Ancestry differences in hypertension

risk, progression and

treatment response

Prof. Kennedy Cruickshank

Cardiovascular Medicine & Diabetes,

King’s College & Health Partners

= Guy’s & St Thomas’ Hospitals,

London, UK

RCP - BHS joint symposium

I have no disclosures

Heart attack and Stroke 20y Incidence by Ethnicity, in west London, 1988-2011

The ‘Sabre’ cohort – Ealing & Brent; Chaturvedi N et al; JACC 2013

N = 2049 630 1517

CHD

STROKE

Frequency distribution of LV mass indexed to

BSA in black vs white women (A) and men (B)

Mark H. Drazner et al.

Hypertension

2005;46:124-129

Representative

samples in the

Dallas Heart study

The ‘null’ hypothesis

More (High) Blood Pressure in W Afrs & Caribbeans but for given BP levels (eg: ‘x’), no difference in outcomes..

- Cruickshank 1989

Inci

den

ce r

ate

o

f o

utc

om

es

Europeans

West African / C’bean

Origins of High Blood Pressure risk

- Genes.. variants?

- Development – life-course exposure..?

- Risk factor burden..?

- Social determinants..?

Trajectories of systolic blood pressure (A) and body mass index (BMI) (B) from 1 to 20 years

of age in offspring of normotensive (blue), pregnancy-induced hypertension (PIH; orange)

and complicated hypertensive (red) pregnancies.

Esther F Davis et al. BMJ Open 2015;5:e008136

©2015 by British Medical Journal Publishing Group

BM

I –kg

/m2

SBP

- m

mH

g

Bogalusa Heart Study: multiple regression

on Longitudinal systolic BP at 15y (n= 182, Af.Am 92)

95 % CI Standard beta

coefficients P value

Birth weight -8.6 to 4.1 - 0.36 <0.01

Height 0.27 to 0.57 0.38 <0.01

BMI 0.30 to 0.85 0.3 <0.01

DWT 04 -1.3 to –0.3 - 0.25 0.01

SBP at 4y 0.08 to 0.44 0.19 <0.01

NB. Ethnic difference in 15y BP ‘accounted for’ by birth weight

Cruickshank et al Circulation 2005;111:1932-37

Adolescence 11-13y & 14-16y, (n=~6000)

CVD, respiratory

& Mental Health

<birth weight linked>

Early life & childhood

+ bio-markers, arterial stiffness, accelerometery, dietary recall, own SEC,

parenting qualitative interviews

21-23y: Pilot study (n=665)

Work life & beyond

health & life trajectories future follow-up linkage of medical data generational studies

DASH - health over the life course, ~1000 in each major ethnic group

100

105

110

115

120

systo

lic B

P (

mm

Hg

)

■ 13.1 - 18.8 kg/m² ♦ 18.8 - 21.9 kg/m² ● 21.9 - 39.5 kg/m²

White Black

Caribbean

Black

African

Indian Pakistani/

Bangladeshi

MixedWhite

Other

Systolic BP by BMI tertiles among adolescent girls

The MRC DASH Study in London Schools 11-13y olds

Harding S, Maynard M, Cruickshank JK. J Hypertension 2006

Systolic blood pressure, by age and ethnicity for males and females: means/percentage and 95% confidence

intervals adjusted for gender and ethnicity

Males Females

100

105

110

115

120

125

130

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

Whites UK BC BA Indian PB Others

Syst

olic

Blo

od

Pre

ssu

re, m

mH

g

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

Whites UK BC BA Indian PB Others

(BC – Black Caribbean, BA – Black African, PB – Pakistani/Bangladeshi)

Systolic blood pressure from adolescence to early adulthood for males and females: association with

longitudinal measures of adiposity* - at ages 21-23y

100

105

110

115

120

11-13y 14-16y 21-23y

Syst

olic

Blo

od

Pre

ssu

re, m

mH

g

Males 95% CI Males

Females 95%CI Females

Coeff 95% CI p-value

Males

Waist to height ratio 32.49 (22.05,42.93) <0.001

BMI, kg/m2 0.17 (0.08,0.26) <0.001

Overweight (Normal weight - Ref)

Overweight/Obese 1.74 (-0.31,3.79) 0.097

Females

Waist to height ratio 32.69 (24.84,40.55) <0.001

BMI, kg/m2 0.54 (0.42,0.65) <0.001

Overweight (Normal weight - Ref)

Overweight/Obese 3.75 (2.00,5.50) <0.001

*Mixed-Effects Linear Regression Model coefficients adjusted for age, ethnicity, waist to height ratio (or BMI, or

overweight status), parental/own employment and currently smoking.

Harding S et al.. Cruickshank JK – in preparation

Relationship of Pulse Wave Velocity with BP

4

6

8

10

12

Pu

lse

Wa

ve

Ve

locity (

m/s

)

80 100 120 140 160 Systolic BP

60 70 80 90 100 Diastolic BP

ρ=0.31 (p<.0001)

ρ=0.18 (p=0.000)

Feasibility Follow-up – ages 22-23y

Many young people

already ‘stiffening’..?

Cruickshank et al Hypertension June 2016

Individual Patient Meta-analysis of Arterial Stiffness and Mortality – an intermediary

outcome not a risk factor..

Ben-Shlomo Y et al, JACC 2014

0

10

20

30

40

%

< 50 60 70 80 90 100 110 120 130 140

Diastolic Blood Pressure mmHg

Manchester 23 Jamaica 15 Cameroon – urban 4.4

% BP (treated)

Cameroon – rural 0.5

Cruickshank et al, J Hypert 2001; 19: 41-46

Age-adjusted blood pressure distributions of

west African-origin populations

Severe childhood malnutrition (SCM) - still a problem…

- 2 distinct forms - based on

presence or absence of nutritional oedema:

Oedematous

(Kwashiorkor & Marasmic-kwashiorkor)

Non-oedematous

(Marasmus)

15

Differences in cardiovascular measures (SD scores)

- controls vs. all Malnutrn survivors at ±30y

16

Measurement

(standardised score)

Controls – all SAM survivors

Difference 95%CI, p-value

Controlled for age and sex

Visceral fat mass -0.09 -0.45 to 0.27, 0.6

Systolic blood pressure -0.22 -0.55 to 0.12, 0.2

Diastolic blood pressure -0.40 -0.71 to -0.08, 0.02

Heart rate 0.21 -0.14 to 0.56, 0.2

Pulse Wave Velocity 0.35 0.06 to 0.65, 0.02

Stroke Volume 0.49 0.15 to 0.82, 0.005

Cardiac Output 0.56 0.23 to 0.90, 0.001

Ejection Fraction -0.41 -0.76 to -0.06, 0.02

LV outflow tract diameter 0.71 0.39 to 1.03, <0.001

Systemic Vascular Resistance -0.69 -1.03 to -0.35, <0.001

LV Mass index -0.02 -0.35 to 0.31, 0.9

Central Systolic BP -0.15 -0.47 to 0.18, 0.4

Tennant-Martin et al – Hypertension 2014

Vascular resistance & LV outflow

tract +/- 30 years after

Malnutrition

17 Tennant-Martin et al – Hypertension 2014

Lifetime transition..?

Recovered

Kwashiorkor, small

baby…

High BP

Type 2 Diabetes

? Population

Risk ?

X 2

Expected

NB

Temperature

difference

Nigerian birth cohort – BP change up to age 3y by maternal

malarial status

Courtesy of Jasmin Farikullah-Mirza, O Ayoola, Clayton P & our team – submitted 2016

Maternal malaria +ve

Unexposed ‘controls’

Increased

Vasc GFs

Sick genes, Sick individuals or

Sick populations with chronic

disease? An example from studying

diabetes & hypertension in African-origin

populations.

JK Cruickshank, J-C Mbanya, R Wilks, B Balkau, N McF Anderson, T Forrester

Int J Epidemiol 2001; 30: 111-117

Developmental & Environmental, rather than genetic-variant, bases for ethnic variation in High Blood Pressure - & Diabetes / Vascular disease

Perspective

1. Natural History of Disease

2. It’s a long way from genome to phenotype:

gene – translation – protein folding, synthesis & turnover – substrate availability & metabolism – tissue lay-down, degradation & repair - individual vs. commonality – social experience & behaviour

- A far cry from genome variants..

What it’s all about is regulation of

gene expression –

not the genome itself

Salt + Aldo excess & remodelling?

Catena C et al, Sechi L

- online, July 2016

PET- CT for Adrenal(micro-)adenoma imaging

Findings: Limited PET with low dose un-enhanced diagnostic CT adrenal glands.

Right Adrenal: >20mm nodule in R adrenal gland, with intense focal tracer uptake

Left Adrenal : Homogenous tracer distribution on the left; no obvious nodularity

IG: 37yo woman, West African descent;

Difficult, poorly controlled BP 3+ line therapy..

R Adrenalectomy - 9/12 later - BP 128/86mmHg on Amlodipine 5mg/d

JKC Guy’s patient – PET-CT courtesy of Morris Brown

NB:

Adrenal Aldosterone-secreting

Adenomas

are somatic, not germ-line

mutations

Beta-blockers ineffective alone for BP in

AfC’bean (& Af) patients

Cruickshank JK et al; BMJ 1988 Nov 5;297:1155-9. Cross-over RCT

Plasma Renin

AfC’bean

Europeans

BP

AfC Eurpns

AfC Eur

Sample of key papers on Ethnic Diffs in Hypert Trtmnt

literature

• Humphreys GS, Delvin DG. Ineffectiveness of propranolol in hypertensive Jamaicans.

Br Med J 1968 ; 2: 601-3.

• Kaplan NM, Kem DC, Holland OB, ..Gomez-Sanchez C. The intravenous furosemide test:

a simple way to evaluate renin responsiveness. Ann Intern Med 1976; 84: 639-45.

• Holland OB, von Kuhnert L, Campbell WB, Anderson RJ. Synergistic effect of captopril

with hydrochlorothiazide for the treatment of low-renin hypertensive black patients.

Hypertension 1983;5:235-9.

• Preston RA, Materson BJ, Reda DJ, Williams DW, Hamburger RJ, Cushman WC, Anderson

RJ. Age-race subgroup compared with renin profile as predictors of blood pressure

response to antihypertensive therapy. Veterans Affairs Coop Study Group on

Hypertensive. JAMA 1998; 280:1168-72.

Heart Failure by ALLHAT treatment

Wright J et al JAMA

2005;293:1595-1608

Systematic review: antihypertensive drug therapy in patients of African and South Asian

ethnicity

Brewster LM, van Montfrans GA, Oehlers GP, Seedat YK

Intern Emerg Med (2016) 11:355–374

32

-20 -10 -5 0 5 10 20 mm Hg

Favours treatment Favours control

(WMD 95%CI)

Effect of different drugs

on

Systolic Blood Pressure in

black people

Ca-blockers

I2 =94%

Diuretics

ACE inhibitors

AT II antagonists

Beta-blockers

-20 -10 -5 0 5 10 20 mm Hg

Favours treatment Favours control Brewster, 2004

Syst Review – RCTs in African-origin Hypertensive pts (only) vs placebo - 1

Brewster LM et al

Syst Review – RCTs in African-origin Hypertensive pts (only) vs placebo – 2, Brewster LM et al

Danish National Registry

Serum K+ and Risk of Death in Hypertension

1. Spironolactone

2. Amiloride ± HCTZ

3+.. Aldosterone synthase antagonism

- every opportunity in African origin patients

‘Simple’ Future Treatments

N. Xiao, J.D. Humphrey, C.A. Figueroa. “Computational Model of 3-D Hemodynamics -

a Full-Body Arterial Network."

Journal of Computational Physics. 2012 DOI: 10.1016/j.jcp.2012.09.016

Flow & Pressure

Variation

across the

Arterial tree..

(modelled)

NB:

Pulse Wave

Velocity

changes

(estimated)

Courtesy of

Dr A Figueroa,

King’s College

The

“Restricted

Vascular

Network”

Hypothesis

for High BP

Intergenerational transmission of CVS RISK -

Examples from Caribbean peoples

Intrauterine Social Position growth early childhood growth adolescent RISK FACTORS Adult Health Outcomes Lifespan--------------------------------------------------------/

maternal

phenotype

genes Individual adaptation Finance

Forces of HISTORY & Society

Slavery & slow escape from post-

emancipation poverty / indentured labour

Nutrition

Upstream & downstream policies to address ethnic inequalities – the heart of public health debate

BP distribution in 35-66y olds –

Kilifi, Kenya; ABPM vs Usual

Etyang, Smeeth, Cruickshank & Scott

BP distribution in 35-66y olds – Kilifi,

Kenya; ABPM vs Usual

Etyang, Smeeth, Cruickshank & Scott

Waist to height ratio by age and ethnicity for males and females: means/percentage and 95% confidence

intervals by gender and ethnicity

0.40

0.42

0.44

0.46

0.48

0.50

0.52

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

Whites UK BC BA Indian PB Others

Wai

st h

eig

ht

rati

o

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

11

-13

y

14

-16

y

21

-23

y

Whites UK BC BA Indian PB Others

Males Females

(BC – Black Caribbean, BA – Black African, PB – Pakistani/Bangladeshi)

Harding S et al.. Cruickshank JK – in preparation

The influence of waist to height ratio at 11-13y and 21-23y on bio-markers at 21-23y

• Allostatic Load: Score derived at 21-23y using high-risk thresholds defined as below the 25th percentile for HDL cholesterol, and above the 75th percentile for all other biomarkers.

• The thresholds were: systolic, 121.0 mmHg; diastolic, 77 mmHg; BMI, 26.8 kg/m2; waist to height ratio male, 0.89, and female, 0.86; total cholesterol, 4.8 mmol/L; HDL cholesterol, 1.8 mmol/L; HbA1c, 37.0 mmol/mol; C-reactive protein, 0.0 mg/L. Linear Regression Model: regression coefficients adjusted for gender, ethnicity, waist to height ratio.

** HbA1c, HDL Cholesterol and Total cholesterol: Linear Regression Model. Regression coefficients adjusted for ethnicity and waist to height ratio at 11-13y and 21-23y.

Allostatic Load HbA1c HDL Cholesterol Total Cholesterol

Coef 95% CI Coef 95% CI Coef 95% CI Coef 95% CI

Males

Waist to height ratio, 11-13y 7.69*** (4.27,11.11) 3.24 (-7.44,13.93) 1.31** (0.49,2.14) -1.47 (-3.55,0.60)

Waist to height ratio, 21-23y - - 0.27 (-9.66,10.21) -1.79*** (-2.55,-1.03) 4.51*** (2.59,6.43)

Females

Waist to height ratio, 11-13y 11.36*** (7.76,17.95) 22.54** (6.36,38.72) 0.25 (-0.90,1.40) -1.95 (-4.31,0.40)

Waist to height ratio, 21-23y - - -0.85 (-11.98,10.28) -1.24** (-2.00,-0.47) 2.93*** (1.31,4.55)

*p<0.05, **p<0.01, ***p<0.001