Age and the Social Stratification of Long-Term Trajectories of Physical Activity

1
Table 2. Americans Changing Lives, 1986-2002
Physical activity over time
Age and the Social Stratification of Long-Term
Trajectories of Physical Activity
Benjamin A. Shaw1,2 Jersey Liang3, Neal Krause3,
Mary Gallant1,2, and Kelly McGeever2 1University
at Albany, School of Public Health 2 University
at Albany, Center for Social and Demographic
Analysis 3University of Michigan, School of
Public Health
BACKGROUND
RESULTS

• Physical activity has well-documented health
  benefits and is considered one of the most
  effective measures for preventing and controlling
  chronic illnesses, enhancing psychological
  well-being, and preventing premature death.
• However, current data from the Behavioral Risk
  Factor Surveillance Study (BRFSS) indicate that
  approximately 50 of all adults do not meet
  recommended levels of regular physical activity
  (CDC 2007) these data also show substantial and
  persistent social stratification in rates of
  physical activity.
• Although these data are useful in charting our
  nations progress towards its public health
  goals, repeated cross-sectional assessments of
  physical activity are insufficient in that they
  reveal little about within-persons changes in
  physical activity over time.
• Assessing within-persons changes in physical
  activity over extended periods of time (i.e.,
  trajectories) can add to our understanding of the
  stratification of physical activity by revealing
  how, and why, various forms of physical activity
  stratification might change as adults at
  different points in the life course, and from
  different birth cohorts, grow older.

• Males, whites, highly educated, and young adults
  were more active, on average.
• Gender and education differences were larger, and
  race differences smaller, among older cohorts.
• On average, levels of physical activity decreased
  within individuals during the follow-up period
• However, significant age differences indicated
  that younger adults increased, while older adults
  decreased.
• Elevated levels of activity among whites
  diminished over time, especially among older
  adults (see figure).
• Gender differences widened over time among older
  adults (see figure).
• Gender-based differences did not remain after
  accounting for time-varying covariates however,
  race-based differences did remain.

Americans Changing Lives, 1986-2002 Physical
activity over time
Coefficient p Coefficient p
FIXED EFFECTS
For intercept (p0i)
Intercept 2.766 .000 1.992 .000
Baseline age -.100 .000 -.092 .000
Education .114 .000 .033 .011
Race (White) .054 .000 .043 .001
Gender (Male) .136 .000 .125 .000
AgeEducation .029 .031 .027 .028
AgeRace -.026 .036 -.018 .105
AgeGender .051 .000 .035 .001
For time slope (p1i)
Intercept -.037 .000 -.010 .298
Baseline age -.034 .001 -.027 .006
Education .001 .955 -.003 .722
Race (White) -.046 .000 -.046 .000
Gender (Male) .016 .050 .008 .311
AgeEducation .006 .553 .003 .741
AgeRace -.030 .003 -.029 .002
AgeGender .014 .085 .008 .342
Time-varying preds
SRH Functional Lims Underweight Overweight Married Support Integration Mastery Self-esteem .066 -.170 -.030 -.085 .094 .052 .113 .014 .011 .000 .000 .421 .000 .000 .000 .000 .148 .339
RANDOM EFFECTS
Intercept (u0i) Time slope (u1i) .288 .028 .000 .000 .219 .027 .000 .000
Level-1 (?ij) .263 .251
CONCLUSIONS

• This study suggests that with increasing age,
  adults may be spending less of their
  discretionary time which itself may actually be
  expanding with age participating in physical
  activities.
• On average, stable or increasing levels of
  activity over time were evident in adults up to
  the baseline age of approximately 33 years, with
  adults older than age 33 at baseline exhibiting
  trajectories that were increasingly negative.
• This transition is perhaps earlier in the life
  course than would be expected if declines in
  physical activity were due only to the onset of
  health and functional problems.
• We also recognize that some of our observed age
  differences may also be due to cohort differences
  in leisure-time physical activity.
• Our findings suggest that excess decline in
  leisure-time physical activity among women is
  primarily due to gender differences in
  time-varying health factors.
• Furthermore, the observed convergence of race
  differences in activity appears to be largely the
  result of declines in rates of physical activity
  among older whites, while rates among blacks
  remain fairly stable.
• This may be a case of old age leveling the
  playing field with respect to activity, a
  healthy survivor effect among blacks, or perhaps
  the results of cultural influences

METHODS

• Data source
• Americans Changing Lives study
• (4 waves 1986-2002)
• 3,360 individuals 9,757 observations
• Mean age at baseline 54.18 (SD 17.60)
• Key measure
• Physical activity How often do you work in the
  garden or yard? Engage in active sports or
  exercise? Take walks?
• Data analysis
• Hierarchical Linear Modeling, with occasions of
  measurement nested within individuals
• Level 1 Model
• Activityij p0i p1iTime p2iZ ?ij
• Level 2 Model
• p0i ?00 ?01Baseline Agei ?02Xi
  ?03Baseline AgeiXi u0i
• p1i ?10 ?11Baseline Agei ?12Xi
  ?13Baseline AgeiXi u1i
• p2i ?20
• Where Z equals time-varying predictors (health,
  social and psychological resources and X equals
  time constant predictors (race1white, 0black
  gender 1male, 0female and education level)
  models also control for attrition, and
  time-varying occupational type (blue collar vs.
  other).

ACKNOWLEDGEMENTS This research was supported by
the grant R01 AG031109-02 (Benjamin Shaw, PI)
from the National Institute on Aging.