Table B1. Parameter Estimates for Dual-Process Model of Puberty and Sleep (N = 1,239)
Table B1. Parameter estimates for dual-process model of puberty and sleep (n = 1,239)
Wake Times / BedtimesBreast
Development / Pubic Hair Development / Breast
Development / Pubic Hair Development
Sleepa
Sleep intercept / 7.70* (.02) / 7.70* (.02) / 22.12* (.02) / 22.12* (.02)
Sleep slope / .06* (.01) / .06* (.01) / .20* (.01) / .20* (.01)
Pubertyb
Pubertal timing / 11.57* (.03) / 11.59* (.03) / 11.57* (.03) / 11.59* (.03)
Pubertal rate / .80* (.01) / .86* (.01) / .80* (.01) / .86* (.01)
Associations
Pubertal timing with sleep intercept / -.04 (.02) / -.02 (.03) / -.09* (.03) / -.08* (.03)
Pubertal timing with sleep slope / .02† (.01) / .02† (.01) / .01 (.01) / .00 (.01)
Pubertal timing with pubertal rateb / -.05* (.01) / -.02 (.02) / -.05* (.01) / -.02 (.02)
Pubertal rate with sleep intercept / .00 (.01) / -.01 (.01) / .00 (.01) / -.01 (.01)
Pubertal rate with sleep slope / -.00 (.00) / -.00 (.00) / -.00 (.00) / -.00 (.00)
Sleep intercept with sleep slopea / .02* (.00) / .02* (.00) / -.00 (.01) / -.00 (.01)
On Sleep Intercepta
Black / .03 (.06) / .03 (.06) / .34* (.06) / .34* (.06)
Hispanic / .19* (.07) / .19* (.07) / .29* (.07) / .30* (.07)
Asian / .26* (.10) / .26* (.10) / .40* (.11) / .40* (.11)
BMI / -.02 (.02) / -.02 (.02) / .03† (.02) / .03† (.02)
Family SES / .01 (.01) / .01 (.01) / -.02 (.01) / -.02 (.01)
MSSM (site) / .04 (.07) / .04 (.07) / .01 (.07) / .01 (.07)
KPNC (site) / -.08 (.05) / -.08 (.05) / -.23* (.06) / -.24* (.06)
On Sleep Slopea
Black / .01 (.02) / .01 (.02) / -.04† (.02) / -.04† (.02)
Hispanic / .01 (.02) / .01 (.02) / -.03 (.02) / -.03 (.02)
Asian / -.01 (.03) / -.01 (.03) / -.03 (.03) / -.03 (.03)
BMI / .01 (.01) / .01 (.01) / -.00 (.01) / -.01 (.01)
Family SES / .00 (.01) / .00 (.01) / .00 (.02) / .00 (.02)
MSSM (site) / -.00 (.02) / -.00 (.02) / -.03 (.02) / -.03 (.02)
KPNC (site) / .02 (.02) / .02 (.02) / -.01 (.02) / -.01 (.02)
On Pubertal Timingb
Black / -.90* (.10) / -1.18* (.10) / -.90* (.10) / -1.18* (.12)
Hispanic / -.49* (.11) / -.40* (.13) / -.49* (.11) / -.40* (.12)
Asian / -.20 (.17) / .62* (.18) / -.20 (.16) / .62* (.18)
BMI / -.33* (.03) / -.33* (.03) / -.33* (.03) / -.33* (.03)
Family SES / -.05 (.10) / -.08 (.02) / -.05 (.02) / -.08 (.02)
MSSM (site) / .28* (.09) / .02 (.12) / .28* (.11) / .02 (.12)
KPNC (site) / .21* (.12) / .01 (.10) / .21* (.09) / .01 (.10)
On Pubertal Rateb
Black / .02 (.03) / -.10* (.04) / .02 (.03) / -.10* (.04)
Hispanic / .00 (.03) / -.02 (.04) / .00 (.03) / -.02 (.04)
Asian / -.07 (.05) / -.18* (.06) / -.07 (.05) / -.18* (.06)
BMI / -.01 (.01) / .03* (.01) / -.01 (.01) / .03* (.01)
Family SES / .01 (.01) / -.00 (.01) / .01 (.01) / -.00 (.01)
MSSM (site) / -.01 (.03) / -.05 (.04) / -.01 (.03) / -.05 (.04)
KPNC (site) / .16 (.03) / -.01 (.03) / .16 (.03) / -.01 (.03)
Model Fit
Freeparameters / 49 / 49 / 49 / 49
-2 Res Log Likelihood / -18461.34 / -18378.74 / -18069.54 / -17988.79
AIC / 41763.43 / 41598.23 / 40979.85 / 40818.33
Note. * p < .01; †p < .05. Analyses were centered at Year 4, but all estimates accounted for individuals’ exact ages at each year. Race/ethnic variables, Black, Hispanic, and Asian are dummy codes comparing against White participants. Covariates include body mass index (BMI) and family socioeconomic status (SES), measured as caregiver years of education. Site variables, MSSM (New York) and KPNC (California), are dummy codes comparing against Cincinnati participants. Unstandardized parameter estimates and standard errors are provided. Wake time and bedtime intercept is presented in hours (e.g., military time with fractions of the hour rather than minutes), and slope is the rate of change in sleep in hours/year. Timing is presented in years of age at sexual maturation rating (SMR) 3, and rate is the change in SMR stages/year at SMR 3. aThe wake time/ bedtime intercept and slope estimatesfor pubic hair, and the two for breast development timing, were estimated from the same dataand are therefore nearly identical (within rounding error) across models. bSimilarly, the timing and rate estimates, associations, and covariate effects are nearly identical (within rounding error) because the same data were included for models including wake times and bedtimes.
Table B2. Multi-group analysis of racial/ethnic differences in associations of pubertal development and wake time
Model 1. Breast Development / Model 2. Pubic Hair DevelopmentBlack
(n = 391) / Hispanic
(n = 372) / White
(n = 418) / Black
(n = 391) / Hispanic
(n = 372) / White
(n = 418)
Wake Timea
Intercept / 7.66* (.04) / 7.84* (.05) / 7.57* (.03) / 7.66* (.04) / 7.84* (.05) / 7.57* (.03)
Slope / .07* (.01) / .06* (.01) / .07* (.01) / .07* (.01) / .06* (.01) / .06* (.01)
Pubertal Developmentb
Timing / 11.12* (.06) / 11.64* (.06) / 11.92* (.05) / 1.90* (.06) / 11.76* (.07) / 11.97* (.05)
Rate / .82* (.02) / .80* (.02) / .83* (.01) / .84* (.02) / .91* (.02) / .96* (.02)
Associations
Timing with intercept / -.01 (.05) / .02 (.05) / -.04 (.03) / .01 (.05) / -.06 (.05) / .02 (.03)
Timing with slope / .03 (.02) / -.07 (.02) / -.01 (.01) / .02† (.02) / .04† (.02) / -.00 (.01)
Timing with rateb / -.06* (.02) / -.04 (.03) / .02 (.01) / -.07† (.03) / .02 (.04) / .03 (.03)
Rate with intercept / .01 (.01) / -.01 (.02) / -.03 (.02) / .00 (.02) / -.03 (.02) / -.01 (.01)
Rate with slope / -.00 (.01) / -.00 (.01) / .01 (.01) / -.01 (.01) / .01 (.01) / -.01 (.01)
Intercept with slopea / .02† (.01) / .03 (.01) / -.01 (.00) / .02† (.01) / .03† (.01) / .02 (.01)
On Wake Time Intercepta
BMI / -.00 (.04) / -.11* (.04) / .04 (.03) / -.00 (.04) / -.11* (.04) / .04 (.03)
Family SES / .00 (.02) / -.03 (.02) / -.02 (.02) / .00 (.02) / -.03 (.02) / -.01 (.02)
On Wake Time Slopea
BMI / -.00 (.01) / .01 (.01) / .02 (.01) / -.00 (.01) / .01 (.01) / .02 (.01)
Family SES / .01 (.01) / -.00 (.01) / .01 (.01) / .01 (.01) / -.00 (.01) / .01 (.01)
On Pubertal Development Timingb
BMI / -.36* (.05) / -.28* (.06) / -.34* (.05) / -.35* (.06) / -.28* (.06) / -.34* (.06)
Family SES / -.11* (.03) / -.04 (.02) / -.03 (.03) / -.10* (.04) / -.08* (.03) / -.04 (.03)
On Pubertal Development Rateb
BMI / -.00 (.02) / .01 (.02) / -.01 (.02) / .05† (.02) / .04 (.02) / .02 (.03)
Family SES / .02† (.01) / .02* (.01) / -.03* (.01) / .03† (.01) / .01 (.01) / -.03 (.01)
Model Fit
-2 Res Log Likelihood / -17136.30 / -16953.19
AIC / 34446.60 / 3408.38
Note.* p < .01; †p < .05.Model Fit: Free parameters = 87 in each model. Covariates include body mass index (BMI) and family socioeconomic status (SES), measured as caregiver years of education. Analyses were centered at Year 4, but all estimates accounted for individuals’ exact ages at each year.aThe wake time/ bedtimeintercept and slope are estimated from the same data for models including breast and pubic hair development timing, and thus are nearly identical (within rounding error) across models. bSimilarly, the timing and rate estimates, associations, and covariate effects are nearly identical (within rounding error) because the same data were included for models including wake times and bedtimes.
Table B3. Multi-group analysis of racial/ethnic differences in associations of pubertal development and bedtime
Model 1. Breast Development / Model 2. Pubic Hair DevelopmentBlack
(n = 391) / Hispanic
(n = 372) / White
(n = 418) / Black
(n = 391) / Hispanic
(n = 372) / White
(n = 418)
Bed Timea
Intercept / 22.26* (.04) / 22.26* (.05) / 21.84* (.03) / 22.26* (.04) / 22.26* (.05) / 21.84* (.03)
Slope / .18* (.01) / .18* (.01) / .23* (.01) / .18* (.01) / .18* (.01) / .23* (.01)
Pubertal Developmentb
Timing / 11.12* (.06) / 11.64* (.06) / 11.92* (.05) / 1.90* (.06) / 11.76* (.07) / 11.97* (.05)
Rate / .82* (.02) / .79* (.02) / .83* (.01) / .84* (.02) / .91* (.02) / .97* (.02)
Associations
Timing with intercept / -.14* (.05) / -.08 (.06) / -.06† (.03) / -.13† (.06) / -.10 (.06) / -.02 (.04)
Timing with slope / .01 (.01) / .01 (.02) / -.00 (.01) / -.00 (.02) / .01 (.02) / -.01 (.01)
Timing with rateb / -.07 (.02) / -.04 (.03) / -.03 (.02) / -.06† (.03) / .02 (.04) / .04 (.03)
Rate with intercept / .01 (.02) / -.02 (.02) / -.01 (.01) / .01 (.02) / -.03 (.02) / -.00 (.01)
Rate with slope / -.01 (.00) / .00 (.01) / -.01 (.00) / -.01 (.01) / .01† (.01) / -.00 (.00)
Intercept with slopea / -.00 (.01) / -.01 (.01) / .00 (.01) / .00 (.01) / -.01 (.01) / .00 (.01)
On Bedtime Intercepta
BMI / .03 (.04) / -.03 (.04) / .08* (.03) / .03 (.04) / -.03 (.04) / .08* (.03)
Family SES / -.02 (.02) / -.05 (.02) / -.01 (.02) / -.02 (.02) / -.05 (.02) / -.01 (.02)
On Bedtime Slopea
BMI / -.01 (.01) / -.00 (.01) / -.01 (.01) / -.01 (.01) / -.00 (.01) / -.01 (.01)
Family SES / .01 (.00) / .00 (.01) / .00 (.01) / .01 (.00) / .00 (.01) / .00 (.01)
On Pubertal Development Timingb
BMI / -.36* (.05) / -.29* (.06) / -.34* (.05) / -.36* (.06) / -.29* (.06) / -.34* (.05)
Family SES / -.10* (.03) / -.04 (.03) / -.03 (.03) / -.10* (.04) / -.08* (.03) / -.04 (.03)
On Pubertal Development Rateb
BMI / .00 (.02) / .01 (.02) / -.02 (.02) / .05† (.02) / .01 (.02) / -.02 (.02)
Family SES / .02† (.01) / .02* (.01) / -.03* (.01) / .03† (.01) / .02* (.01) / -.03* (.01)
Model Fit
-2 Res Log Likelihood / -16735.32 / -16555.97
AIC / 33644.65 / 33285.93
Note.* p < .01; †p < .05.Model Fit: Free parameters = 87 in each model. Covariates include body mass index (BMI) and family socioeconomic status (SES), measured as caregiver years of education. Analyses were centered at Year 4, but all estimates accounted for individuals’ exact ages at each year.aThe wake time/ bedtime intercept and slope are estimated from the same data for models including breast and pubic hair development timing, and thus are nearly identical (within rounding error) across models. bSimilarly, the timing and rate estimates, associations, and covariate effects are nearly identical (within rounding error) because the same data were included for models including wake times and bedtimes.