CALegTaskfObesDiab22 Aug07

Obesity and Diabetes Epidemiology: Implications for Control & Prevention: 

Obesity and Diabetes Epidemiology: Implications for Control & Prevention Antronette K. (Toni) Yancey, MD, MPH Professor, Department of Health Services, Co-Director, Center to Eliminate Health Disparities UCLA School of Public Health

Health challenges have changed, but our health care system has not:Clinical encounter influences only ~10% of health status: 

Health challenges have changed, but our health care system has not: Clinical encounter influences only ~10% of health status Factors influencing gain in life expectancy: 1900-19992 National spending for population-based prevention21

Unhealthy eating and inactivity are leading causes of death in the U.S.: 

Unhealthy eating and inactivity are leading causes of death in the U.S. HHS estimates that unhealthy eating and inactivity contribute to 310,000 to 580,000 deaths each year. That’s 5 times more than are killed by guns, HIV, and drug use combined.1 The typical American diet is too high in saturated fat, cholesterol, salt, & refined sugar and too low in fruit, vegetables, whole grains, calcium& fiber. Such a diet contributes to four of the seven leading causes of death and increases the risk of numerous diseases, including: heart disease diabetes cancer high blood pressure obesity osteoporosis stroke 60% of Americans are at risk for health problems related to lack of physical activity (i.e. get less than 30 minutes of moderate activity 5 or more times per week). 2 Leading Contributors to Premature Death1 Leading Causes of Death3 (Diet is a leading risk factor for causes of death shown in bold or green.)

Slide4: 

Obj. 19-2 Total White Female Male Black Female Male Mexican American Female Male Target Adult Obesity: 1988-94 to 1999-2000 0 10 20 30 40 50 Note: Data are for ages 20 years and over, age adjusted to the 2000 standard population. Obesity is defined as BMI >= 30.0. Black and white exclude persons of Hispanic origin. Persons of Mexican-American origin may be any race. Source: National Health and Nutrition Examination Survey, NCHS, CDC. Percent 1988-94 Race/Ethnicity

Slide5: 

Age-Adjusted Prevalence of Overweight & Obese by Race – NHANES Adults Left bars = BMI 25.0 or higher; right bars = BMI 30.0 or higher

Slide6: 

Source: National Center for Health Statistics Death rates by cause for persons aged 45 to 65, 1995 Men Women Deaths per 100,000 persons

Slide7: 

Age-adjusted, 1998 data Source: National Center for Health Statistics, Health US 2000, table 31 Years of Potential Life Lost to Diabetes YPLL before age 75 y per 100,000 population

U.S. Childhood Obesity Epidemic Trends: 

U.S. Childhood Obesity Epidemic Trends Obesity prevalence in U.S. children and adolescents by age and time frame, 1963-2004

Type 2 Diabetes Incidence Among Adolescents: 

Type 2 Diabetes Incidence Among Adolescents 1982: 0.7 / 100,000 per year 1994: 7.2 / 100,000 per year 2004: 13.9-16.0 / 100,000 per year Sources: Pinhas-Hamiel O et al. J Pediatr 1996;128:608-15 Daniels et al, Circulat 2005;111;1999-2012 >20-fold increase

?Influence on Life Expectancy: 

?Influence on Life Expectancy For the first time in thousands of years, this generation of children may live shorter lives than their parents* Chances of developing diabetes at birth for African-American and Latino children is about twice that of white children (1 in 2 vs. 1 in 4)** *Olshansky et al., NEJM 2005;352(11):1138-45 ** Daniels et al, Circulat 2005;111;1999-2012

Prevalence of Obesity among LA County Adults by Ethnicity, 1997-2005: 

Prevalence of Obesity among LA County Adults by Ethnicity, 1997-2005

Staying healthy is easier for some than for others…: 

Staying healthy is easier for some than for others… UPPER SES LOWER SES Education College GED or HS Housing Own / Safe Rent / Safe? Physical activity Gyms /Parks Parks?, “move insecure” Advertising Sparse Pervasive Neighborhood stores Fruit/Veg, food secure Drugs/Alcohol, food insecure Police Helpful Abusive Healthcare Private Doc ER, VA Sick leave Accrued None Leisure priority Exercise Rest Work conditions Safe, hi decis. lat., Hazardous, lo decis. lat., no +flex time no flex time Child care Nanny/hi-qual facil. Family/neighbor, lo-qual facil. Elder/disabled care HHW/hi-qual facil. Family/neighbor, lo-qual facil. Criminal just. sys. Little contact Much contact Premature M&M Low High

Years of Potential Life Lostby Ethnicity (per 100,000): 

Years of Potential Life Lost by Ethnicity (per 100,000)

Slide16: 

Energy Intake Energy Expenditure Energy Balance Individual Factors Behavioral Settings Social Norms and Values Communities Worksites Health Care Schools and Child Care Home Demographic Factors (e.g., age, sex, SES, race/ethnicity) Psychosocial Factors Gene-Environment Interactions Other Factors Government Public Health Health Care Agriculture Education Media Land Use and Transportation Communities Foundations Industry Food Beverage Retail Leisure and Recreation Entertainment Physical Activity Sectors of Influence Food & Beverage Intake

Excess physical environmental risk in underserved communities:: 

Excess physical environmental risk in underserved communities: targeted/exploitative marketing excess fast food outlets few supermarkets limited healthful shelf choices & poorer quality produce high-fat food availability (home, church poorer public/less private transportation distance to private fitness facilities few worksite fitness opportunities few/poor neighborhood recreation facilities lesser neighborhood safety poorer public/less private transportation Adapted from Kumanyika S. Obesity in Minority Populations. In Fairburn G & Brownell K, Eating Disorders and Obesity. A Comprehensive Handbook, 2002.

Buford Highway projectEstimating projected increases in walking: 

Buford Highway project Estimating projected increases in walking Before After Avg LOS = D (4.1) Est’d LOS = B- (2.4) Avg min. walked/week = 51 Est’d min. walked/week = 62 -175 Increased walking based on improved walkability

Excess economic environmental risk in underserved communities:: 

Excess economic environmental risk in underserved communities: low neighborhood demand for healthy food choices low family incomes other household expenses little homegrown food financial incentives for under-resourced schools by commercial vendors limited investment in parks/recreation facilities fitness facility fees cost of exercise equipment less stable employment patterns fewer trained PE instructors larger PE classes poorly equipped facilities

CA PA/PE Assessment Study: 

CA PA/PE Assessment Study TCE-funded, 3-year study Random sample of 10 school districts from the CDE Public School Directory stratified by region of the state (southern, central, northern), tertile of Fitnessgram scores & SES Average district—12,000 students, incl. 43% Latino and 53% White Mean FRPL eligibility 58% (SES measure) $100 incentive for participation 77/200 schools responded (38.5% response rate, R=26.7%-58.8%)

CA PA/PE Assessment Study: 

CA PA/PE Assessment Study Three data collection methods: Principal’s (or most knowledgeable designee) survey—45% principals, 48% PE teachers Environmental audit of sample of responding schools (facilities, recess, PE class) Stakeholder survey (e.g., teachers, PTA members, school board members, fitness-related non-profit organization staff)

School Physical Activity & Physical Education (PE) Assessment Study:Principal survey findings: 

School Physical Activity & Physical Education (PE) Assessment Study: Principal survey findings Participation rates lowest among kindergartners & 10th-12th graders Lowest rates of adherence to mandated PE minutes in primary grades, K-3rd Mean student-to-teacher ratio in PE classes is 41-52 Only ~2 in 5 elementary schools report that all of their PE instructors were certified

Sample: 

Sample Aim for 3 schools in 10 school districts (some replacements due to recruiting problems and absence of elementary schools in one “High School Districts”) 2 classes observed by at least one rater at each school 18 classes at 9 elementary schools 20 classes at 10 middle schools 20 classes at high school Adapted SDSU SOFIT measure with additional observations for training and inter-rater reliability assessment Intervals originally coded as 0 (student cannot be observed) were recoded as 2 (standing)

Slide26: 

Avg. amount of PE class time spent in MVPA by school level School Level The amount of P.E. class time that students were physically active was slightly higher in higher grades but there was a great variation within each school level.

Slide27: 

% PE class time in MVPA by % FRPL-eligible & by district avg. FitnessGRAM score

Slide28: 

Avg. amount of PE class time in MVPA by class size (secondary schools only) Class Size The amount of P.E. class time that students were physically active was less in larger classes. N=6 N=12 N=12 N=10

Rel. between PE Quality (%class t in MVPA) & API Score in High & Low SES Schools: 

Rel. between PE Quality (%class t in MVPA) & API Score in High & Low SES Schools

Excess sociocultural environmental risk in underserved communities:: 

Excess sociocultural environmental risk in underserved communities: traditional cuisine/cultural anchoring prevalent obesity/norms body image ideals female roles fasting-feasting/ (perceived) food insecurity higher stressor levels/ comfort foods cultural attitudes about PA, rest fears about safety female roles cultural reverence for cars hairstyle-related concerns about sweating increased “screen” time

Physical Activity Levels, %L.A. County Adults, 1999: 

Physical Activity Levels, % L.A. County Adults, 1999

Physical Inactivity Levels: TV viewing/computer use, %L.A. County Adults, 1999: 

Physical Inactivity Levels: TV viewing/computer use, % L.A. County Adults, 1999

Media Project: six-city outdoor advertising content analysis: 

Media Project: six-city outdoor advertising content analysis Funded by CA DHS, UT, Penn, NYU & RWJF Cities: LA, Philadelphia, Austin, Sacramento, Fresno, New York City Comparing high & low SES predominantly black, Latino, & white neighborhoods (all 6 categories not available in all cities, e.g., high SES black in Sacramento and Fresno) Utilizing secondary data from CHIS, LACHS, grocery store scanner (MOU with major supermarket chain) purchase data for correlational analyses

Media Project: six-city outdoor advertising content analysis: 

Media Project: six-city outdoor advertising content analysis Digital cameras and GPS units used to record a photo and location coordinates of all ads in zip code Electronic abstraction form captures domains including: content, size, nature of appeal, density, visibility, ethnicity of subjects, weight status of subjects

Media Project: six-city outdoor advertising content analysis: 

Media Project: six-city outdoor advertising content analysis Examples Large billboards (14’ x 48’) Medium-sized billboards (12’ x 25’, “30 sheets”) Small billboards (6’ x 12’, “8 sheets”) Convenience and supermarket displays (“two-sheets”) One-sheet posters Bus bench advertisements Bus shelter advertisements Murals Sandwich board Exterior subway or train station advertisements

Preliminary findings: 

Preliminary findings Absence of billboards and near-absence of other outdoor advertising in affluent white neighborhoods—existing ads unrelated to weight Essentially no outdoor advertising of PA-promoting goods & services in any community, but large amount sedentary entertainment & transportation ads in low-income communities Pervasiveness of advertising in low-income white & Latino communities, but more fast food, sugar-sweetened and alcoholic beverages in latter City of LA has moratorium on new billboards, but in low-income Latino comm., large # of new side of building ads similarly framed Findings must be interpreted in light of historical covenants, fewer ads trad. In unincorp. areas

Preliminary Findings: 

Preliminary Findings

Unhealthy Food Ad Density near Schools/Day Care Centers(# within ¼ mi of sugary beverage/fast food ads): 

Unhealthy Food Ad Density near Schools/Day Care Centers (# within ¼ mi of sugary beverage/fast food ads)

Capturing targeted media influence: 

Capturing targeted media influence Social action theory (Kumanyika & Ewart) and social learning theory (Bandura) emphasize influence of role modeling and social comparisons with “like” others on self-efficacy and behavior

Capturing targeted media influence: 

Capturing targeted media influence A TV content analysis study found that more than 4 times as many “black prime time” actors were overweight as actors on “general audience” shows Tirodkar MA, Jain A. Food messages on African American television shows.Am J Public Health 2003;93:439-41.

Obesity as social contagion: 

Obesity as social contagion Social network analysis of 32 years of Framingham cohort study data on 12,000 adults weight gain/loss/maintenance over time influenced by friends & family, extending to 3 degrees of separation, independent of immediate physical environment (Christakis & Fowler, NEJM, 2007) Concluded that each individual influences the social norm for her/his circle Thinness as well as obesity “contagious”

Self-Perceived Overweight by Ethnicity & Gender, % LA County Adults: 

Self-Perceived Overweight by Ethnicity & Gender, % LA County Adults

Physical Activity Levels: TV viewing>2 hrs/d vs. regular PA, %California adolescents, 2001: 

Physical Activity Levels: TV viewing>2 hrs/d vs. regular PA, % California adolescents, 2001

Need for Focus on Behavior vs. Weight: 

Need for Focus on Behavior vs. Weight Cultural influences on obesity definition Behavioral economics (immediate availability governs choices) Human tendency to “do better” when we feel better, not worse, about ourselves Evolutionary “hard-wiring” Stress management/mood elevation benefits Suggests need to focus on re-integrating PA and increasing healthy food availability, appeal, pricing competitiveness, and modeling

Influence of Self-Perceived Weight Status on PA,% LA County Adults: 

Influence of Self-Perceived Weight Status on PA, % LA County Adults Overall, regardless of BMI, those perceiving themselves as overweight more sedentary than those with average wt. self-perception (45% vs. 30%) Influence most pronounced for males and normal weight individuals Overwt. self-perception not assoc. with sedentariness among white women, the only one of the 6 ethnic-gender groups included in which BMI<25 normative In multivariate analysis, self-perceived overweight, not BMI, predicted sedentary behavior (OR=1.40, CI 1.19, 1.64) Yancey, Simon et al., Obes (Res) 2006;14:980-8. Yancey, Wold et al., Am J Prev Med, 2004;27:146-52.

SB19 Early Implementation Study: Baseline demographics: 

SB19 Early Implementation Study: Baseline demographics

Distribution of dieting classification, by BMI-for-age: 

Distribution of dieting classification, by BMI-for-age

Slide53: 

N = 1,767 7th graders OR =0.79, 95% CI: .72, .86, p < .0001, controlling for age, gender and subsidized lunch eligibility

Health behavior correlates of not trying to control weight:: 

Health behavior correlates of not trying to control weight: Those not trying to control weight vs. other diet status categories ate fast food less often (p = .04), ate sports bars less often ( p = .004), drank sports drinks less often (p = .001), drank sweetened juices less often (p =.02) and participated in more minutes of physical activity in PE class (p = .03). They also did better academically than other students (p = .005). *note: logistic regression models included age, gender and meal subsidy eligibility as covariates

Lesser Effectiveness of Key Environmental Interventions in Underserved Groups: Example: 

Lesser Effectiveness of Key Environmental Interventions in Underserved Groups: Example Posting of Signs Promoting Stair Usage (suburban Baltimore mall) Overall, stair use increased from 4.8% to 6.9%, 7.2%, depending upon which of 2 signs used Among whites, increased from 5.1% to 7.5%, 7.8% Among blacks, changed from 4.1% to 3.4%, 5.0% Among n’l wt, inc from 5.4% to 7.2%, 6.9% Among overwt, inc from 3.8% to 6.3%, 7.8% Andersen, Franckowiak, Snyder et al., Ann Int Med, 1998;129:363-369.

California Department of Transportation, District 7 Headquarters100 S. Main Street, Los Angeles, CA: 

California Department of Transportation, District 7 Headquarters 100 S. Main Street, Los Angeles, CA

Study locations; time of day: 

Study locations; time of day Stairwells #4 and #7 in CalTrans building at 100 S. Main Street Observers were randomly assigned to one stairwell, then switched to 2nd stairwell Additional observers were posted at building entry to permit sociodemographic comparisons of employees to stair users—SOPAWS (Systematic Observation of PA in Worksite Stairs) Observations conducted during working hours of weekdays, from 8 am-5 pm

Ethnic distribution of stair users, by gender & ethnicity: 

Ethnic distribution of stair users, by gender & ethnicity

Ethnic distribution of building entrants, by gender & ethnicity: 

Ethnic distribution of building entrants, by gender & ethnicity

Stair Use Differences Between Gender and Ethnic Groups: 

Stair Use Differences Between Gender and Ethnic Groups

Shift in health promotion field (Spectrum of Prevention): 

Shift in health promotion field (Spectrum of Prevention) The most effective and sustainable PH intervention approaches of the past two decades are the more “upstream” ones (structural/environmental vs. individual-level), involving social norm change: Tobacco control Alcohol consumption and driving Littering & recycling Seat belt & child safety seat usage

Spectrum of Prevention:Health behavior change model: 

Spectrum of Prevention: Health behavior change model Level 1: Strengthening individual knowledge and skills Level 2: Promoting community education Level 3: Educating service providers Level 4: Fostering coalitions and networks Level 5: Changing organizational practice Level 6: Influencing policy and legislation

Lift Offs Work!: the Rapidly Growing Evidence Base : 

Lift Offs Work!: the Rapidly Growing Evidence Base Documented individual and organizational receptivity to integrating PA on paid work time Contribute meaningfully to daily accumulation of MVPA Motivational “teachable moment” linking sedentariness to health status for inactive folks Improvements in clinical outcomes from as little as one 10-min. break/day—BP, BMI, waist circ., mood, attention span, cumulative trauma disorders “Spill-over” or generalization to inc. active leisure Favorable cost-benefit ratio, eg, L.L. Bean mfg plant