Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Sep 1.
Published in final edited form as: J Adolesc Health. 2023 Jun 8;73(3):543–552. doi: 10.1016/j.jadohealth.2023.04.014

A Developmental Science Approach to Informing Age Subgroups in Adolescent and Young Adult Cancer Research

Elizabeth J Siembida a,b, Kaitlyn M Fladeboe c,d, Edward Ip e,f, Brad Zebrack g,h, Mallory A Snyder i, John M Salsman f,j
PMCID: PMC10524106  NIHMSID: NIHMS1898864  PMID: 37294255

Abstract

Purpose:

Adolescent and young adult (AYA; diagnosed ages 15-39) cancer survivors are developmentally heterogenous, and this population consists of at least three distinct theoretically informed subgroups: adolescents, emerging adults, and young adults. However, there are limited evidence-based recommendations for delineating the validity of these subgroups in cancer-specific research. We sought to inform recommended chronological age ranges for each subgroup based on developmental processes.

Methods:

The data were collected using a 2x3 stratified sampling design (on- vs. off-treatment; ages 15-17, 18-25, 26-39) and a cross-sectional survey. AYAs (N=572) completed three subscales of the Inventory of Dimensions of Emerging Adulthood (Identity Exploration, Experimentation/Possibilities, Other-Focused), and we used regression tree analyses to identify distinct shifts in mean subscale scores that would indicate unique subgroups. Models included (a) chronological age, (b) chronological age + cancer-related variables, and (c) chronological age + sociodemographic/psychosocial variables as predictors of each developmental measure.

Results:

The recommended age ranges for AYA survivors receiving active treatment were consistent with prior research: adolescents ages 15-17, emerging adults ages 18-24, and young adults ages 25-39. Models for off-treatment survivors suggested four distinct subgroups: adolescents ages 15-17, emerging adults ages 18-23, and ‘younger’ (ages 24-32) and ‘older’ young adults (ages 33-39). No sociodemographic or psychosocial variables meaningfully shifted these recommendations.

Discussion:

Our results suggest that three developmental subgroups remain appropriate for on-treatment survivors, but a second young adult subgroup (ages 33-39) emerged for off-treatment survivors. Therefore, development disruptions may be more likely to occur or manifest in post-treatment survivorship.

Keywords: cancer, adolescent and young adult, development, developmental science

In 2006, the National Cancer Institute and Lance Armstrong (now Livestrong) Foundation’s Adolescent and Young Adult Oncology Progress Review Group defined adolescent and young adult (AYA) cancer survivors as individuals diagnosed between 15-39 years [1]. Although they recommended that narrower subpopulations of AYAs should be examined whenever possible, few studies use subgroups [2–4], and subgroup definitions are not consistent [2, 5–8]. The most typical delineation is three subgroups – adolescents, emerging adults, and young adults – but there is limited evidence to support these subgroups in cancer-related research. Identifying appropriate subgroups within the AYA population will provide an evidence-based recommendation for individuals conducting AYA cancer research.

Developmental science provides strong theoretical and methodological foundations to accomplish this goal. The Life Course Health Development (LCHD) framework posits that healthy development, including physical and mental well-being, is influenced by the timing of health-related exposures in the life course [9]. Adolescence, emerging adulthood, and young adulthood are distinct developmental periods, and a cancer diagnosis in any of these periods may shift an individual’s developmental trajectory away from the expected timing; AYA survivors may delay educational plans, take longer to achieve financial independence, or delay adult social roles like marriage or parenthood [10,11]. Without evidence-based recommendations for defining each life stage, it can be difficult to design research studies that elucidate these variations and develop appropriate interventions.

Each developmental group encompassed within the AYA age range is associated with unique developmental processes – or the cognitive, emotional, and social changes that occur across the lifespan – that vary within and across sociodemographic characteristics (e.g., race/ethnicity, gender) [12]. Adolescence (commonly ages 13-17) is characterized by improvements in cognitive development [13– 16]; increases in emotional lability and impulsivity [17]; and prioritization of peer relationships as a mechanism for belonging, identity exploration, and social skill development [18]. Emerging adulthood (commonly ages 18-25) is characterized by change and instability [19,20]. Individuals experiment with identities and values, and pursue multiple vocational paths [12,20]. Emerging adults continue to develop the self-regulation, decision-making, and complex social abilities associated with the final phases of frontal lobe development [21–23]. Young adulthood (commonly ages 26-39) marks the end of this turbulent period as individuals settle into adult roles, complete cognitive and emotional development begun in adolescence, and embark on building careers and stable romantic partnerships [24–26]. Finally, young adulthood is increasingly characterized by conflicting demands of work and family [27]. It is likely that individuals diagnosed with cancer in each of these life stages will have unique experiences and priorities.

We sought to inform recommended chronological age ranges for AYA subgroups for use in AYA cancer research by examining developmental processes among AYA survivors purposefully recruited to span the entire AYA age spectrum. First, we empirically identified chronological age ranges that delineated distinct subgroups based on three developmental process measures critical to these stages of life. Second, we explored if adding cancer-related factors influenced the identified chronological age ranges. Finally, we explored if adding sociodemographic and psychosocial factors influenced the recommended chronological age ranges. Our hope is this study will catalyze additional attempts to create evidence-based recommendations for defining developmental subgroups in AYA cancer research, and inform standardized definitions to allow for cross-study comparisons.

Methods

Data Collection and Study Sample

Data were collected from a sample of AYAs using stratified sampling by two dimensions: treatment status (on- vs. off-treatment) and current age group (adolescents [15-17 years], emerging adults [18-25 years], young adults [26-39 years]). AYAs were eligible if they were diagnosed with cancer (any stage) between the ages of 15-39, were currently receiving treatment or less than five years posttreatment, were between the ages of 15-39 at time of survey, could read and speak English, had internet access, and could provide consent. AYAs were excluded if they were diagnosed with basal cell carcinoma.

Eligible AYAs were identified, recruited, and consented by Opinions 4 Good (Op4G), an online research panel, for the cross-sectional online survey. Op4G partners with national health-related non-profit organizations to recruit panel members from high-, middle-, and low-socioeconomic strata by contacting organization donors, volunteers, and the communities they serve. Respondents could opt out of the survey at any time. They selected a non-profit organization for an Op4G donation as an incentive. To ensure data quality, we eliminated surveys suggestive of invalid responding (e.g., survey completion time < one-third median completion time) and excluded surveys missing >10% of the items. The study procedures were approved by the Northwestern University Institutional Review Board (Protocol #: IRB00035377), and data from this study is available from the corresponding author upon reasonable request.

Measures

Developmental processes.

To explore the chronological age ranges that delineate subgroups, we used three subscales of the Inventory of Dimensions of Emerging Adulthood (IDEA) [28]: Identity Exploration (7 items; Cronbach’s alpha [α]=0.80; example item: “Is this period of your life a time of defining yourself?”); Experimentation/Possibilities (5 items; α=0.74; example item: “Is this period of your life a time of trying out new things?”); and Other-Focused (3-items; α=0.73; example item: “Is this period of your life a time of commitments to others?”). AYAs were asked to think about this present time in their life and indicate if a particular item describes it well. Response options ranged from Strongly Disagree to Strongly Agree. The IDEA subscales have been used in prior research to distinguish between subgroups of emerging adults [29,30]. The Identity Exploration and Experimentation/Possibilities subscales represent developmental processes most central to emerging adulthood. These subscales were selected because a distinct increase in scores may indicate the transition from adolescence to emerging adulthood. Similarly, the Other-Focused subscale was included because individuals are expected to prioritize others’ needs more as they transition from emerging adulthood into young adulthood; a distinct increase in this subscale may indicate the transition from emerging adulthood to young adulthood. For each subscale, we calculated a mean score (Range=1 to 4), with higher scores indicating the developmental process more strongly describes their life stage.

Cancer variables.

To align with the LCHD framework, we selected cancer-related variables that may influence developmental timing. Treatment status was defined as on- vs. off-treatment. Treatment modality was dichotomized into one treatment modality vs. combined treatment modalities (i.e., self-report of receiving two or more different types of treatment). Cancer type was measured using a dichotomous variable comparing hematologic malignancies (Hodgkin lymphoma, leukemia, myeloma, non-Hodgkin’s lymphoma) vs. primary solid tumors (bladder, bone tumors + sarcomas, brain, breast, central nervous system tumor, cervical, colorectal, esophageal, head and neck, hepatobiliary, kidney, lung, melanoma, ovarian, stomach, testicular, thyroid).

Sociodemographic and psychosocial variables.

We analyzed sociodemographic (primarily social milestones) and psychosocial variables that may influence developmental timing. We measured four sociodemographic characteristics via self-report: education level (≤ high school; some college; ≥ college), sex (male, female), marital status (single vs. married/living with a partner), and childrearing status (currently raising child <18 years vs. not currently raising child <18 years). We measured two psychosocial variables using single items. Subjective maturity was assessed using the item “Do you think you’ve reached adulthood?” Participants were able to answer “Yes,” “No,” or “In some respects yes, and some respects no.” This item has been previously used to differentiate younger and older AYAs [20,31,32]. We measured the impact of cancer with a single global impact item, “Overall, how much has having your illness affected your views about yourself and your life?” Response options ranged from “not at all” to “very much,” with higher scores representing greater impact of illness on their life [33,34].

Statistical Analysis

The exploration of chronological age ranges was conducted using regression tree models [35]. The tree-based model is guided by a partition of the predictor space into nonoverlapping segments which correspond to the terminal nodes or leaves of the tree. The partitioning is done recursively, and at each step, the parent node is split into child nodes through selection of a predictor variable and a split value that minimizes the variability in the response across the child nodes [36]. When there was more than one predictor in the model, the process would continue for the next selected predictor and cycle through the entire list of predictors. The iterative procedure results in a full-grown tree, which is then pruned using cross-validation to avoid overfitting. Quality of model fit was based on the Averaged Square Error (ASE); ASE is the total squared deviation between expected and observed values divided by the sample size. Lower values of ASE indicate better model fit, and we selected the tree model with the lowest ASE on cross-validation. Across our regression tree models, ASE scores ranged from 0.32 to 0.48, and represented reasonable model fit. We also used the metric of Variable Importance to report the most important variable as determined by the tree model. This metric is data-specific but can be used to assess relative predictive power of different variables (e.g., age vs. cancer variables) within the same model. Finally, we used a range of possible age subgroup sample sizes to calculate the detectable effect sizes between subgroups. With our sample size, we can detect small to moderate effect sizes (0.24-0.40; per Cohen’s criteria).

We created separate models for each of the three IDEA subscales as a dependent variable, and then applied the tree-based nonparametric regression using (a) chronological age only (the baseline model), (b) chronological age together with the cancer variables, and (c) chronological age together with the sociodemographic and psychosocial variables, respectively, as independent variables (predictors). The regression-tree procedure was implemented in SAS v9.4 PROC HPSPLIT.

Results

Overview of Sample

The final sample (Table 1) includes 572 AYAs with a mean current age of 23.6 years [Standard Deviation (SD)=7.1]. Across the total sample, AYAs reported an average score of 2.76 (SD=0.63) on the Identity Exploration subscale, 2.78 (SD=0.70) on the Experimentation/Possibilities subscale, and 2.40 (SD=0.80) on the Other-Focused subscale.

Table 1.

Sociodemographic and clinical characteristics of overall sample

Overall Sample N = 572
Sex (N, %)
  Male 323 (56)
  Female 249 (44)
Current Living Situation (N, %)
  Live alone 123 (22)
  Live with others 449 (79)
Health Insurance Coverage (N, %)
  No 104 (18)
  Yes 468 (82)
Race (N, %)
  White 441 (77)
  Black 49 (9)
  Asian 24 (4)
  Native Hawaiian/Pacific Islander 4 (1)
  Native American or Alaska Native 10 (2)
  Mixed racial background 33 (6)
  Other race/decline to answer 11 (2)
Ethnicity
  Hispanic 100 (17)
  Not Hispanic 465 (81)
  Decline to answer 7 (1)
Cancer Diagnosis (N, %)
  Hematologic cancers 125 (24)
  Solid tumors 437 (76)
Treatment Status
  On-treatment 294 (51)
  Off-treatment 278 (49)
Two or more treatment types (N, %)
  No 255 (51)
  Yes 245 (49)
Education
  No college 300 (52)
  Some college 109 (19)
  College grad or higher 163 (29)
Marital Status
  Single 454 (79)
  Married/living with a partner 118 (21)
Childrearing Status (N, %)
  Raising child <18 years 105 (18)
  Not raising child <18 years 467 (82)
Subjective Maturity (N, %)
  No, have not achieved adulthood 32 (6)
  In some ways yes, some ways no 332 (58)
  Yes, have achieved adulthood 208 (36)
Overall Impact of Cancer (M, SD) 3.51 (1.01)
Open in a new tab

Base Regression Tree Model

Our first set of regression tree models, the baseline model, included only current chronological age as a predictor variable.

Identity exploration.

The Identity Exploration model (Figure 1a) identified a distinct increase in mean scores at 17.2 years (Mean [M]=2.53 at <17.2 years) and at 33.2 years (M=2.82 between 17.2-33.1 years, M=3.06 at ≥33.2 years).

Figure 1.

Figure 1.

Figure 1.

Figure 1.

Open in a new tab

Regression tree models with Identity Exploration as the dependent variable. (a) Base model with chronological age as the only predictor variables. (b) Cancer model includes chronological age, treatment status, cancer type, and treatment modality as predictor variables. (c) Sociodemographic and psychosocial model includes chronological age, sex, education level, marital status, childrearing status, subjective maturity, and overall impact of cancer as predictor variables.

Experimentation/Possibilities.

The regression tree model for Experimentation/Possibilities (Figure 2a) identified a distinct increase in mean scores at 16.2 years (M=2.48 at <16.2 years, M=2.85 at ≥16.2 years).

Figure 2.

Figure 2.

Figure 2.

Figure 2.

Open in a new tab

Regression tree models with Experimentation/Possibilities as the dependent variable. (a) Base model with chronological age as the only predictor variables. (b) Cancer model includes chronological age, treatment status, cancer type, and treatment modality as predictor variables. (c) Sociodemographic and psychosocial model includes chronological age, sex, education level, marital status, childrearing status, subjective maturity, and overall impact of cancer as predictor variables.

Other-Focused.

Finally, the regression tree model for Other-Focused (Figure 3a) identified distinct increases in mean scores at ages 17.2, 24.1, and 33.2 years. For AYAs <17.2 years, the mean score on the Other-Focused subscale was 1.92, which increased to 2.29 for AYAs between 17.2 and 24.1 years, then increased further to 2.75 for AYAs between 24.1 and 33.2 years, and finally increased again to 3.04 for AYAs ≥33.2 years.

Figure 3.

Figure 3.

Figure 3.

Figure 3.

Open in a new tab

Regression tree models with Other-Focused as the dependent variable. (a) Base model with chronological age as the only predictor variables. (b) Cancer model includes chronological age, treatment status, cancer type, and treatment modality as predictor variables. (c) Sociodemographic and psychosocial model includes chronological age, sex, education level, marital status, childrearing status, subjective maturity, and overall impact of cancer as predictor variables.

Together, results from the baseline models suggest there are four distinct subgroups of AYAs: adolescents (ages 15-17), emerging adults (ages 18-24), ‘younger’ young adults (ages 25-32), and ‘older’ young adults (ages 33-39). Supplemental Table 1 provides a description of the sample categorized by these chronological age ranges, demonstrating notable increases in many social milestones (e.g., childrearing) among ‘older’ young adults compared to ‘younger’ young adults.

Recommended Age Cut-Offs When Including Cancer Variables

Building on our base models, the second set of regression tree models added cancer-specific variables (treatment status, treatment modality, cancer type) in addition to chronological age as predictors. Supplemental Table 2 provides Variable Importance scores for all variables.

Identity Exploration.

For Identity Exploration (Figure 1b), chronological age emerged as the most important predictor variable (Variable Importance=4.69), and the inclusion of the cancer variables did not shift the recommended age ranges significantly. For AYAs <17.2 years, the mean score on the Identity Exploration subscale was 2.49; increased to 2.78 for AYAs between 17.2 and 24.1 years; further increased to 2.88 for AYAs between 24.1 and 33.2 years; and finally increased again to 3.10 for AYAs ≥31.2 years. The cancer variables were most informative for the adolescent population cut-offs (<17.2 years). Within this subgroup, Identity Exploration scores were higher in individuals with hematologic malignancies (M=2.64) and among off-treatment AYAs with solid tumors (M=2.54) compared to on-treatment AYAs with solid tumors (M=2.21).

Experimentation/Possibilities.

For Experimentation/Possibilities (Figure 2b), both chronological age and treatment status were important predictor variables (Variable Importance scores >4.00 for both variables). This suggests the age ranges delineating developmental subgroups vary for on vs. off-treatment AYAs. Among off-treatment AYAs, there was a distinct change in mean scores at age 15.2 years (M=2.65 at <15.2 years) and at 30.1 years (M=2.92 between 15.2-30.1 years, M=3.07 at >30.1 years). In contrast, among on-treatment AYAs, we saw distinct changes in mean scores at 17.2 years (M=2.28 at <17.2 years) and again at 24.1 years (M=2.68 between 17.2-24.1 years, M=2.86 at >24.1 years).

Other-Focused.

Finally, chronological age remained the most important predictor in the Other-Focused model (Figure 3b), and the results aligned with our base model suggesting delineation between subgroups at 17.2, 24.1, and 33.2 years. For AYAs <17.2 years, the mean score on the Other-Focused subscale was 1.93, which increased to 2.34 for AYAs between 17.2 and 24.1 years, then increased further to 2.74 for AYAs between 24.1 and 33.2 years, and finally increased again to 3.05 for AYAs ≥33.2 years.

Overall, these results suggest the chronological age ranges identified in our base models remain accurate for off-treatment AYAs and samples that include both on- and off-treatment AYAs. However, for samples of only on-treatment AYAs, our results suggest the more typical three subgroup delineation (adolescents, emerging adults, and young adults) is most appropriate.

Recommended Age Cut-Offs When Including Sociodemographic and Psychosocial Variables

Our third set of regression tree models also built on the base model by adding sociodemographic and psychosocial variables (education level, sex, marital status, childrearing status, subjective maturity, and global impact of cancer) as predictors in addition to current chronological age. Supplemental Table 3 provides Variable Importance scores for all variables.

Identity Exploration.

For the Identity Exploration subscale (Figure 1c), chronological age remained the most important predictor (Variable Importance=4.55). The same delineation at age 17.2 years (M=2.52 at <17.2 years, M=2.87 at ≥17.2 years) was identified with some meaningful differences in mean scores based on the sociodemographic and psychosocial variables. Among the adolescents (<17.2 years), individuals who reported that cancer had a high impact on their life reported higher mean scores on Identity Exploration (M=2.84) than those who reported less impact on their life (M=2.45). Among the emerging and young adults (≥17.2 years), AYAs who perceived they had reached adulthood reported higher mean scores (M=2.99) than AYAs who did not feel fully like an adult (M=2.76). This difference was not present among AYAs who reported not feeling fully like an adult but were married or cohabitating with a partner.

Experimentation/Possibilities.

The Experimentation/Possibilities model (Figure 2c) remained unchanged from our base model after adding in the sociodemographic and psychosocial variables; chronological age remained the only important predictor (Variable Importance=3.51). There was a distinct increase in mean scores at age 16.2 years (M=2.48 at <16.2 years, M=2.86 at ≥16.2 years).

Other-Focused.

Finally, the results of the Other-Focused model (Figure 3c) also found chronological age was the most important predictor (Variable lmportance=9.26). There were similar cut-offs to our base model with scores distinctly increasing at ages 24.1 and 33.2 years. For AYAs <24.1 years, the mean score on the Other-Focused subscale was 2.09; increased to 2.75 for AYAs between 24.1 and 33.2 years; and increased again to 3.04 for AYAs ≥33.2 years. However, the sociodemographic and psychosocial variables highlighted important differences within the developmental subgroups. Among AYAs <24.1 years, perceiving one had reached adulthood (M=2.56) or having some college education (M=2.33) was associated with higher mean scores on the Other-Focused subscale. Among AYAs ≥24.1 years, AYAs between 24.1 and 33.2 years who were married or cohabitating with a partner (M=2.91) reported significantly higher scores than AYAs who were single (M=2.63).

Taken together, results from these models suggest that although sociodemographic and psychosocial variables (most notably, perceived adult status and romantic partner status) influence AYAs’ developmental processes, the chronological age ranges identified in the base model (i.e., adolescents ages 15-17, emerging adults ages 18-23, ‘younger’ young adults ages 24-33, and ‘older’ young adults ages 33-39) remain the best delineation. Table 2 provides a summary of the recommended age cut-offs for each of our presented models.

Table 2.

Recommended age cut-offs for subgroups based on developmental processes

Identity Exploration Experimentation & Possibilities Other-Focused
Cut-Off Adol to EAa Cut-Off EA to YAb Cut-Off YA to EsAc Cut-Off Adol to EAa Cut-Off EA to YAb Cut-Off YA to EsAc Cut-Off Adol to EAa Cut-Off EA to YAb Cut-Off YA to EsAc
Base model 17.2 33.2 NA 16.2 NA NA 17.2 24.1 33.2
Addition of cancer variables
 Off-Treatment 17.2 24.1 31.1 30.1 NA NA 17.2 24.1 33.2
 On-Treatment 17.2 24.1 NA
Addition of psychosocial variables 17.2 NA NA 16.2 NA NA 24.1 33.2 NA
Open in a new tab
a

Age at which the adolescent subgroup ends and emerging adult subgroup begins

b

Age at which the emerging adult subgroup ends and young adult subgroup begins

c

Age at which the young adult subgroup ends and established adult subgroup begins

Abbreviations: Adol = adolescent; EA = emerging adult; YA = young adult; EsA = established adult; NA = not applicable

Discussion

This is the first empirical study to identify recommended chronological age ranges for developmental subgroups among AYA cancer survivors, and potential variations based on cancer- or psychosocial-related factors. Overall, our results suggest that the three developmental subgroups commonly used in AYA cancer research remain appropriate for AYAs receiving treatment. However, our results suggest the addition of a fourth subgroup (approximately ages 33-39) may be important for AYAs off-treatment or for mixed samples.

The potential addition of a fourth subgroup (ages 33-39) aligns with recent advances in developmental science highlighting differences between the experiences of adults in their mid-twenties to adults in their early/mid-thirties (referred to as, “established adulthood”) [27]. The developmental distinction between these groups occurs across physical, cognitive, and social domains. For example, adults ages 20-30 experience frequent job changes, are typically childfree, and are entering their first committed romantic partnerships [37–39]. In contrast, adults ages mid-30s to 40s are often making decisions about having or raising young children, settled on their career path, and are in long-term committed relationships [38,39] Experiencing cancer may further affect these groupings as AYAs focus solely on cancer treatment, but then refocus on these developmental processes in survivorship.

We found three developmental subgroups (adolescence, emerging adulthood, young adulthood) were most appropriate for AYAs on-treatment, whereas the new four subgroup recommendation was most appropriate for AYAs off-treatment or in mixed samples. These differences may be due to interruptions in developmentally expected social roles during treatment (e.g., leaving full-time schooling) and delays in achievement of certain milestones (e.g., autonomy from parents) [10,11]. Thus, off-treatment AYAs may be “catching up” to their peers and completing normative milestones that were missed or delayed due to diagnosis and treatment [42]. Researchers may consider their target population and specific research question when deciding if three or four subgroups is most appropriate.

The sociodemographic and psychosocial variables we examined did not shift recommendations. However, these variables exemplified the heterogeneity in the timing of developmental processes. For example, Identity Exploration was higher among emerging and young adults who perceived they had reached adulthood compared to emerging and young adults who did not, but this difference did not remain among those who were married or cohabitating with a partner. Research has described normative, typical developmental trajectories and approximate chronological age equivalents for important social milestones [14,20,27]. But, as the LCHD framework articulates, there is heterogeneity within these typical trajectories with some individuals meeting milestones at slightly younger or slightly older ages, and this remains true for AYAs with cancer. Cancer is a critical health-related exposure that can accelerate or decelerate an individual’s trajectory, emphasizing that even within relatively similar subgroups individuals will remain developmentally heterogenous.

Findings should be interpreted considering the limitations. First, our study was cross-sectional; we could not examine intra-individual change in development from onto off-treatment or as AYAs age “out of” and “into” different subgroups. Second, we did not collect data from a comparison group of non-cancer AYAs to directly test how cancer may shift developmental processes. Third, we were unable to examine if developmental subgroups varied by intersecting identities (e.g., Black AYAs off-treatment), and we did not measure important social determinants of health. It is likely that social determinants of health like housing insecurity, exposures to discrimination or childhood trauma, or access to healthcare will influence an AYA’s shifting into or out of developmental subgroups. Additionally, the original research defining developmental subgroups was performed in non-Hispanic White, educated, cisgender populations. Critical reviews have emphasized that processes underlying development likely vary across minoritized and marginalized populations, but limited research exists to elucidate these differences [12]. Fourth, we only used three of the six IDEA subscales for our analysis. We eliminated the negativity/instability, self-focused, and feeling in-between subscales because we believe they conflate meaningfully with the cancer experience. Finally, our sample size and use of a self-report methodology limited the extent to which we could examine the cancer-specific variables, and key variables may have been missed. Future research should examine differences in developmental subgroups by individual cancer types, and utilize electronic health records to robustly explore treatment intensity.

The study represents one of the first attempts at providing empirically supported chronological age cut-offs for subgroups in AYA cancer research. We provide novel recommendations regarding the use of four developmental subgroups (adolescents, emerging adults, young adults, established adults), which aligns with recent trends in developmental science. Treatment status was a key cancer factor that shifted recommended age range cut-offs, emphasizing the importance of considering the target population when designing AYA research studies. We hope future research will expand sociodemographic and psychosocial variables to include nuanced examinations of social determinants of health, sex and gender identity, and cancer experience variables. Future work should also study AYA subgroups longitudinally and examine objective clinical factors (e.g., disease stage, treatment type) using electronic health records to tailor AYA cancer research.

Supplementary Material

1

Implications and Contribution Summary Statement:

These results suggest that the AYA cancer population may include four distinct developmental subgroups (adolescents ages 15-17, emerging adults ages 18-23, young adults ages 24-33, established adults ages 33-39), and research in on-treatment AYAs may consider alternate subgroupings (adolescents ages 15-17, emerging adults ages 18-24, young adults ages 25-39).

Acknowledgements:

The authors would like to thank the adolescent and young adult survivors who generously devoted their time to teach us about their experiences. The corresponding author, Dr. Siembida, also confirms that all individuals who contributed meaningfully to this manuscript has been include in the authorship list.

Funding:

Research reported in this publication was supported by pilot funding from the Department of Medical Social Sciences at the Northwestern University Feinberg School of Medicine. Analyses were supported by the Biostatistics Shared Resource of the Wake Forest Baptist Comprehensive Cancer Center’s NCI Cancer Center Support Grant P30CA012197. Drs Fladeboe, Zebrack, and Salsman are currently supported, in part, by unrelated grants from the National Institutes of Health (Fladeboe: K99CA267481; Zebrack: R01CA261752; Salsman: R01CA242849, R01CA218398). The opinions herein represent those of the authors and not necessarily their funders.

Abbreviations:

AYA

adolescent and young adult

LCHD

Life Course Health Development

Op4G

Opinions 4 Good

IDEA

Inventory of Dimensions of Emerging Adulthood

ASE

Averaged Square Error

SD

Standard Deviation

M

Mean

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of Interest Statement: The authors have no financial or other conflicts of interest to declare, and no funders influenced the writing of this manuscript. The first draft of this manuscript was written by Elizabeth Siembida.

References

  • 1.National Cancer Institute. Closing the gap: Research and cancer care imperatives for adolescents and young adults with cancer (NIH Publication No. 06-6067). Bethesda, MD: Department of Health and Human Services, National Institutes of Health, National Cancer Institute, and the LIVESTRONG Young Adult Alliance; August 2006. Accessed April 4, 2023 from: https://www.cancer.gov/types/aya/research/ayao-august-2006.pdf [Google Scholar]
  • 2.McCarthy MC, McNeil R, Drew S, et al. Psychological distress and posttraumatic stress symptoms in adolescents and young adults with cancer and their parents. Journal Adoles Young Adult Oncol 2016;5(4):322–329. DOI: 10.1089/jayao.2016.0015 [DOI] [PubMed] [Google Scholar]
  • 3.Bøhn SH, Lie HC, Reinertsen KV, et al. Lifestyle among long-term survivors of cancers in young adulthood. Support Care Cancer 2021;29:289–300. DOI: 10.1007/s00520-020-05445-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Chao C, Xu L, Bhartia S, et al. Cardiovascular disease risk profiles in survivors of adolescent and young adult (AYA) cancer: The Kaiser Permanente AYA cancer survivors study. J Clin Oncol 2016;34:1626–1633. DOI: 10.1200/JCO.2015.65.5845 [DOI] [PubMed] [Google Scholar]
  • 5.Kwak M, Zebrack BJ, Meeske KA, et al. Trajectories of psychological distress in adolescent and young adult patients with cancer: A 1-year longitudinal study. J Clin Oncol 2013;31:2160–2166. DOI: 10.1200/JCO.2012.45.9222 [DOI] [PubMed] [Google Scholar]
  • 6.Kwak M, Zebrack BJ, Meeske et al. Prevalence and predictors of post-traumatic stress symptoms in adolescent and young adult cancer survivors: A 1-year follow-up study. Psychooncology 2013;22:1798–1806. DOI: 10.1002/pon.3217 [DOI] [PubMed] [Google Scholar]
  • 7.Kirchhoff AC, Lyles CR, Fluchel M, et al. Limitations in health care access and utilization among long-term survivors of adolescent and young adult cancer. Cancer 2012;118:5964–72. DOI: 10.1002/cncr.27537 [DOI] [PubMed] [Google Scholar]
  • 8.Betts AC, Murphy CC, Shay LA, et al. Polypharmacy and prescription medication use in a population-based sample of adolescent and young adult cancer survivors. J Cancer Survivor 2022;epub. DOI: 10.1007/s11764-021-01161-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Halfon N, Forrest CB. The emerging theoretical framework of life course health development. In: Halfon N, Forrest CB, Lerner RM, Faustman EM, (eds). Handbook of Life Course Health Development Cham, Switzerland: Springer Nature, 2018:19–43. [PubMed] [Google Scholar]
  • 10.Kirchhoff AC, Yi J, Wright J, et al. Marriage and divorce among young adult cancer survivors. J Cancer Surviv 2012;6(4):441–450. DOI: 10.1007/s11764-012-0238-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Gunnes MW, Lie RT, Bjørge T, et al. Reproduction and marriage among male survivors of cancer in childhood, adolescence and young adulthood: a national cohort study. Br J Cancer 2016;114(3):348–356. DOI: 10.1038/bjc.2015.455 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Syed M, Mitchell LL. Race, ethnicity, and emerging adulthood: Retrospect and prospects. Emerg Adulthood 2013;1(2):83–95. DOI: 10.1177/2167696813480503 [DOI] [Google Scholar]
  • 13.Collins WA, Steinberg L. Adolescent development in interpersonal context. In: Eisenberg N, Damon W, Lerner RM, eds. Handbook of Child Psychology: Vol. 3, Social, Emotional, and Personality Development Hoboken, NJ: John Wiley & Sons, Inc., 2006:1003–1067. [Google Scholar]
  • 14.McNeely C, Blanchard J. The teen years explained: A guide to health adolescent development. Baltimore, MD: Center for Adolescent Health at Johns Hopkins Bloomberg School of Public Health, 2009. [Google Scholar]
  • 15.Mello ZR, Bhadare D, Fearn EJ, et al. The window, the river, and the novel: Examining adolescents’ conceptions of the past, the present, and the future. Adolescence 2009;44(175):539–556. [PubMed] [Google Scholar]
  • 16.Piaget J Intellectual evolution from adolescence to adulthood. Hum Dev 2008;51:40–47. (Reprinted from Hum Dev 1972;15:1-12) DOI: 10.1159/000112531 [DOI] [Google Scholar]
  • 17.Chambers RA, Taylor JR, Potenza MN. Developmental neurocircuity of motivation in adolescence: A critical period of addiction vulnerability. Am J Psychiatry 2003;160:1041–1052. DOI: 10.1176/appi.ajp.160.6.1041 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Barrett DE The three stages of adolescence. The High School Journal 1996;79(4):333–339. [Google Scholar]
  • 19.Wood D, Crapnell T, Lau L, et al. Emerging adulthood as a critical stage in the life course. In: Halfon N, Forrest CB, Lerner RM, Faustman EM, (eds). Handbook of Life Course Health Development Cham, Switzerland: Springer Nature, 2018:123–143. [PubMed] [Google Scholar]
  • 20.Arnett JJ Emerging adulthood: A theory of development from the late teens through the twenties. Am Psychol 2000;55(5):469–480. DOI: 10.1037/0003-066X.55.5.469 [DOI] [PubMed] [Google Scholar]
  • 21.Eluvathingal TJ, Hasan KM, Kramer L, et al. Quantitative diffusion tensor tractography of association and projection fibers in normally developing children and adolescents. Cereb Cortex 2007;17:2760–2768. DOI: 10.1093/cercor/bhm003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Steinberg L Should the science of adolescent brain development inform public policy? Am Psychol 2009;64(8):739–750. DOI: 10.1037/0003-066X.64.8.739 [DOI] [PubMed] [Google Scholar]
  • 23.Turner-Henson A Understanding adolescents: A guide for researchers. J Neurosci Nurs 2005;37(3):164–168. [PubMed] [Google Scholar]
  • 24.U.S. Census Bureau. Number, timing and duration of marriages and divorces. Available at: https://www.census.gov/newsroom/press-releases/2021/marriages-and-divorces.html. Accessed on January 6, 2023.
  • 25.National Center for Health Statistics. Age at first birth. Available at: https://www.cdc.gov/nchs/nsfg/key_statistics/b.htm#birthsmothers. Accessed on January 6, 2023.
  • 26.U.S. Bureau of Labor Statistics. Employment status of the civilian noninstitutional population by age, sex, and race. Available at: https://www.bls.gov/cps/cpsaat03.htm. Accessed on January 6, 2023.
  • 27.Mehta CM, Arnett JJ, Palmer CG, Nelson LJ. Established adulthood: A new conception of ages 30 to 45. Am Psychol 2020;75(4):431–444. DOI: 10.1037/amp0000600. [DOI] [PubMed] [Google Scholar]
  • 28.Reifman A, Colwell MJ, Arnett JJ. Emerging adulthood: Theory, assessment and application. J Youth Dev 2007;2(1):Epub. DOI: 10.5195/jyd.2007.359. [DOI] [Google Scholar]
  • 29.Arnett JJ, Mitra D. Are the features of emerging adulthood developmentally distinctive? A comparison of ages 18-60 in the United States. Emerg Adulthood 2018;8(5):1–8. DOI: 10.1177/2167696818810073 [DOI] [Google Scholar]
  • 30.Lanctot J, Poulin F. Emerging adulthood features and adjustment: A person-centered approach. Emerg Adulthood 2017;6(2):1–13. DOI: 10.1177/2167696817706024 [DOI] [Google Scholar]
  • 31.Arnett JJ Emerging adulthood(s): The cultural psychology of a new life stage. In: Jensen LA, ed. Bridging cultural and developmental approaches to psychology: New syntheses in theory, research, and policy. Oxford University Press, 2011:255–275. [Google Scholar]
  • 32.Arnett JJ, Žukauskienė R, Sugimura K. The new life stage of emerging adulthood at ages 18–29 years: Implications for mental health. Lancet Psychiatry. 2014;1(7):569–576. DOI: 10.1016/S2215-0366(14)00080-7 [DOI] [PubMed] [Google Scholar]
  • 33.Yanez B, Garcia SF, Victorson D, Salsman JM. Distress among young adult cancer survivors: A cohort study. Support Care Cancer. 2013;21(9):2403–2408. DOI: 10.1007/S00520-013-1793-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Lai J, Garcia SF, Salsman JM, Rosenbloom Cella D. The psychosocial impact of cancer: Evidence in support of independent general positive and negative components. Qual Life Res 2012:21(2):195–207. DOI: 10.1007/s11136-011-9935-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Breiman L, Friedman J, Olshen RA, Stone CJ. Classification and Regression Trees. Belmont, CA: Wadsworth, 1984. [Google Scholar]
  • 36.The HPSPLIT procedure overview. https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.4/stathpug/stathpug_hpsplit_overview.htm. Last retrieved July 26, 2022.
  • 37.Bureau of Labor Statistics. Median years of tenure with current employer for employed wage and salary workers by age and sex, selected years, 2008–2018. Available at: https://www.bls.gov/news.release/tenure.t01.htm. Accessed on January 6, 2023.
  • 38.Organisation for Economic Co-operation and Development. Marriage and divorce rates. Available at: https://www.oecd.org/els/family/SF_3_1_Marriage_and_divorce_rates.pdf. Accessed on January 6, 2023.
  • 39.Organisation for Economic Co-operation and Development. Age of mothers at childbirth and age-specific fertility. Available at: https://www.oecd.org/els/soc/SF_2_3_Age_mothers_childbirth.pdf. Accessed on January 6, 2023.
  • 40.Patterson P, McDonald FE, Zebrack B, Medlow S. Emerging issues among adolescent and young adult cancer survivors. Semin Oncol Nurs 2015;31(1):53–59. DOI: 10.1016/j.soncn.2014.11.006. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1898864-supplement-1.xlsx (25.6KB, xlsx)

RESOURCES