of melanoma is informed by a deeper understanding of the development of melanocytic nevi because nevi are an established risk factor for melanoma7-10 and basal cell carcinoma,11 and many of the same factors confer risk for nevi and skin cancers.12-20 The extant evidence indicates that nevi increase with age21-28 and are more numerous among boys than girls20,21,23,25,27 (although this can vary by body site24), those with lighter phenotypes (light skin, eyes, and hair),21,23-25,29,30 and those who receive higher levels of exposure to UV light.20,24,27,31-34 However, this evidence is based largely on studies that are cross-sectional,24 are longitudinal but focused on a relatively short period of childhood or adolescence,23,27,30 examined nevi intermittently across a range of years,21,26,32 or evaluated nevi only of a certain size or on certain body sites.21,23,26,27,30,31 Characterizing nevus count trajectories during childhood may suggest developmental windows of variable risk and important time points for intervention. Identification of high-risk demographic, phenotypic, or exposure groups has implications for tailoring interventions for skin cancer prevention.21,22
A previous study reported median nevus counts among non-Hispanic white children aged 3 to 8 years in the Colorado Kids Sun Care Program cohort who were born in 1998 in the Denver region of Colorado and attended 1 or more annual skin examinations between 2001 and 2006.25 In general, counts increased by 4 to 6 nevi each year. Between 6 and 8 years of age, boys had 4 to 5 more nevi on average than did girls. The sex difference was apparent for total nevi, nevi smaller than 2 mm, and nevi on body sites chronically exposed to the sun, but not for nevi 2 mm or larger or nevi on body sites intermittently exposed to the sun. The current analysis extends the 2009 report25 by analyzing nevus development from age 3 to 16 years and using statistical techniques that take into account correlated data, test for linear and nonlinear patterns of development, and evaluate sex and ethnicity (Hispanic and non-Hispanic white) differences in acquisition patterns for total nevus counts and counts on body sites intermittently and chronically exposed to UV light.
The Colorado Kids Sun Care Program cohort has been described previously.25 Among 1595 children from the Denver region of Colorado who enrolled through 2 waves of recruitment at birth and at 5 to 6 years of age,35,36 this analysis includes 1085 non-Hispanic and Hispanic white participants who completed at least 1 skin examination between June 1, 2001, and October 31, 2014. The Colorado Multiple Institutional Review Board and the Kaiser Permanente of Colorado Institutional Review Board reviewed and approved study procedures. Parents provided written informed consent; children 7 years or older provided written informed assent.
Using established protocols, trained and monitored health care professionals and study staff (N.L.A., J.G.M., S.T.M., J.A., R.P.D., and L.A.C.) recorded nevus counts, hair color, and eye color during annual skin examinations.25 Examiners recorded nevi of all sizes on a standard body map and participated in duplicate examinations each year to determine interrater reliability, which ranged from 0.78 to 0.97 between 2001 and 2014. Parents reported their child’s race/ethnicity as either non-Hispanic white, Hispanic white, black, Asian or Pacific Islander, Native American, or other. Owing to funding gaps, only a subsample of the cohort participated in skin examinations in 2009 (Table 1).
Statistical analysis was conducted between February 1, 2015, and August 31, 2017. We conducted analyses separately for Hispanic and non-Hispanic white participants after excluding participants from other racial/ethnic groups because of their small numbers and substantially lower nevus counts.25 Outcomes included nevus counts on all body sites, sites chronically exposed to the sun (face, anterior neck, lateral forearms, dorsa of hands, and posterior neck on boys only), and sites intermittently exposed to the sun (chest and abdomen, back and shoulders, lateral upper arms, legs, dorsa of feet, medial aspects of arms, palms, bottoms of feet, and posterior neck on girls only).18
We logarithmically transformed nevus counts to satisfy regression assumptions about normally distributed residuals and constant variance of residuals over the range of covariate values. We evaluated whether mean nevus counts varied by the number of completed skin examinations or age at last examination. No such variation was observed, suggesting that the data were consistent with data missing at random.37 We used SPSS for Windows, version 23 (SPSS Inc), to perform variable transformations and descriptive analyses, and SAS SAS/STAT software, version 9.4, of the SAS System for Windows (SAS Inc) for generalized linear mixed model regression analyses and plots. Generalized linear mixed model analyses were conducted using SAS PROC GLIMMIX, which accounts for correlated observations within a child over time.38 To facilitate model convergence, we specified a fourth-order orthogonal polynomial model with a random intercept and random slopes for the linear through quartic trends. We regarded the fourth-order polynomial model (fitted in both means and covariances) as sufficiently complex to describe our data. We used effect cell coding for participant sex to further facilitate model convergence.39 For each outcome, we conducted likelihood ratio tests to evaluate the statistical significance (α = .05, 2-sided) of the set of 4 time trend × sex interaction terms. If the set of interaction terms was significant, all were retained and model evaluation was terminated. Otherwise, all interaction terms were dropped and a main effects–only random effects model was reestimated. Inspection of conditional residual plots and influence statistics for each model confirmed that statistical assumptions were satisfactorily met and that there were no outliers or influential observations. We made predictions based on the population average form of each model and back transformed to the original nevus count scale. For each nevus outcome and sample subgroup, we present R2 values (total variance explained) for the model overall, as well as R2 values and associated P values for each term within the model. Contributions of each covariate to the total variance were estimated using an R2 statistic developed for mixed models.40
To assess sex differences in longitudinal trends as a function of body site among non-Hispanic white youths, the model included all the time trend main effects and sex × time trend interactions described, in addition to a set of 3-way interaction terms involving sex × each time trend × body site (chronic vs intermittent exposure to the sun). Finally, we reestimated our analysis of total nevus counts among non-Hispanic white participants in 2 sensitivity analyses: one that excluded cases (n = 113) from a health care professional in 2013 who was found to have undercounted nevi, and another that excluded 2009 data when only a subsample (n = 167) participated owing to funding constraints and a decrease in median nevus counts was observed. Those analyses sought to determine whether each subsample unduly affected overall trends.
Table 1 shows the characteristics of the youths included in this analysis. Except for 2009, when only a subsample was invited to participate, annual participation between 2004 and 2014 ranged from 48.2% to 76.1% in the white subgroup. Participation between 2001 and 2003 was substantially lower. The median number of skin examinations undergone by a child was 7 (range, 1-14). Across the years, white participants were largely non-Hispanic (221 of 268 [82.5%] to 116 of 132 [87.9%]) and evenly distributed by sex. Approximately 70% had higher-risk eye colors (blue or green), and approximately 55% had higher-risk hair colors (lighter). Unadjusted geometric mean total nevus counts increased over time, reaching a mean of 79.0 in 2014 at age 16 years. Exceptions included slight dips in total, chronic, and intermittent mean nevus counts in 2009 (11 years of age) and 2013 (15 years of age), which may be explained by data collection artifacts (described in Data Analysis).
Figure 1A shows trends in total body nevus counts for non-Hispanic white boys and girls from general linear mixed model analyses. Although the major trend was linear (R2 = 0.92; P < .001) among both non-Hispanic boys and girls, there was deviation from linearity (quadratic, R2 = 0.54; P < .001; cubic, R2 = 0.09; P < .001; and quartic, R2 = 0.04; P = .04) (Table 2). Significant sex differences showed that boys accumulated nevi at a faster rate than did girls beginning at approximately 6 years of age and continuing through 16 years of age (sex, R2 = 0.01; P = .04; sex × quadratic trend, R2 = 0.01; P = .01; sex × cubic trend, R2 = 0.01; P = .04). Non-Hispanic children gained about 3 to 5 nevi per year from 3 through 7 years of age; after that, they gained a mean of 6 to 7 nevi per year.
Nevus counts were lower among Hispanic white children but also followed a largely linear trend (R2 = 0.93; P < .001) that varied by sex (sex × linear trend, R2 = 0.11; P = .01). There was also some evidence of deviation from linearity (quadratic time trend, R2 = 0.32; P < .001) (Table 2 and Figure 1B). Between 3 and 10 years of age, white Hispanic boys accumulated nevi at a faster rate than did girls. Thereafter, the rate of nevus acquisition among girls was the same as or higher than that among boys in this subgroup. On average, Hispanic children acquired 1 to 3 nevi per year from 3 to 8 years of age. Starting around 9 years of age, this increased to 4 to 5 nevi per year. At 16 years of age, nevus acquisition appeared to decrease to around 3 per year.
Nevus counts on sites chronically exposed to the sun increased over time but leveled off by 16 years of age among non-Hispanic white youths (Figure 2A). Accumulation of nevi was higher among boys than among girls at all ages (R2 = 0.12; P < .001) and included linear (R2 = 0.88; P < .001), quadratic (R2 = 29; P < .001), and quartic (R2 = 0.06; P < .001) time trends, which varied by sex (sex × linear trend, R2 = 0.001; P = .03; sex × quadratic trend, R2 = 0.02; P < .001; sex × cubic trend, R2 = 0.03; P < .001) (Table 2). Trends were likewise largely linear among Hispanic white participants (R2 = 0.87; P < .001) with some evidence of curvature (quadratic trend, R2 = 0.28; P < .001; quartic trend, R2 = 0.05; P = .04) and showed a leveling off at the oldest ages (Figure 2B). Median counts among Hispanic white youths were higher among boys than among girls (R2 = 0.02; P = .01), but there were no statistically significant sex differences in patterns of accumulation of nevi (Table 2).
Nevus counts among non-Hispanic white youths followed a strong linear pattern (R2 = 0.90; P < .001) throughout the study period, with evidence of some upward curvature at the oldest ages (quadratic trend, R2 = 0.53; P < .001; cubic trend, R2 = 0.24; P < .001) (Figure 3A and Table 2). Counts were higher among girls than among boys (R2 = 0.01; P = .05), but there were no sex differences in patterns of accumulation (ie, no sex × time trend interactions). A similarly linear pattern was observed among Hispanic white participants (R2 = 0.88; P < .001), although the rate of nevus accumulation was slower, and there was evidence of curvature (quadratic trend, R2 = 0.17; P < .001) (Figure 3B). Time trends differed by sex (sex × linear trend, R2 = 0.09; P = .02); Hispanic white boys and girls had similar counts through 10 years of age, but increases thereafter were steeper for girls, with counts surpassing those of boys from 11 to 16 years of age. Among non-Hispanic white participants, sex differences in nevus count trends did not significantly vary between body sites chronically exposed to the sun and body sites intermittently exposed to the sun (data not shown).
Findings were similar when we repeated the analyses of total body nevus counts among non-Hispanic white youths after excluding cases from the health care professional who undercounted nevi in 2013 (n = 113) and cases from the subsample that participated in 2009 data collection (n = 167).
This study characterizes nevus development in a large cohort of non-Hispanic and Hispanic white youths between 3 and 16 years of age in Colorado. To our knowledge, no previous longitudinal studies of children have reported full-body counts of nevi of all sizes annually for a 13-year span.25,41 By use of general linear mixed model regression analyses, our findings point to developmental stages of greater and lesser nevus acquisition, illuminating ages of potential vulnerability for increasing risk of melanoma and suggesting increased attention to preventive measures during those periods of greater vulnerability.
Nevus acquisition across all body sites was largely linear between 3 to 16 years of age. There was, however, evidence of curvature suggestive of variability in the rate of nevus acquisition. Departures from linearity were apparent for body sites chronically exposed to the sun, where nevus acquisition leveled off among both non-Hispanic and Hispanic white youths at older ages. Curvature was also evident for sites intermittently exposed to the sun among older adolescents.
There are several possible explanations for the different patterns observed on body sites chronically vs intermittently exposed to the sun. First, the plateau in midadolescence is consistent with a previous study suggesting that nevus development on body sites chronically exposed to the sun is limited by a “saturation level” for UV damage, after which ongoing exposure to UV light has little effect.42 In this scenario, skin that is intermittently exposed to the sun may take longer to reach threshold levels. Second, it is possible that skin that is chronically exposed to the sun adapts to persistent UV exposure in a way that is less favorable for nevus formation. For example, DNA repair pathways and other skin defense mechanisms may be augmented in skin that is chronically exposed to the sun. There is evidence that DNA repair mechanisms are better in darker skin.43 Finally, skin that is intermittently exposed to the sun may be particularly affected by increases in UV exposure that often occur in adolescence (especially among girls) related to changes in clothing styles, intentional tanning, and reduced sun protection.44 Ongoing observation of nevus acquisition patterns through adolescence and young adulthood is necessary to determine the continuity of these trends, and additional research at the molecular and behavioral levels is necessary to distinguish among possible mechanisms.
The present findings are consistent with those of previous reports showing lower counts of nevi in Hispanic white children compared with non-Hispanic white children25,41 and reports from other studies showing differences by sex.20,21,23,25,27 Nevus counts on all body sites and on sites that were chronically or intermittently exposed to the sun were lower in all years among Hispanic white youths compared with non-Hispanic white youths. These findings converge with considerably lower rates of melanoma among Hispanic populations compared with non-Hispanic white populations.1,45 Moreover, beginning at approximately 6 years of age and continuing through 16 years of age, non-Hispanic white boys accumulated nevi at a faster rate than did non-Hispanic white girls, overall and on body sites chronically exposed to the sun. We observed a similar trend in nevus counts among Hispanic white participants for body sites chronically exposed to the sun. These findings are congruent with overall higher rates of melanoma for males. In contrast, median nevus counts on body sites intermittently exposed to the sun were higher among girls compared with boys, across all ages among non-Hispanic white participants and beginning at 11 years of age among Hispanic white participants. This difference may reflect patterns of clothing and other UV exposure behaviors that increase risk among adolescent girls and may be associated with the higher incidence of melanoma among Hispanic and non-Hispanic females compared with their male counterparts prior to 55 years of age.45 Our study includes behavioral data that will be analyzed to elucidate this observation.
Although we did not observe a peak or plateau of total body nevus counts, it is generally expected that one will be observed at some point in early adulthood. Childhood and adolescent skin may be distinguished from adult skin by its inherent growth potential that may provide a fertile microenvironment for nevus formation after UV exposures that ultimately result in activating mutations in oncogenes such as BRAF and NRAS that initiate nevi.46 Exposure to UV light induces expression of growth factors that affect melanocyte proliferation.47 These same growth factors may be expressed at higher levels in childhood and adolescence than at older ages.
Our findings support the commonly accepted notion that exposure to UV light during childhood and adolescence should be minimized to reduce the risk of melanoma later in life. Although true for all youths, the need to limit exposure to UV light is even more imperative among non-Hispanic white youths and especially boys, who accumulate nevi on body sites chronically exposed to the sun at a faster rate. The findings also highlight the importance of adolescence in the development of nevi on body sites intermittently exposed to the sun, where counts continued to increase through 16 years of age and did so at a faster rate among girls than among boys regardless of ethnicity. The differences in nevus acquisition in skin that is chronically vs intermittently exposed to the sun are consistent with the dual pathways hypothesis that suggests major differences in melanomas associated with long-term high levels of UV exposure on skin that is chronically exposed to the sun vs melanomas associated with intermittent high doses of UV light on skin that is occasionally exposed to the sun.42 Overall, the patterns observed in our study underscore the continued need for prevention efforts focused on modifying clothing norms and UV exposure practices such as tanning, which may be more prevalent among girls than boys. In addition, given the increasing awareness about the need for targeted melanoma prevention among Hispanic individuals,45 these results are an important step in characterizing risk trajectories in this demographic group.
The present findings must be considered in light of the study limitations. First, the study was conducted with children living in Colorado, a state with a high altitude and more than 300 days of sunshine each year, and with rates of melanoma that exceed the national rate.48 Nevus development may be different among children living in other geographical regions. Second, the study included a relatively small number of Hispanic white youths. The patterns identified in that subgroup should be replicated in a larger sample to ensure that important variations in nevus development over time and by sex were not missed owing to a lack of statistical power, especially for nevi on body sites chronically exposed to the sun. Finally, continued longitudinal observation of this cohort is critical to determine whether there is a sustained plateau of nevus development on body sites chronically exposed to the sun, a continued increase on body sites intermittently exposed to the sun, and whether sex differences persist in nevus counts on both body sites.
These limitations notwithstanding, the strengths of the present investigation lie in its longitudinal design involving full-body nevus examinations with high interrater reliability over a long follow-up period from early childhood through 16 years of age in a large cohort that includes Hispanic and non-Hispanic white youths. We used powerful statistical techniques that account for correlation of observations within children and allow for testing of nonlinear trends in nevus acquisition.
These findings represent an important advance in the understanding of nevus development among Hispanic and non-Hispanic white youths and a contribution to the extant literature on risk factors for melanoma and the science of skin cancer prevention. Our results are novel in showing that nevus counts on body sites intermittently exposed to the sun are higher among non-Hispanic white girls than among non-Hispanic white boys between 3 and 16 years of age and are higher among Hispanic white girls than among Hispanic white boys between 11 and 16 years of age. To our knowledge, we are also the first US study to report a leveling off of nevus development on sites chronically exposed to the sun during adolescence. Future analyses from this cohort will add to this knowledge base by examining how trajectories of nevus development during childhood and adolescence vary as a function of differences in genotype, phenotype, UV exposure patterns, and sun protection practices.
Accepted for Publication: July 13, 2018.
Corresponding Author: Lori A. Crane, PhD, Department of Community and Behavioral Health, University of Colorado Anschutz Medical Campus, 13001 E 17th Pl, Box B119, Bldg 500, Room C3000D, Aurora, CO 80045 (firstname.lastname@example.org).
Published Online: September 12, 2018. doi:10.1001/jamadermatol.2018.3027
Author Contributions: Drs Asdigian and Barón had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Asdigian, Morelli, Crane.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Asdigian, Box, Crane.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Asdigian, Barón, Muller.
Obtained funding: Crane.
Administrative, technical, or material support: Asdigian, Mokrohisky, Aalborg, Dellavalle, Daley, Berwick, Box, Crane.
Supervision: Asdigian, Daley, Box, Crane.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported in part by grant RO1-CA74592 from the National Cancer Institute (Dr Crane). Dr Muller was supported by grants R01-GM121081 and 1R25GM111901 from the National Institute of General Medical Sciences and grant 1G13LM011879 from the National Library of Medicine. The University of Colorado Department of Dermatology provided funding for data collection in 2009.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: We thank the parents and children who participated in the Colorado Kids Sun Care Program between 1998 and 2014, and the nurses and physicians who conducted the skin examinations. Myles Cockburn, PhD, Cancer Prevention and Control Program, University of Colorado Cancer Center, University of Colorado Anschutz Medical Campus, provided comments on an earlier version of this article; he was not compensated for his contributions. Ned Calonge, MD, MPH, The Colorado Trust, collaborated in the original study design and participant recruitment at Kaiser Permanente of Colorado; he was compensated for his role as a coinvestigator.