The Role of Midlife Obesity in Late Life Health Outcomes

advertisement
The Role of Midlife Obesity in
Late Life Health Outcomes
Brenda C. Spillman
The Urban Institute
Supported by funding from the Office of Disability,
Aging, and Long-Term Care Policy, ASPE/DHHS
THE URBAN INSTITUTE
Background
„
Prevalence of overweight and obesity in the U.S.
population is growing
„
Recent studies have found
–
Relative to nonobese elders, obese elders have
similar mortality rates
‹ higher disability rates and have not enjoyed the declines seen
the nonobese population
‹ higher Medicare spending
‹
–
„
Baby boomers have higher rates of obesity in midlife and
earlier than previous age cohorts
Most studies of the elderly population examine only
weight in old age
THE URBAN INSTITUTE
2
Presentation Aims
„
Examine prevalence of current and mid-life
obesity in the 65+ Medicare population
„
Examine the association between
life obesity and late-life health, disability, and
spending outcomes
„
Show how conclusions differ when midlife
obesity can be observed
„
Discuss implications for data collection and
policy
THE URBAN INSTITUTE
3
Data and Sample
„ 1999
–
–
National Long Term Care Survey
Merged Medicare/vital statistics data for a
year follow-up period
Community characteristics from the Area
Resource File and other data merged with
sample
„ Analysis
sample: ~ 5,000 community
residents aged 65+ with and without
disability
THE URBAN INSTITUTE
4
Obesity Measures
„
Current BMI and BMI at age 50 from self-report
height and weight and conventional weight
categories
–
–
–
–
„
Underweight = bmi < 18.5
Normal weight 18.5 - < 25
Overweight 25 - < 30
Obese 30+
“Long-term” weight coded
–
–
–
–
–
Formerly obese, long-term obese, became obese
Long-term under/normal weight
Long term overweight
All other combinations of weight gain/loss not including obesity
Missing current weight/height or weight at age 50
THE URBAN INSTITUTE
5
Outcomes Examined
„ 5-year
mortality (linked vital statistics
„ Self-report
chronic conditions (diabetes,
hypertension, heart disease, arthritis)
„ Chronic
disability (any ADL/IADL
disability, 3+ ADL disabilities)
„ Total
Medicare spending in the 5 years
after interview (linked claims)
THE URBAN INSTITUTE
6
Estimation Models
„
Probit models for 5-year survival, presence of
chronic conditions, and disability
„
Two-part model for Medicare spending
„
Control variables:
–
–
–
Socioeconomic characteristics (age, gender, race,
Hispanic origin, education, income, supplemental
insurance)
Region, urban/rural residence, local health care
market characteristics
Months of follow-up period survived for Medicare
spending models
THE URBAN INSTITUTE
7
Prevalence of Long-term Weight Classes
(at age 50 and at interview)
Currently
obese
Long-term weight
Formerly obese
Long-term obese
Became obese
Long-term under/normal weight*
Long-term overweight
All other weight gain or loss patterns
Missing BMI now and/or at age 50
Percent
4.0
6.5
9.6
29.6
18.7
21.5
10.1
Obese
at
age 50
* Reference group for models
THE URBAN INSTITUTE
8
Long-term weight and health outcomes
Formerly
obese
Long-term
obese
Became
obese
-0.028
-0.050
Five-year mortality
0.144 **
Any chronic disability
0.113 **
0.121 **
0.064 **
Disability in 3 or more ADLs
0.044 **
0.051 **
0.023 **
Diabetes
0.189 **
0.205 **
0.140 **
High blood pressure in the last year
0.143 **
0.197 **
0.224 **
Heart attack/other heart problem in the last year
0.076 **
0.101 **
Arthritis
0.155 **
0.325 **
Any Medicare spending in the next 5 years
Conditional Medicare spending in the next 5 years
0.001
0.416 **
0.006
0.348 **
-0.010
0.208 **
-0.006 *
0.118
Marginal effects for probabilities, percent change for conditional spending
**(*) P-value <= 0.05(0.10) Reference group is long-term under/normal weight
THE URBAN INSTITUTE
9
Disability Rate and Expected Spending
Long-term Weight
Any chronic
disability
3+ ADLs
Per capita
annual
Medicare
spending
All persons age 65+
15.6%
6.8%
$6,307
Long-term under/normal weight
12.4
5.1
5,553
All obese/formerly obese
20.3
8.5
6,963
Formerly obese
28.6
12.9
8,304
Long-term obese
22.3
10.0
7,594
Became obese
15.5
5.7
5,952
THE URBAN INSTITUTE
10
% of Aggregate Disability and Spending
Attributable to Obesity in Midlife or Late Life
Any chronic
disability
3+ ADLs
Medicare
spending
All obese/formerly obese
12.6%
13.6%
4.7%
Formerly obese
3.6%
4.1%
1.6%
Long-term obese
5.4%
6.3%
2.3%
Became obese
3.6%
3.2%
0.8%
THE URBAN INSTITUTE
11
Summary of Findings
„
Midlife obesity is associated with significantly
worse health and disability outcomes and with
higher spending.
„
Former obesity is associated with impaired
survival, but long-term and later onset obesity are
not
„
Becoming obese after midlife appears to be less
harmful, but also is associated with significantly
worse health and disability outcomes and higher
Medicare costs
THE URBAN INSTITUTE
12
Implications
„
Continued upward trends in middle-age obesity, as seen
the Baby Boom generation could
moderate recent downward trends in disability
– increase prevalence of old age chronic disease and Medicare
–
„
Estimates based only on current weight status bias
estimates and understate
Health benefits of stable normal or lower weight
– Detrimental effects of obesity for old age health outcomes
–
„
Results support the importance of initiatives to increase
primary prevention in middle age and earlier.
THE URBAN INSTITUTE
13
Download