Diabetes Treatments

Epidemiology of Diabetes and its Vascular Lesions, Elsevier, 1978 WHO Expert Panel on Diabetes 1979. Kelly West, MD (1925-1980) …

UNDERSTANDING THE RISING TIDE OF DIABETIC CVD
THE STRONG HEART STUDY

Barbara V Howard, PhD MedStar Research Institute

Kelly West, MD 1925-1980
George Lynn Cross Research Professor of Biostatistics and Epidemiology, University of Oklahoma Health Sciences Center Epidemiology of Diabetes and its Vascular Lesions, Elsevier, 1978 Chair, Board of Regents, National Library of Medicine Chair, IDF International Epidemiology Group WHO Expert Panel on Diabetes 1979

Kelly West, MD 1925-1980
1972 - Initiated the Lawton Area Indian Project to study and control diabetes in American Indians Diabetes in American Indians and Other Native Populations in the New World Diabetes, October, 1974 Late 70s - congressional appropriation for diabetes treatment programs for American Indians

The Strong Heart Study
A study of cardiovascular disease in American Indians, supported by the National Heart, Lung, and Blood Institute and the Indian Health Service

Outline
The Strong Heart Study The epidemic of CVD in diabetes Risk factors - conventional and new

Variations in CHD in American Indians by IHS Area - 1975
Rate/100,000
500 400 300 200 100 0
DAK ALA ALB BEM BIL NAS NAV OK PHX POR TUC
USA

ICD-9 Codes 390448

Hospital Admissions
Phoenix PHS Indian Hospital, 1957-1966
Myocardial infarction n 61 No 28 7 1 5 7 13 61 459 115 16 82 115 213 1000 All hospital admissions N 24,121 362 220 105 84 65 164 1000

Tribe Pima Apache Navajo Papago Hopi Other SW Indians Total

Sievers, Ann Intern Med 1967;67:800

Strong Heart Study Objectives
Measure rates of CVD Measure levels of risk factors Determine effect of diabetes on CVD Follow up to evaluate determinants of CVD Follow up to evaluate changes in risk factors
Rapid City

Strong Heart Study Population
4549 American Indians ages 45-74 years
NORTH DAKOTA

Arizona: Pima/Maricopa/Papago in the Gila River, Salt River, and Ak-Chin Indian communities Oklahoma: Apache, Caddo, Comanche, Delaware, Fort Sill Apache, Kiowa, and Wichita South/North Dakota: Oglala Sioux and Cheyenne River Sioux SD and the Spirit Lake Tribe in the Fort Totten area ND

ARIZONA

Spirit Lake Bismarck

Salt River Indian Community Phoenix Ak-Chin Ak-Chin

SOUTH DAKOTA
Rapid City Pine Ridge Cheyenne River Eagle Butte

Gila River Indian Community

Oglala

OKLAHOMA
Oklahoma City Anandarko Lawton

Strong Heart Study
1988-2003

Community Mortality
Study
3574-year-old men and women Deaths between 1984 and 1994 Medical record review of all possible CVD deaths

Physical Examinations of Cohort
1989-91, 1993-5 and 1998-9 ECG, Cardiac and Carotid ECHO Anthropometry, BP, ABI, PFT Blood, urine and DNA samples Medical and medication history Diet, PA, QOL, smoking, alcohol

Strong Heart Study
1988-2003

Cohort Surveillance
Yearly contact Medical record review for all cause and CVD Mortality and nonfatal CVD

Family Study
120 families of 30 members each Examination similar to that of cohort 10cM genetic map

Community Involvement
Study design and implementation American Indian investigators and staff Medical care to participants Data used for community health initiatives Education of community youth Participation in community health initiatives

The Strong Heart Study
Sgufk Ibthag Muschuma Oodham Natsu witu Pihi Comanche Tawahe Cante Suta Woospe Lakota

The Rising Tide of CVD

CVD and Total Mortality Rates
SD/ND vs Strong Heart Study, Men, 1986
per 10,000

600 500 400 300 200 100 0 S/ND SHS S/ND SHS S/ND SHS
CVD Mortality
Total Mortality
45-54 55-64 65-74

CVD and Total Mortality Rates
Arizona vs Strong Heart Study,
Men, 1986
per 10,000

600 500 400 300 200 100 0 AZ SHS AZ SHS AZ SHS
45-54 55-64 65-74

CVD Mortality

Total Mortality

CHD Incidence in American Indians
Compared to ARIC Population
Rate per 1000 person-years

CHD 45-64 years
ARIC SHS

20

10

Women

Men

CHD includes fatal and nonfatal events plus revascularization

Prevalence of Atherosclerotic Plaque in SHS and ARIC/CHS
Prevalence
100 90 80 70 60 50 40 30 20 10 0 SHS ARIC/CHS
P001 P001 P001 P001 P001

45-49

50-54

55-59

60-64

65-69

70-74

75-79

79

Age years

Trends in CVD Mortality
The Strong Heart Study
/ 3 years
80

Women
60 40 20 0
Arizona Oklahoma Dakota Arizona

Men

Oklahoma

Dakota

1984-1986

1991-1993

Prevalence of Diabetes
Strong Heart Study, by Gender and Center
100 80

Women

Men

60
40 20 0 AZ OK ND/SD
Diabetes

AZ
IGT

OK

ND/SD

CVD by Gender and Diabetic Status
Percent/82 year Follow up Number/1000 per year

Nondiabetic
Men
Women

170 85 235 208

21 10 30 25

Diabetic
Men Women

Contribution of Diabetes to CHD
Strong Heart Study
Women Hazard Ratio Prevalence PAR 63 60 76 Men 31 50 51

CVD rates in American Indians are now
higher than in the general US population The rates are increasing The
majority of CVD occurs in people with diabetes

Effect of Epidemic of Diabetes on Duration-Related Complications
Cumulative DM cases Prevalence of complications among those with DM with DM15-year duration

450 No of Cases

50 40

300

Percent

30 20 10

150

0 0 5 10 15 20 25 30 35 40 45 50
Time years

0

16 new cases/yr Death occurs after 30-year DM; Complication incidence 50 after 15-year DM

Change in US CHD Death Rate
1950-2010 projection
Change
20

0

-20

-40

CHD

-60

1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010

NIH, NHLBI Morbidity Mortality 1996 Chartbook on Cardiovascular, Lung, and Blood Diseases, USDHHS, 1996

Conventional Risk Factors New Risk Factors

CVD Risk Factors
HR
Age yr Gender Smoke SBP 10mm LDL 30mg HDL 5mg
104 p 00001 213 p 00001 139 p 001 112 p 00001 112 p 001 094 p 001 102 p 00001 112 p 015 134 p 001 118 p 00001 124 p 00001 092 p 0001

10

125

150

10

125

150

No Diabetes 86- year follow up n 261

Diabetes 82- year follow up n 420

Distribution of LDL Cholesterol in American Indians

LDL Cholesterol mg/dl

Distribution of Systolic Blood Pressure in Diabetic and Nondiabetic American Indians
Diabetic
Diabetic

Systolic blood pressure mmHg

d i a b e t Nondiabetic i c s

Systolic blood pressure mmHg

mm Hg

RR for Total Cholesterol in Framingham vs Strong Heart Study
Men FHS Ref 119 166 193 Men SHS Ref 163 231 287 Women Women FHS SHS Ref Ref 123 109 128 155 171 257

160-199 200-239 240-279 280

Implementation of Findings
Health Care Providers
Increased awareness of CVD because of SHS Development of Indian-specific Framingham predictive equation

Prevention Stop Atherosclerosis in Native Diabetics Study
3-year randomized intervention for LDL and BP usual targets LDL 75mg/dl, BP 115/75 endpoint carotid and coronary ECHO

Conventional Risk Factors New Risk Factors

Age- and Center-Adjusted Hazard Ratios for Albuminuria as Predictor of CVD
80 60 40 20 10
Women Men

Microalbuminuria

80 60 40 20 10
Women

Macroalbuminuria

Men

Albuminuria and LV Structure and Function
No Albuminuria N685 LV Mass gms LV Hypertrophy Fractional Shortening Ejection Fraction 155 36 a 23 a 35 6 a 64 8 a Micro-Alb N519 160 39 a 31 b 35 6 a 63 9 a Macro-Alb N372 181 50 b 49 c 33 8 b 60 12 b

Values with same superscript are not significantly different

Prevalence of Abnormal
Diastolic Function Based on Albuminuria Status
40
P0008 P0001 NS

30 20

10

0

normal

micro

macro

Adjusted covariates: Age, gender, BMI, SBP, HgbA1C, coronary artery disease, ejection fraction, LV mass

Distribution of Fasting Insulin Levels
Strong Heart Study

IR and MS as Predictors of Diabetes

30
35

25

Percent
1st Tertile 2nd tertile Tertile of HOMA IR 3rd Tertile

30 Percent 25 20 15 10 5 0

20 15 10 5 0 Absent Present

ATP III Metabolic Syndrome

IR and MS as Predictors of CVD
10 9
10 9 8 7 6 5 4 3 2 1 0 1st Tertile 2nd tertile Tertile of HOMA IR 3rd Tertile

8

Percent

7 6 5 4 3 2 1 0 Absent Present

Percent

ATP III Metabolic Syndrome

The insulin resistance syndrome is not a strong determinant of CVD in non-diabetic American Indians Among the CVD risk factors and parameters of CV function, only lipoproteins and PAI-1 change significantly with increasing IR IR leads to diabetes before it leads to CVD

Inflammatory Markers?

Age- and Center-Adjusted Hazard Ratios for Fibrinogen as Predictor of CVD
80 60 40 20 10
Women Men

Elevated Fibrinogen and ECHO Abnormalities
OR adj ECHO
LVH Art Stiffness Low Contr

95 CI 110-173 111-188 070-157 122-187

138 145 105
151

Any of above

CRP - Population Comparisons
Study

SHS NHANES PRINCE CHS IRAS NHS MRFIT

Median/ Mean 38

Pop based Yes

Comments

Am Indian US 40 years US 21 years US AV 77 years Diabetes Women High-Risk Men

21 205 267 38 28 195

Yes Yes Yes No No No

CRP and CVD
Model RR for CRP 1 2 3 4 12 129 122 115 95 CI 105-138 112-149 105-140 099-135 Covariates none
gender, age, center, BMI, smoke, meds plus HT, DM, albuminuria plus PAI-1, fibrinogen

CRP high in populations with high rates of obesity and diabetes CRP, therefore, is a weaker predictor of CVD Fibrinogens stronger predictive value may reflect its relation to thrombosis as well as inflammation

Adiponectin ACRP30/ adipoQ
Produced by adipocytes, but plasma levels lower in obesity Sequence homologies to collagen and complement factor C1q Metabolic actions - related to increased insulin sensitivity Relations to inflammation Structural homologies to TNF
Inhibits NF-B signaling

Adiponectin as Predictor of CHD:
Multivariate Analysis
Model 1 Model 2 Model 3
Waist, age, FAT , SBP, albuminuria
HDL, LDL QUICKI

0

025

05

075

1

125

15

175

Odds Ratio CHD [95 CI]

Adiponectin as a Predictor of CHD:
Subgroups Model
2
Male DM Male No DM Female DM Female No DM

0

025

05

075

1

125

15

175

Odds Ratio CHD [95CI]

Adiponectin is: Related to BMI -, fasting insulin -, HDL Protective against later rises in fasting BG Positively related to albumin excretion

Markers of Pre-Clinical Disease?

Abnormalities of Cardiac Structure and Function in Diabetic and Non-Diabetic SHS Participants Without Known Coronary Disease
30

All p0001

Prevalence

25 20 15 10 5 0
LVH Low Contr Abnl Filling
Diabetes Non-DM

Relation of LV Mass/Height27 to Subsequent CVD and Non-CVD Mortality
3 25 2 15 1 05 0 Non-CVD Death CVD Death
p 001

Proportion dead

Normal LV Mass LV Hypertrophy

Adjusted for covariates

LV Function and CVD Mortality in Diabetes
100
11

98

Cumulative Survival

Cumulative Survival

10

96

94

EF
92

9

40 40-54

E/A15
No

90 0

55
10 20 30 40 50

8 -10 0 10 20 30 40 50 60 70

Yes

Months

Time months

LVEF

Abnormal LV Filling

Univariate Relative Risk of Mortality in Diabetes
10 9 8 7 6 5 4 3 2 1 0
207

p0001
954

ST Depression PCA Ratio QTc

Relative Risk

p0001 p0001
317 p0018 468

p0001 p00004
211 195

Cardiovascular Mortality

All-Cause Mortality

Multivariate Relative Risk of
Mortality in Diabetes
40 35 30
p00009
368

p0006
261

ST Depression PCA Ratio QTc
p0002 p0001 p0242
133 203

Relative Risk

25 20 15 10 05

p0082
187

236

00 Cardiovascular Mortality

All-Cause Mortality

Adjusted for age, sex, BMI, BP, HDL, LDL, TGs, albuminuria, alcohol use, smoking, prevalent CHD, and center

Application to the Clinical ECG
50 uV of ST depression 1/2 mm half
a small box The QTc is the longest interval from onset of QRS to the end of the T-wave PCA cannot be measured from ECG alone

Strong Heart Family Study
Phenotypes include ECG/BP, ECHOs, lipid and thrombotic markers, and glucose tolerance Covariates include demographic factors, lifestyle, reproductive history, and medical history

Numbers of Examined Relative Pairs
Strong Heart Family Study
Relationship
Parent-offspring Siblings Half-siblings Avuncular Grandparent-grandchild Grand avuncular First cousins First cousins once removed Second cousins Other Total Relative Pairs DA N326 336 330 86 954 129 395 1142 1091 487 621 5571 OK N310 245 319 77 802 35 139 843 409 57 322 3248 AZ N345 319 350 44 721 12 261 866 909 233 145 3975 SHFS N 981 900 999 207 2477 291 795 2851 2409 777 1088 12794

Heritabilities
for CVD Risk Factors
Strong Heart Family Study
Phenotype Proportion of Variance Due to Genes SE
44 54 39 50 40 23 29 44 23 007 007 006 007 007 006 008 008 006

BMI WHR LDL-C HDL-C lnTG SBP F Glu ln Ins ln Fibrinogen

Heritability h2 of LV Mass and Geometry in SHS-FS
045 04 035

Heritability

03 025 02 015 01 005 0
LV Mass LVIDd RWTd Overall Center

Heritability for QTL needed for power 80 in SHS

Genotype by Diabetes Status Interaction for CVD Risk Factors
Trait BMI Fat WHR HDL-C TG g, D 523 81 559 69 041 07 897 134 373 94 g, ND 581 55 614 57 058 06 390 48 g D, ND 34 p0002 52 p0005 69 p0353 02 p0096

924 108 62 p0307

Genetic and Environmental Correlations Between Diabetes and CVD Risk Factors
Risk Factor BMI Fat WHR HDL-C TG Fibrinogen PAI 1 SBP Genetic Environmental 55 14 10 13 38 14 11 14 58 15 13 14 -37 27 -32 18 65 21 33 13 40 17 20 11 67 17 18 10 57 21 11 11

SUMMARY
There is a rising tide of CVD in diabetes
LDL and blood pressure are strong risk factors Albuminuria, fibrinogen and ECHO/ECG abnormalities are also strong risk factors

Current strategy for prevention
Aggressive control of LDL and BP

SUMMARY, cont
Future directions for understanding the impact of
diabetes on CVD
Understand mechanism of cardiac abnormalities Explore further the inflammatory/thrombotic axis Study adipocyte mediators

Future direction - Genetics
Localize and identify genes contributing to CVD risk and measure their effects Identify genes that modulate therapeutic responses

The Strong Heart Study Investigators
Lyle Best, MD Linda Cowan, PhD Richard Devereux, MD Richard Fabsitz, PhD Jeffrey Henderson, MD WmJames Howard, MD Elisa Lee, PhD Albert Lewis Robert Lindsay, MD Kari North, PhD Jean MacCluer, PhD Peter Okin, MD Vittorio Palmieri, MD Helaine Resnick, PhD Everett Rhoades, MD David C Robbins, MD Giacomo Ruotolo, MD, PhD Thomas Welty, MD Jeunliang L Yeh, PhD Ellie Zephier, RD

The Strong Heart Study Field Staff
Marcia OLeary, BSN Lillian Brown Cherie Kessler Wendy Lawrence Francine Red Willow Taqueer Ali, PhD Linda Poolaw Karen Kimberly Stephanie Gomez Betty Jarvis, BSN Bert Lewis Rosinna Briones Mary Rybka Nanette Oram

Source:diabeteseducationandresearchcenter.org