Close
Executive SummaryIntroductionFindingsSmoking and Obesity- A Public Health Success and ChallengeExplore How the Prevalence of Obesity and Smoking Has ChangedComparison with Other NationsCore MeasuresBehaviorsCommunity & EnvironmentPolicyClinical CareOutcomesSupplemental MeasuresState Summaries
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
US SummaryAppendixData Sources and MeasuresMethodology2016 Model DevelopmentScientific Advisory CommitteeThe TeamConclusion
AlabamaAlaskaArizonaArkansasCaliforniaColoradoConnecticutDelawareDistrict of ColumbiaFloridaGeorgiaHawaiiIdahoIllinoisIndianaIowaKansasKentuckyLouisianaMaineMarylandMassachusettsMichiganMinnesotaMississippiMissouriMontanaNebraskaNevadaNew HampshireNew JerseyNew MexicoNew YorkNorth CarolinaNorth DakotaOhioOklahomaOregonPennsylvaniaRhode IslandSouth CarolinaSouth DakotaTennesseeTexasUtahVermontVirginiaWashingtonWest VirginiaWisconsinWyomingUS SummaryAppendixData Sources and MeasuresMethodology2016 Model DevelopmentScientific Advisory CommitteeThe TeamConclusionRankings Calculation
For each measure, the raw state-level data are obtained from secondary sources (Table 5) and presented as a “value.” The most current data available as of October 2016 were included in the analysis. The score for each state is based on the following formula:
This “z score” indicates the number of standard deviations a state value is above or below the national value. A 0.00 indicates a state has the same value as the nation. States with higher values than the national value have a positive score; states below the national value have a negative score. To prevent an extreme score from exerting excessive influence, the maximum score for a measure is capped at +/- 2.00. If a US value is not available from the original data source for a measure, the mean of all state values is used with the exception of measures from the Behavioral Risk Factor Surveillance System (BRFSS). For BRFSS measures, the median of the state values is used for the US value to conform to the Centers for Disease Control and Prevention methodology. Overall score is calculated by adding the scores of each measure multiplied by its assigned weight (the percentage of the total overall ranking).
Each of the five major categories of the America’s Health Rankings model of health (behaviors, community & environment, policy, clinical care, and outcomes) are assigned different weights (Table 7). Measure weights can be found at http://www.americashealthrankings.org/AR16/about.
The overall ranking is the ordering of each state according to the overall score. The ranking of individual measures is the ordering of each state according to the measure’s value, with the exception of immunizations--adolescents and infectious disease, which are ranked according to score. Ties in values are assigned equal ranks. Not all changes in rank are statistically significant.
Weights by Model Category
Table 7
BRFSS data were analyzed using Stata v14.1 to account for the complex survey design. Responses of “refused”, “don’t know,” or “not sure” were excluded from the analysis, but are reflected in standard error and confidence interval estimates. For subpopulation measures, “refused”, “don’t know,” or “not sure” responses were coded as missing. For calculating subpopulation estimates, the population of interest is specified in a manner that avoids deletion of cases. This ensures an accurate variance estimation.
Population estimates for measures from BRFSS were calculated using the specified survey weights and represent the non-institutionalized adult population. Discrepancies between prevalence estimates and population estimates are likely due to random sampling error and nonrandom response biases.
BRFSS made two changes in 2011 to improve their survey methodology. Due to these changes, 2011 to 2015 BRFSS data is not comparable to previous years.
Trends in Smoking and Obesity Prevalence, 2012 to 2016
Five-year trends in smoking and obesity prevalence in the United States were examined to highlight differences across states and levels of education. Using 2011 to 2015 BRFSS data, state and national prevalence estimates of smoking and obesity among adults aged 18 years and older were calculated for each edition year from 2012 to 2016. State and national prevalence estimates of smoking and obesity among adults aged 25 years and older were also calculated for each of four self-reported education levels (less than high school education, high school graduate, some college, and college graduate). Statistically significant differences between groups were determined by non-overlapping 95% confidence intervals. Because of changes in BRFSS methodology prior to 2011 (2012 edition), data was limited to 2011 to 2015 (2012 to 2016 editions) to allow comparability.
Average rate of change per year (slope of a least-squares tted line) in prevalence over the five-year period (2012 to 2016) was calculated for the analysis. The average prevalence over the same five-year period was calculated for comparison with average annual rate of change by state and education level. Using rate of change per year better accounts for expected year-to-year variation in the measures compared with reporting relative or absolute differences between editions 2012 and 2016.