Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Dec 14.
Published in final edited form as: Dev Psychobiol. 2022 Nov;64(7):e22311. doi: 10.1002/dev.22311

Demographic and health predictors of telomere length during adolescence

Darlene A Kertes 1,2, John Leri 1, Ke Duan 1, Jake Tarrence 3, Christopher Browning 3, Rita Pickler 4, Jodi Ford 4
PMCID: PMC9749139  NIHMSID: NIHMS1854660  PMID: 36282763

Abstract

Telomere length (TL) is proposed to play a mechanistic role in how the exposome affects health outcomes. Little is known about TL during adolescence, a developmental period during which precursors of adult-onset health problems often emerge. We examined health and demographic sources of variation in TL in 899 youth aged 11–17. Demographic and health information included age, sex, race, household income, caregiver age and marital status, youth tobacco exposure, body mass index, pubertal status, health problems, medication use, and season of data collection. Genomic DNA was extracted from saliva, and T/S ratios were computed following qPCR. Age, race, season of data collection, and household income were associated with the telomere to single copy (T/S) ratio. We found that T/S ratios were larger at younger ages, among Black youth, for saliva collected during autumn and winter, and among households with higher income. Analyses stratified by race revealed that the association between age and T/S ratio was present for Black youth, that season of collection was present across races, but that family demographic associations with T/S ratio varied by race. The results provide information for future telomere research and highlight adolescence as a potentially important developmental phase for racial disparities in telomere shortening and health.

Keywords: adolescence, child development, race, season, socioeconomic status, telomere

1 |. INTRODUCTION

The life history of environmental exposures, collectively referred to as the exposome, affects the trajectory of long-term health. Collectively, the exposome can accelerate aging and increase the risk for morbidity and mortality. Telomeres, noncoding DNA sequences that cap the end of chromosomes, can index cellular aging, are affected by the exposome, and have consequences for health outcomes (Srinivas et al., 2020). Telomere length (TL) is believed to reflect the physiological effects of the exposome and potentially play a role in mediating or moderating effects of the exposome on health outcomes.

A robust literature links TL, the exposome, and health outcomes among adults. Psychological stress, discrimination, and traumatic experiences are associated with shorter TL as assessed from saliva, blood, or buccal cells (see for reviews Coimbra et al., 2017; Colich et al., 2020; Mathur et al., 2016; Rentscher et al., 2020). Race/ethnic differences are also observed, such that Black Americans lose telomeres at a faster rate than White Americans from birth to adulthood (Hunt et al., 2008; Roux et al., 2009). Shorter TL has subsequently been linked to adverse health outcomes such as depression, anxiety, posttraumatic stress disorder, cancer, diabetes, Alzheimer’s disease, and cardiovascular disease (see for review Smith et al., 2019).

In recent years, there has been increased scientific attention on factors contributing to individual variation in TL during early life development. Prenatal exposure to maternal stress is associated with newborn cord blood TL (Bosquet Enlow et al., 2018; Marchetto et al., 2016; Verner et al., 2020) as well as with the rate of telomere shortening during infancy assessed via saliva and blood (Bosquet Enlow et al., 2020; de Fluiter et al., 2021). Postnatally, early life adversity is associated with shorter TL in multiple tissue types during childhood and early adolescence (Coimbra et al., 2017; Colich et al., 2020; Rentscher et al., 2020).

Compared to adulthood and infancy and early childhood, relatively little is known about TL during adolescence. During adolescence, changes in peer interactions, cognitive maturation, the emergence of behavioral autonomy, and alterations in peer dynamics (Caballero et al., 2016; Jackson & Goossens, 2020) are observed along with physical maturation. From a lifespan perspective, adolescence is second only to infancy in the degree of rapid developmental change and heightened susceptibility or sensitivity to environmental influences with lifetime physiological and psychological effects (Rapee et al., 2019). Adolescence is also the developmental phase during which early indicators of many adult-onset health problems typically emerge. It is critical to document the effects of the exposome on TL during adolescence so that the intermediary role of TL on health trajectories can be fully understood.

To accurately assess the effect of the exposome on TL, foundational information on individual- and family-level demographics that predict variation in TL is needed. In this report, we document associations between demographic and health variables commonly assessed in epidemiological studies with salivary TL among a large sample of adolescents. The results serve to identify population-level factors that warrant inclusion in studies focused on specific environmental risk and protective factors for the developmentally dynamic but under-studied period of adolescence. Based on documented racial disparities in telomere shortening across the lifespan, demographic associations were tested first for the full sample and then stratified by race to examine within-race demographic variables associated with TL. As the extant literature on TL has historically varied in laboratory methods and quality control checks utilized, this report details the assay techniques and analytic pipeline in a manner designed to enhance reproducibility (Lindrose & Drury, 2020).

2 |. METHODS

2.1 |. Participants and procedure overview

Participants were youth aged 11–17 and their primary caregiver. Participants were recruited using a mix of vendor- and school-based address lists from the Columbus, Ohio area, specifically a contiguous space within the Franklin County core bounded by the outer belt of Interstate 270. Among eligible households contacted, one randomly chosen child and a primary caregiver were selected to participate. Study-related procedures and analyses were reviewed and approved by the institutional review boards at Ohio State University and the University of Florida.

Written parental consent and youth assent were obtained by trained interviewers prior to the start of the study. The study was conducted at the participants’ home to reduce participant burden. Following consent, interviewer- and self-administered surveys were administered to focal youth and their caregivers separately. Trained field technicians collected saliva samples from youth. Saliva specimens were transported from the participant’s home to the survey research center, where they were stored at −20°C and then transferred on dry ice to the −80°C university freezers until assay.

2.2 |. Measures

Demographics were measured using caregiver and adolescent self-reports. Adolescents reported their age, sex, and race. The caregiver reported demographic information, including the caregiver age, annual household income (below $30,000 per year, between $30,000 and $60,000 per year, and above $60,000 per year), and marital status (single, married, cohabitating, separated or divorced, widowed). Due to the small number of widowed caregivers (n = 10), separated/divorced and widowed caregivers were combined into one category for analysis.

General health was assessed from caregiver reports. Caregivers reported on the adolescents’ history of medical diagnoses via a checklist and free-response items based on the National Center for Health Statistics (Centers for Disease Control & Prevention, 2012). The checklist items included allergies, anxiety problems, arthritis, asthma, attention deficit disorder, autism spectrum disorders, autoimmune disorders, behavior and conduct problems, cancer, cerebral palsy, depression, epilepsy, high blood pressure, intellectual disabilities, learning disabilities, diabetes, and a free response for other health conditions. The most reported health conditions were allergies (n = 253), asthma (n = 149), attention deficit hyperactivity disorder (ADHD; n = 157), and anxiety (n = 115). Prescription and over-the-counter medications were also reported by parents. Steroid medications for asthma/allergies were reported for 101 youth; other prescription medications were reported for 306 youth. Given that this was a community-based sample not selected for health conditions, medical diagnoses and prescription medications were summed and dichotomized into any/no conditions and any/no medications. The caregiver also reported on second-hand exposure to tobacco smoke inside the home (dichotomized as present/absent), and youth self-reported tobacco use over the past 12 months (dichotomized).

Based on prior reports of seasonal variation in telomere length (Beaulieu et al., 2017; Rehkopf et al., 2014; Tennyson et al., 2018), the date of salivary DNA collection was used to create an indicator of season. Northern Meteorological Seasons (e.g., Garay et al., 2019) were used to define winter as December through February, spring as March through May, summer as June through August, and fall as September through November, with season collapsed into autumn/winter and spring/summer for analytic purposes.

2.3 |. Salivary telomere length

Passive drool saliva was collected in an OGD-500 collection device (DNA Genotek Inc., Ottawa, Canada). Samples were frozen at −80°C at Ohio State University, transported on dry ice, and assayed at the University of Florida. Whole saliva samples were stored at −80°C from the end of collection until assay (3 years) and did not undergo freeze–thaw cycles until DNA extraction and dilution to working concentrations. Genomic DNA was isolated from 500 μl of saliva via ethanol precipitation and suspended in TE buffer. Initial genomic DNA quality and quantity were determined using UV spectrophotometry (Nanodrop 2000; Thermo Fisher Scientific) and gel electrophoresis (1% TBE) and confirmed via fluorometry (Qubit 3.0; Thermo Fisher Scientific) and an Agilent TapeStation 2200 (Santa Clara, California, USA).

Genomic DNA telomere length was assessed using quantitative polymerase chain reactions (qPCR). The telomere primers were telg (5′-ACACTAAGGTTTGGGTTTGGGTTTGGGTTTGGGTTAGTGT-3′) and telc (5′-TGTTAGGTATCCCTATCCCTATCCCTATCCCTATCCCTAACA-3′) used at a final concentration of 500 nM (Eurofins Genomics, Louisville, Kentucky). The primers for the single-copy gene (albumin) were Albugcr2 (5′-CGGCGGCGGGCGGCGCGGGCTGGGCGGCCATGCTTTTCAGCTCTGCAAGTC-3′) and Albdgcr2 (5′-GCCCGGCCCGCCGCGCCCGTCCCGCCGAGCATTAAGCTCTTTGGCAACGTAGGTTTC-3′) at a final concentration of 225 nM (Eurofins Genomics). The final reaction mix was 10 μl consisting of 5 μl of SsoFast EvaGreen Supermix (Bio Rad, Hercules, California, USA), 4 μl of primer mix, and 1 μl of DNA diluted to 10 ng/μl. A five-point serial dilution (25, 12.5, 6.25, 3.12, 1.56 ng/μl) of HeLa genomic DNA (New England BioLabs, Ipswich, Massachusetts, USA) was used as a reference DNA to create the standard curve. The same reference DNA was used for all qPCR runs. Nontemplate controls and a positive control DNA sample, comprising pooled genomic DNA extracted from saliva, were included on each plate.

All qPCR runs were carried out on an Applied Biosystems StepOne-Plus Real-Time PCR System with a 96-well heating block. The cycling conditions for the telomere reaction were as follows: holding stage: 95°C for 15 min, cycling stage 1 (two cycles): 98°C for 2 s, 49°C for 30 s, cycling stage 2: (34 cycles): 98°C for 2 s, 59°C for 30 s, 74°C for 15 s. Signal acquisition: 72°C for 1 min, 95°C for 15 s. The cycling conditions for the single gene reaction were as follows: holding stage: 95°C for 15 min, cycling stage 1 (2 cycles): 98°C for 2 s, 49°C for 30 s, cycling stage 2 (38 cycles): 98°C for 2 s, 59°C for 30 s, 84°C for 15 s. Signal acquisition: 72°C for 1 min, 95°C for 15 s.

Telomere (T) and single-gene (S) reactions were run on separate plates. The sample position was kept constant across the T and S plates. Between the T and S reactions, diluted working samples were maintained at 4°C for approximately 1 hour. Unknown samples were run in triplicate. Unknown samples for which the cycle threshold (Ct) of all three replicates was outside the bounds of the standard curve were rediluted to 10 ng/μl and assayed again. Unknown samples outside of the standard curve on both assay attempts were excluded from analysis (n = 74). Grubbs’ method (Grubbs, 1969) was used to calculate outlying values within unknown sample triplicates. Specifically, raw Ct values greater than 1.15 values away from the mean Ct value were excluded from downstream analyses (Shalev et al., 2013).

Relative telomere to single-copy gene (T/S) quantities were calculated according to the procedure described by Pfaffl (2001) and detailed by Hellemans and Vandesompele (2011). The slope of the standard curve included on each plate was calculated as y = intercept + slope × x, where y was the Ct value and x was the log of the input DNA concentration (e.g., log(25 ng)). Reaction efficiencies (RE) were calculated via the slope of the standard curve (RE = 10(1/−slope)) such that a value of 2 was equivalent to 100% RE. The average RE for the telomere reaction was 1.87 (mean R2 = 0.99) and 1.90 (mean R2 = 0.99) for the albumin reaction. Relative quantities (RQs) were calculated (RQ = REΔCt) for each sample by finding the difference between the mean Ct value within unknown sample triplicates and the mean Ct value of all unknown samples on the plate. The telomere RQ was then divided by the albumin RQ to produce a T/S ratio for each sample. The unknown T/S ratios were then standardized to the 12.5 ng/μl standard (interplate CV for T = 2.69%; S = 1.93%) included on each plate to yield a calibrated, normalized relative quantity (CNRQ). The intraclass correlation of T/S ratios from a randomly selected subset of 95 samples was 0.62.

T/S ratios were adjusted to account for the effect of well position based on Eisenberg et al. (2015). First, the mean CNRQ value was calculated for each of the 24 unknown-sample positions. The respective mean T/S ratio for each sample position was then subtracted from the CNRQ of each unknown sample. The mean CNRQ for all unknown samples was then added to the T/S ratio quantity for each unknown sample.

Prior to analysis, T/S ratio data were examined for skew and outliers. A positive skew was observed, and log10 transformation was applied. T/S ratios were mean centered and scaled by standard deviation. Values greater than four standard deviations from the mean were winsorized (n = 8).

2.4 |. Statistical analyses

Phenotype data and salivary DNA samples were collected from 1080 adolescents. Telomere data were successfully generated for 995 youth. Of those, 899 participants who self-identified as either Black/African American (N = 411) or White (N = 488) comprised our final analytic sample. Adolescents whose self-identified race was Asian (n = 17), Hispanic/Latino (n = 53), or other (n = 26) were not included in the analyses due to the small numbers in these racial groups to avoid Type I or Type II error.

Statistical analyses were conducted using R version 4.0.5. To accommodate missing phenotype values, multiple imputation was conducted prior to statistical modeling using the “mice” package (van Buuren & Groothuis-Oudshoorn, 2011). Categorical variables with missing data were imputed using polytomous logistic regression (Venables & Ripley, 2002). Continuous variables with missing data were imputed using predictive mean matching (Little, 1988). Regression models were fit on all imputed datasets (n = 100), and model estimates were pooled across imputations (Rubin, 1987). Hierarchical regression analyses were executed by entering model variables into three sequential blocks. Youth age, sex, race, body mass index, and season of saliva collection were entered into the first block. Caregiver age, marital status, and household income were entered in a second block. Youth health diagnoses, medication status, cigarette use, secondhand exposure to tobacco smoke, and pubertal development were entered in a third block. Likelihood ratio tests were used to assess and compare model fit (Meng & Rubin, 1992).

3 |. RESULTS

Demographic data for adolescents and households, along with general health and seasonality, are presented in Table 1. Regression analyses with multiple imputation were conducted with the full sample of 899 youth. The initial model with adolescent demographics was significant (F (5, 892) = 16.67, p < .001) and accounted for 9.71% of the variance in the T/S ratio. As shown in Table 2, T/S ratio was significantly associated with age, race, and the season during which saliva was collected. Specifically, T/S ratio was negatively associated with age, indicating that older age was associated with shorter telomeres. T/S ratio was larger among Black youth than among White youth. In addition, T/S ratio was larger when measured from saliva samples collected in the fall/winter months compared to those collected in the spring/summer months. For the second block of caregiver demographics, the model was significant (𝒳2(6) = 2.93, p = .007, ΔR2 = 0.018) and accounted for an additional 1.80% of the variance in T/S ratio. Only family income was significantly associated with T/S ratio in adolescence, such that T/S ratio was significantly larger for youth whose annual household income was above $30,000 compared to families with household income below $30,000. Among the caregiver demographic predictors, T/S ratio was not associated with caregiver age or marital status. For the health predictors entered on Block 3, the model was not significant (𝒳2(5) = 0.67, p = .645, ΔR2 = 0.004). Exposure to tobacco smoke in the household, having one or more medical health diagnoses, or using one or more prescription medications were not associated with T/S ratio. Medical diagnoses and prescription medication use were reanalyzed using continuous rather than dichotomized measures, which did not affect the pattern of results observed.

TABLE 1.

Demographic information

Variable All participants Black White
N 899 411 488
Age 14.39 (1.87) 14.27 (1.85) 14.53 (1.88)
Male 465 (52%) 189 (46%) 223 (46%)
Body mass index 23.96 (7.15) 24.99 (7.71) 23.22 (6.63)
Season of data collection
Fall/winter 434 (48%) 207 (50%) 227 (47%)
Spring/summer 465 (52%) 204 (50%) 261 (53%)
Caregiver age 45.76 (8.31) 43.77 (9.65) 47.43 (6.54)
Caregiver annual household income
<$30,000 382 (34%) 197 (52%) 85 (18%)
$30,000-$60,000 209 (25%) 125 (33%) 84 (18%)
>$60,000 384 (41%) 55 (15%) 293 (63%)
Relationship status
Married 496 (56%) 121 (31%) 375 (77%)
Cohabitating 87 (10%) 54 (14%) 33 (7%)
Single 164 (19%) 152 (38%) 12 (2%)
Separated/divorced/widowed 134 (15%) 68 (17%) 66 (14%)
Pubertal development 3.02 (0.65) 3.06 (0.66) 2.97 (0.62)
Medical diagnosis (1+) 525 (59%) 240 (60%) 285 (59%)
Prescription medication (1+) 360 (42%) 149 (39%) 211 (43%)
Smoke tobacco (previous 12 months) 35 (4%) 13 (3%) 22 (5%)
Household tobacco smoke exposure (ever) 206 (23%) 134 (33%) 72 (15%)
Open in a new tab

Note: Table displays the mean and standard deviation for continuous variables and frequencies and proportions for discrete variables. Missing data are excluded when calculating proportions.

TABLE 2.

Demographic and health variables predicting T/S ratio

Model 1 Model 2 Model 3
Variable β SE p β SE p β SE p
Age −0.023 0.016 .161 −0.034 0.017 .042 −0.043 0.021 .042
Race
Black Reference Reference Reference
White −0.488 0.060 < .001 −0.614 0.074 < .001 −0.614 0.074 < .001
Sex
Female Reference Reference Reference
Male 0.064 0.060 .281 0.066 0.059 .268 0.087 0.066 .192
Body mass index −0.003 0.005 .500 −0.001 0.005 .787 −0.001 0.005 .777
Season of saliva collection
Fall/winter Reference Reference Reference
Spring/summer −0.254 0.060 < .001 −0.260 0.060 < .001 −0.259 0.060 < .001
Caregiver age 0.007 0.004 .080 0.007 0.004 .094
Household income
<$30,000 Reference Reference
$30,000-$60,000 0.182 0.082 .027 0.176 0.083 .034
>$60,000 0.186 0.089 .037 0.178 0.093 .045
Caregiver marital status
Married Reference Reference
Single 0.061 0.115 .598 0.060 0.116 .604
Cohabitating −0.120 0.103 .244 −0.118 0.103 .254
Other (separated, divorced, widowed) −0.137 0.094 .144 −0.123 0.094 .191
Pubertal development 0.039 0.065 .549
Health diagnosis (yes) −0.112 0.069 .103
Prescription and steroid medication (yes) 0.018 0.068 .797
Tobacco use in previous year (yes) 0.070 0.153 .647
Household tobacco smoke exposure (yes) −0.005 0.077 .948
Open in a new tab

Note: Bold values indicate significance at p < .05.

Given the extant literature on racial disparities in telomere length and the observed race difference in T/S ratio in the full initial analyses, regression analyses were repeated stratified by race to examine within-race demographic variables associated with T/S ratio. The results of regression models stratified by youth race are shown in Table 3. Similar to the full sample, among Black adolescents (n = 411), significant effects were observed for Block 1 (F (4, 406) = 2.58, p = .035, R2 = 0.029) and Block 2 (𝒳2(6) = 2.29, p = 0.033, ΔR2 = 0.034) but not Block 3 (𝒳2(5) = 1.19, p = .311, ΔR2 = 0.014). Consistent with the full sample, older age was associated with smaller T/S ratios, T/S ratio was marginally larger for saliva collected during the fall/winter months than during the spring/summer months, and youth living in households with annual household income between $30,000 and $60,000 had larger T/S ratios than those with annual household income less than $30,000. Other demographic or health variables were not significantly associated with T/S ratio among Black youth.

TABLE 3.

Demographic and health variables predicting T/S ratio among black youth

Model 1 Model 2 Model 3
Variable β SE p β SE p β SE p
Age −0.036 0.022 .100 −0.053 0.022 .018 −0.074 0.026 .005
Sex
Female Reference Reference Reference
Male 0.031 0.081 .698 −0.009 0.081 .908 0.048 0.090 .589
Body mass index 0.001 0.007 .899 0.002 0.007 .729 0.002 0.007 .743
Season of saliva collection
Fall/winter Reference Reference Reference
Spring/summer −0.218 0.081 .007 −0.217 0.081 .007 −0.212 0.081 .009
Caregiver age 0.007 0.004 .110 0.007 0.004 .105
Household income
<$30,000 Reference Reference
$30,000-$60,000 0.201 0.093 .032 0.213 0.095 .026
>$60,000 0.212 0.138 .125 0.255 0.144 .076
Caregiver marital status
Married Reference Reference
Single 0.143 0.140 .308 0.167 0.141 .239
Cohabitating −0.056 0.112 .618 −0.021 0.115 .856
Other (separated, divorced, widowed) 0.054 0.130 .676 0.099 0.132 .452
Pubertal development 0.115 0.084 .171
Health diagnosis (yes) −0.159 0.094 .092
Prescription and steroid medication (yes) 0.100 0.096 .299
Tobacco use in previous year (yes) 0.043 0.228 .850
Household tobacco smoke exposure (yes) 0.102 0.090 .258
Open in a new tab

Note: Bold indicate values indicate significance at p < .05.

Among White youth (n = 488), the results showed significant effects for variables entered on Block 1 (F (4, 482) = 4.01, p = .003, R2 = 0.036) but not Block 2 (𝒳2(6) = 1.29, p = .258, ΔR2 = 0.017) or Block 3 (𝒳2(5) = 0.56, p = .734, ΔR2 = 0.006). As shown in Table 4, White youth from whom saliva was collected during the fall/winter months had larger T/S ratios than those from whom saliva was collected during the spring/summer months. Adolescents whose parents were separated, divorced, or widowed had significantly smaller T/S ratios than those whose caregivers were married. Post hoc analyses using Tukey’s HSD test comparing T/S ratios across all categories of caregiver marital status showed that the group difference was only significant between “separated/divorced/widowed” and “married” (p = .047, 95% CI [−0.654, −0.003]). Adolescents with caregivers who reported their marital status as single or cohabitating did not differ from adolescents with married parents or those whose parents were separated, divorced, or widowed. No other demographic or health variables were significantly associated with T/S ratio among White youth.

TABLE 4.

Demographic and health variables predicting T/S ratio among White youth

Model 1 Model 2 Model 3
Variable β SE p β SE p β SE p
Age −0.007 0.023 .760 −0.012 0.025 .628 −0.005 0.034 .885
Sex
Female Reference Reference Reference
Male 0.169 0.088 .056 0.159 0.088 .073 0.142 0.100 .155
Body mass index −0.011 0.008 .163 −0.007 0.008 .379 −0.005 0.009 .554
Season of saliva collection
Fall/winter Reference Reference Reference
Spring/summer −0.278 0.087 .001 −0.283 0.087 .001 −0.278 0.087 .002
Caregiver age 0.003 0.007 .700 0.002 0.008 .806
Household income
<$30,000 Reference Reference
$30,000-$60,000 0.145 0.148 .331 0.123 0.150 .414
>$60,000 0.151 0.129 .243 0.119 0.135 .379
Caregiver marital status
Married Reference Reference
Single −0.013 0.193 .945 0.030 0.198 .882
Cohabitating −0.140 0.301 .642 −0.108 0.303 .720
Other (separated, divorced, widowed) −0.278 0.137 .043 −0.274 0.138 .048
Pubertal development 0.043 0.101 .672
Health diagnosis (yes) −0.034 0.098 .724
Prescription and steroid medication (yes) −0.036 0.098 .713
Tobacco use in previous year (yes) 0.126 0.208 .545
Household tobacco smoke exposure (yes) −0.143 0.136 .296
Open in a new tab

Note: Bold indicate values indicate significance at p < .05.

4 |. DISCUSSION

This study addressed current gaps in knowledge about demographic and health-related factors associated with TL in adolescents, the least commonly studied age group with TL to date. Key findings from this study are discussed in terms of their implications for a basic understanding of demographics and TL during adolescence, methodological considerations in terms of important factors when conducting research on TL in development, and health disparities research.

Consistent with studies in other age developmental phases of life (Bosquet Enlow et al., 2020; Chen et al., 2011), our results documented that the T/S ratio declines with age, even within the limited range of 11–17 years of age. On average, older adolescents had smaller T/S ratios than younger adolescents. This association was stable even after including other demographic and health indicators. This provides further support for the link between TL and chronological age and, by extension, cellular senescence. This is noteworthy given the dearth of information on TL during adolescence relative to other developmental periods.

Another key contribution of this study was that, on average, Black adolescents had larger T/S ratios than White participants. In the United States, Black newborns and children have longer telomeres than their White counterparts (Drury et al., 2015; Needham et al., 2017; Selvaraju et al., 2021). However, telomere shortening is accelerated among Black Americans (Diez Roux et al., 2009; Hunt et al., 2008). TL, although associated with chronological age, is more accurately described as a marker of biological age and is a purported mechanism of racial health disparities (Clendinen & Kertes, 2022). In analyses of T/S ratio stratified by race, we observed that the association of age with T/S ratio was robust for Black youth. Our findings are consistent with purported accelerated cellular aging among Black Americans relative to White Americans. The more pronounced age effect among Black youth suggests a role for environmental stressors (Roux et al., 2009). A separate study with children showed Black/White race differences in leukocyte TL that were more pronounced at lower socioeconomic status levels. (Needham et al., 2020). Among Black adults, racial discrimination is associated with accelerated leukocyte TL shortening over time (Rentscher et al., 2020). In our study, the stronger age effect among Black youth may be a byproduct, at least in part, of accumulated chronic stressors that disproportionately affect Black Americans and point to potentially modifiable aspects of the social environment that may be important to consider in preventing the onset of TL-linked racial health disparities.

One potential mechanism is the downstream effects of chronic activation of the stress-sensitive hypothalamic-pituitary-adrenocortical system, which shifts cellular resources toward responding to immediate threats via glucose release and away from mechanisms of cellular growth and repair, including reduced telomerase activity (Choi et al., 2008). Stress processes involving the sympathetic-adrenomedullary system simultaneously trigger downstream effects on the release of proinflammatory cytokines and other components of the immune system that impact oxidative stress and DNA damage, to which DNA telomeric ends are particularly vulnerable. With TL and accelerated TL shortening linked to a host of diseases and disorders with observed health disparities, our results are consistent with the notion of heightened biological wear and tear among Black youth that may increase the risk for morbidities in adulthood impacted by cellular aging.

Salivary T/S ratio was also associated with the season during which saliva samples were collected. Saliva samples collected in the fall to winter months (September through February) had larger T/S ratios than those collected in the spring to summer (March through August). To the best of our knowledge, this is the first study to document a season effect in salivary samples. Our findings are consistent with seasonality effects previously documented in blood from human and nonhuman primates living in tropical environments (Beaulieu et al., 2017; Rehkopf et al., 2014; Tennyson et al., 2018). One possible explanation for seasonal TL variation is cell type composition, which mirrors TL variation observed in peripheral blood leukocytes and microbial conditions linked to seasonal precipitation (Beaulieu et al., 2017). These findings point to the need to control for season of tissue collection in developmental research involving telomere length and with salivary TL assessment more broadly, even among populations located outside of tropical environments. Otherwise, differential timing of participant recruitment and sampling may unintentionally introduce error or bias results in either cross-sectional or longitudinal data collection.

T/S ratio was also associated with household income. Adolescents residing in households with an annual income below $30,000 had smaller T/S ratios than adolescents residing in households with higher income. These findings are consistent with other reports of TL and indicators of socioeconomic status during early life (Bosquet Enlow et al., 2018; Martens et al., 2020; Needham et al., 2012). The mechanisms by which household income adversely affects TL are not well understood. It has been suggested that associations between economic adversity and TL may be explained by intermediate factors such as tobacco use or obesity (Cohen et al., 2013; Hiscock et al., 2012). In the present study, the T/S ratio was not associated with adolescent tobacco use, household tobacco exposure, body mass index, diagnosed medical conditions, or medication use. These findings point to the need to comprehensively assess aspects of the social environment that underlie or moderate socioeconomic associations with TL during development. As low household income is often stable during early life and typically co-occurs with other life stressors, these findings also speak to the need for a more comprehensive assessment of the role of early life adversity on TL during development.

Race stratified analyses of the demographic and health factors associated with T/S ratio identified within-race associations of household demographics with T/S ratio. Among Black youth, a smaller T/S ratio was associated with older age, household income below $30,000, and season of saliva collection. These findings were identical to those from the analyses collapsing across race. Some differences were observed for analyses conducted with White youth only compared to those conducted among Black youth only or to the analyses collapsed across race. Among White youth, the season of collection was associated with T/S ratio, but as indicated above, age was not. Household demographics associated with T/S ratio also differed for White youth. Parent marital status, but not household income, was significantly associated with T/S ratio, such that White youth with separated, divorced, or widowed parents had smaller T/S ratios than those whose caregivers were married. This difference may be related to racial differences in household income and marital status, which were observed in our community-based sample and are observed in the broader U.S. population (U.S. Census Bureau, 2021). Alternatively, discrepancies in the race-stratified analyses may reflect unmeasured differences in the psychosocial effects of financial and family composition by racial group. These results indicate a concern for generalizability across demographic groups in TL research, particularly given the historical overrepresentation of White, middle-class families in developmental research.

It is also worth noting the demographic and health indicators that were not associated with TL among adolescents. We did not observe any meaningful differences between males and females, by pubertal status (controlling for age), caregiver age, presence/absence or number of diagnosed health conditions, and adolescent use or number of medications taken, either when analyzed collapsed across race or within race. Moreover, no associations were observed for either adolescent tobacco use or household exposure to tobacco smoke. The potential link between tobacco exposure and TL in adulthood is not yet clear. One meta-analysis of ever vs. never smokers and number of pack-years in adults suggests an association with tobacco exposure (Astuti et al., 2017); however, in a different meta-analysis focused on longitudinal designs, there was no association of smoking with telomere attrition (Bateson et al., 2019). The large sample in the current analysis lends confidence that these demographic and health indicators do not have a major effect on TL during adolescence and are unlikely to be confounding factors in research using TL measures with adolescents.

The findings from this study should be interpreted considering the study strengths and limitations. Methodological rigor was maximized for adolescent- and caregiver-report measures by conducting all assessments in the participant’s home to facilitate completion and comprehension. The rigor of the telomere assays, quality control checks, and analytic pipeline of T/S ratio data was maximized by adhering to the best reporting practices as recommended by the NIH Telomere Research Network (Lindrose & Drury, 2020). The large sample allowed for within-race analyses. This includes within-race examinations among Black Americans, who show accelerated telomere shortening compared to White Americans, have been historically marginalized in behavioral and biomedical science research, and who carry a disproportionate burden of disease (Hunt et al., 2008). Potential limitations of the study are that due to the community-based sample unselected for a particular disease or disorder, clinical assessments of health and disease were not conducted. As racial disparities in health care access and appropriate diagnoses persist in the United States, it is possible that other health indicators not measured in this study may relate to TL during adolescence.

In sum, the results from this study yield methodological and conceptual contributions to an understanding of TL during adolescence. Similar to other ages, telomere length is shorter with advancing age, particularly among Black youth. From a methodological perspective, this speaks to the need to consider both age and race in TL research. Conceptually, the results also suggest that adolescence may mark an important developmental phase in accelerated telomere shortening among Black Americans, potentially contributing to racial health disparities in adulthood. Our results also provide evidence that season of data collection is an essential covariate in TL research when studying developmental populations and/or using saliva as a tissue source to avoid potential confounds. Finally, household-level demographics may be important considerations in studies of TL research, especially those related to indices of socioeconomic status, calling attention to the dearth of information about the mechanisms by which demographics are related to cellular aging. Together, this foundational knowledge offers a basis for more robust examinations of the social environment on telomere length and telomere attrition during the physically, cognitively, and socially dynamic period of adolescence.

ACKNOWLEDGMENTS

We gratefully acknowledge the families who participated in the study, without whom this work would not be possible. Funding sources: R01MD011727, R01NR019008, R01DA032371, R21DA034960, P2CHD058484, William T. Grant Foundation.

Footnotes

CONFLICT OF INTEREST

The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available upon reasonable request.

REFERENCES

  1. Astuti Y, Wardhana A, Watkins J, & Wulaningsih W (2017). Cigarette smoking and telomere length: A systematic review of 84 studies and meta-analysis. Environmental Research, 158, 480–489. 10.1016/j.envres.2017.06.038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bateson M, Aviv A, Bendix L, Benetos A, Ben-Shlomo Y, Bojesen SE, Cooper C, Cooper R, Deary IJ, Hägg S, Harris SE, Kark JD, Kronenberg F, Kuh D, Labat C, Martin-Ruiz CM, Meyer C, Nordestgaard BG, Pennix BWJH, & Nettle D (2019). Smoking does not accelerate leucocyte telomere attrition: A meta-analysis of 18 longitudinal cohorts. Royal Society open science, 6(6), 190420. 10.1098/rsos.190420 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Beaulieu M, Benoit L, Abaga S, Kappeler PM, & Charpentier MJE (2017). Mind the cell: Seasonal variation in telomere length mirrors changes in leucocyte profile. Molecular Ecology, 26(20), 5603–5613. 10.1111/mec.14329 [DOI] [PubMed] [Google Scholar]
  4. Bosquet Enlow M, Bollati V, Sideridis G, Flom J, Hoxha M, Hacker M, & Wright R (2018). Sex differences in effects of maternal risk and protective factors in childhood and pregnancy on newborn telomere length. Psychoneuroendocrinology, 95, 74–85. 10.1016/j.psyneuen.2018.05.025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bosquet Enlow M, Kane-Grade F, De Vivo I, Petty C, & Nelson C (2020). Patterns of change in telomere length over the first three years of life in healthy children. Psychoneuroendocrinology, 115, 104602. 10.1016/j.psyneuen.2020.104602 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Caballero A, Granberg R, & Tseng KY (2016). Mechanisms contributing to prefrontal maturation during adolescence. Neuroscience & Biobehavioral Reviews, 70, 4–12. 10.1016/j.neubiorev.2016.05.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Centers for Disease Control and Prevention (2012). 2011 National survey of children’s health questionnaire. https://www.cdc.gov/nchs/data/slaits/2011NSCHQuestionnaire.pdf
  8. Chen W, Kimura M, Kim S, Cao X, Srinivasan SR, Bernenson GS, Kark JD, & Aviv A (2011). Longitudinal versus cross-sectional evaluations of leukocyte telomere length dynamics: Age-dependent telomere shortening is the rule. The Journals of Gerontology: Series A, 66(3), 312–319. 10.1093/gerona/glq223 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Choi J, Fauce SR, & Effros RB (2008). Reduced telomerase activity in human T lymphocytes exposed to cortisol. Brain, Behavior, and Immunity, 22(4), 600–605. 10.1016/j.bbi.2007.12.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Clendinen C, & Kertes DA (2022). Discrimination and heath: Fostering better health for Black American youth. Policy Insights from the Behavioral and Brain Sciences, 9, L3–L10. 10.1177/23727322211073796 [DOI] [Google Scholar]
  11. Cohen AK, Rai M, Rehkopf DH, & Abrams B (2013). Educational attainment and obesity: A systematic review. Obesity Review, 14(12), 989–1005. 10.1111/obr.12062 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Coimbra BM, Carvalho CM, Moretti PN, Mello MF, & Belangero SI (2017). Stress-related telomere length in children: A systematic review. Journal of Psychiatric Research, 92, 47–54. 10.1016/j.jpsychires.2017.03.023 [DOI] [PubMed] [Google Scholar]
  13. Colich NL, Rosen ML, Williams ES, & McLaughlin KA (2020). Biological aging in childhood and adolescence following experiences of threat and deprivation: A systematic review and meta-analysis. Psychological Bulletin, 146(9), 721–764. 10.1037/bul0000270 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. de Fluiter KS, Codd V, Denniff M, Kerkhof GF, van Beijsterveldt IALP, Breij LM, Samani NJ, Abrahamse-Berkeveld M, & Hokken-Koelega ACS (2021). Longitudinal telomere length and body composition in health term-born infants during the first two years of life. PLoS One, 16(2), e024600. 10.1371/journal.pone.0246400 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Diez Roux AV, Ranjit N, Jenny NS, Shea S, Chushman M, Fitzpatrick A, & Seeman T (2009). Race/ethnicity and telomere length in the Multi-ethnic Study of Atherosclerosis. Aging Cell, 8(3), 251–257. 10.1111/j.1474-9726.2009.00470.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Drury SS, Esteves K, Hatch V, Woodbury M, Borne S, Adamski A, & Theall KP (2015). Setting the trajectory: Racial disparities in newborn telomere length. The Journal of Pediatrics, 166(5), 1181–1186. 10.1016/j.jpeds.2015.01.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Eisenberg DTA, Kuzawa CW, & Hayes MG (2015). Improving qPCR telomere length assays: Controlling for well position effects increases statistical power. American Journal of Human Biology, 27(4), 570–575. 10.1002/ajhb.22690 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Garay SM, Savory KA, Sumption LA, Penketh RJA, Jones IR, Janssen AB, & John RM (2019). Seasonal variation in salivary cortisol but not symptoms of depression and trait anxiety in pregnant women undergoing an elective caesarean section. Psychoneuroendocrinology, 108, 14–19. 10.1016/j.psyneuen.2019.05.029 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Grubbs FE (1969). Procedures for detecting outlying observations in samples. Technometrics, 11(1), 1–21. 10.1080/00401706.1969.10490657 [DOI] [Google Scholar]
  20. Hellemans J, & Vandesompele J (2011). qPCR data analysis-unlocking the secret to successful results. PCR troubleshooting and optimization: The essential guide, 1, 13. [Google Scholar]
  21. Hiscock R, Bauld L, Amos A, Fidler JA, & Munafo M (2012). Socioeconomic status and smoking: A review. Annals of the New York Academy of Sciences, 1248(1), 107–123. 10.1111/j.1749-6632.2011.06202.x [DOI] [PubMed] [Google Scholar]
  22. Hunt SC, Chen W, Gardner JP, Kimura M, Srinivasan SR, Eckfeldt JH, Berenson GS, & Aviv A (2008). Leukocyte telomeres are longer in African Americans than in whites: The National Heart, Lung, and Blood Institute Family Heart Study and the Bogalusa Heart Study. Aging Cell, 7(4), 451–458. 10.1111/j.1474-9726.2008.00397.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Jackson S, & Goossens L (Eds.). (2020). Handbook of adolescent development. Psychology Press. [Google Scholar]
  24. Lindrose A, & Drury S (2020) Minimum reporting recommendations for PCR-based telomere length measurement. 10.31219/osf.io/9pzst [DOI] [Google Scholar]
  25. Little RJA (1988). Missing data adjustments in large surveys. Journal of Business Economics and Statistics, 6(3), 287–296. 10.1080/07350015.1988.10509663 [DOI] [Google Scholar]
  26. Marchetto NM, Glynn RA, Ferry ML, Ostojic M, Wolff SM, Yao R, & Haussmann MF (2016). Prenatal stress and newborn telomere length. American Journal of Obstetrics and Gynecology, 215(1), 94.e1–94.e8. 10.1016/j.ajog.2016.01.177 [DOI] [PubMed] [Google Scholar]
  27. Martens DS, Janssen BG, Bijnens EM, Clemente DBP, Vineis P, Plusquin M, & Nawrot TS (2020). Association of parental socioeconomic status and newborn telomere length. JAMA Network Open, 3(5), e204057. 10.1001/jamanetworkopen.2020.4057 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Mathur MB, Epel E, Kind S, Desai M, Parks CG, Sandler DP, & Khazeni N (2016). Perceived stress and telomere length: A systematic review, meta-analysis, and methodologic considerations for advancing the field. Brain, Behavior, and Immunity, 54, 158–169. 10.1016/j.bbi.2016.02.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Meng XL, & Rubin DB (1992). Performing likelihood ratio tests with multiply-imputed data sets. Biometrika, 79(1), 103–111. 10.1093/biomet/79.1.103 [DOI] [Google Scholar]
  30. Needham BL, Fernandez JR, Lin J, Epel ES, & Blackburn EH (2012). Socioeconomic status and cell aging in children. Social Science & Medicine, 74(12), 1948–1951. 10.1016/j.socscimed.2012.02.019 [DOI] [PubMed] [Google Scholar]
  31. Needham BL, Hicken MT, Govia IO, Mitchell C, & Abdou CM (2017). Maternal social disadvantage and newborn telomere length in archived dried blood spots from the Michigan Neonatal Biobank. Biodemography and Social Biology, 63(3), 221–235. 10.1080/19485565.2017.1300520 [DOI] [PubMed] [Google Scholar]
  32. Needham BL, Salerni S, Roberts E, Boss J, Allgood KL, & Mukherjee B (2020). Do black/white differences in telomere length depend on socioeconomic status? Biodemography and Social Biology, 64(4), 287–312. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Pfaffl MW (2001). A new mathematical model for relative quantification in real-time RT–PCR. Nucleic Acids Research, 29(9), e45–e45. 10.1093/nar/29.9e45 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Rapee RM, Oar EL, Johnco CJ, Forbes MK, Fardouly J, Magson NR, & Richardson CE (2019). Adolescent development and risk for the onset of social-emotional disorders: A review and conceptual model. Behaviour Research and Therapy, 123, 103501. 10.1016/j.brat.2019.103501 [DOI] [PubMed] [Google Scholar]
  35. Rehkopf DH, Dow WH, Rosero-Bixby L, Lin J, Epel ES, & Blackburn EH (2014). Seasonal variation of peripheral blood leukocyte telomere length in Costa Rica: A population-based observational study. American Journal of Human Biology, 26(3), 367–375. 10.1002/ajhb.22529 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Rentscher KE, Carroll JE, & Mitchell C (2020). Psychosocial stressors and telomere length: A current review of the science. Annual Review of Public Health, 41, 223–245. 10.1146/annurev-publhealth-040119-094239 [DOI] [PubMed] [Google Scholar]
  37. Roux AVD, Ranjit N, Swords Jenny N, Shea S, Cushman M, Fitxpatrick A, & Seeman T (2009). Race/ethnicity and telomere length in the Multi-Ethnic Study of Atherosclerosis. Aging Cell, 8(3), 251–257. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Rubin DB (1987). Multiple imputation for nonresponse in surveys. John Wiley & Sons. [Google Scholar]
  39. Selvaraju V, Phillips M, Fouty A, Babu JR, & Geetha T (2021). Telomere length as a biomarker for race-related health disparities. Genes, 12(1), 78. 10.3390/genes12010078 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Shalev I, Moffitt TE, Sugden K, Williams B, Houts RM, Danese A, Mill J, Arseneault L, & Caspi A (2013). Exposure to violence during childhood is associated with telomere erosion from 5 to 10 years of age: A longitudinal study. Molecular Psychiatry, 18(5), 576–581. 10.1038/mp.2012.32 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Smith L, Luchini C, Demurtas J, Soysal P, Stubbs B, Hamer M, Nottegar A, Lawlor RT, Lopez-Sanchez GF, Firth J, Koyanagi A, Roberts J, Willeit P, Waldhoer T, Loosemore M, Abbs AD, Johnstone J, Yang L, & Veronese N (2019). Telomere length and health outcomes: An umbrella review of systematic reviews and meta-analyses of observational studies. Ageing Research Reviews, 51, 1–10. 10.1016/j.arr.2019.02.003 [DOI] [PubMed] [Google Scholar]
  42. Srinivas N, Rachakonda S, & Kumar R (2020). Telomeres and telomere length: A general overview. Cancers, 12(3), 558. 10.3390/cancers12030558 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Tennyson RL, Gettler LT, Kuzawa CW, Haynes MG, Agustin SS, & Eisenberg DTA (2018). Lifetime socioeconomic status and early life microbial environments predict adult blood telomere length in the Philippines. American Journal of Human Biology, 30(5), e23145. 10.1002/ajhb.23145 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. U.S. Census Bureau (2021). Current Population Survey, 2021 Annual Social and Economic (ASEC) Supplement. https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar21.pdf
  45. van Buuren S, & Groothuis-Oudshoorn K (2011). MICE: Multivariate imputation by chained equations in R. Journal of Statistical Software, 45(3), 1–67. 10.18637/jss.v045.i03 [DOI] [Google Scholar]
  46. Venables WN, & Ripley BD (2002). Modern applied statistics with S (4th ed.). Springer. 10.1007/978-0-387-21706-2 [DOI] [Google Scholar]
  47. Verner G, Epel E, Lahti-Pulkkinen M, Kajantie E, Buss C, Lin J, Blackburn E, Räikkönen K, Wadhwa P, & Entringer S (2020). Maternal psychological resilience during pregnancy and newborn telomere length: A prospective study. The American Journal of Psychiatry, 178(2), 183–192. 10.1176/appi.ajp.2020.19101003 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Data Availability Statement

The data that support the findings of this study are available upon reasonable request.

RESOURCES