Genetic-Environmental Interaction: Implications for Osteoporosis Prevention Strategies

1
Genetic-Environmental InteractionImplications
for Osteoporosis Prevention Strategies
2
How many Australians have osteoporosis?
Current
Increased BMD by 10
x1000
x1000
3
Annual incidence of fractures in Australia?
All fractures
Hip fractures
x1000
x1000
Year
Year
4

• Can we
• predict,
• reduce,
• prevent,
• eliminate
• osteoporosis and fractures?

5
Aetiology

• Mendelian
• Chromosomal aetiology
• Multifactorial aetiology with high heritability
• Multifactorial aetiology with low heritability
• Infectious aetiology
• Environmental aetiology

6
Determinants of BMD

• Genetics Environment
• Lumbar spine 0.778 0.222
• Femoral neck 0.764 0.236
• Total body 0.786 0.214

7

• Can we use environmental factors to predict
  fracture?
• Can we use genetic factors to predict fracture?

8
Criteria

• Validity and available of tests
• Public health impact
• Magnitude of association between risk factor and
  fracture
• Interaction between known environmental factors
  and genes
• Availability of safe and efficacious treatment
• Confidentiality, ethics

9
Risk Factors for Hip Fx in Females
Risk factor Relative Risk
Prevalence AR
Osteoporosis 1 (Y/N) 7.9 (3.9 -
16.1) 0.26 0.65 Body sway (75th pct) 3.6 (1.8
- 7.0) 0.25 0.48 Previous falls (Y/N) 3.5
(1.8 - 6.9) 0.31 0.44 Any of the three factors
22.9 (3.1 - 34.7) 0.60 0.93
1 Defined by FNBMD
10
Distribution of BMD incidence of fractures
11
Familial Relative Risk of Fracture

• Intraclass correlation
• RR of BMD r0.8 r0.9
• _________________________________________________
• 5 1.14 1.16
• 6 1.17 1.20
• 7 1.21 1.24
• 8 1.24 1.28
• _________________________________________________

12
Strategies for Prevention of Osteoporosis

• Population-based strategy
• High risk strategy
• Genetic-environmental based strategy ?

13
Some Epidemiological Concepts

• Population Attributable Risk (PAF)
• proportion by which the incidence rate of disease
  in the population would be reduced if the risk
  factors were eliminated
• Positive Predictive Value (PPV)
• Risk of disease among individuals with the
  presence of a risk factor

14
General Formulation of a Screening Model

• Parameters
• Lifetime risk of fracture (d)
• Prevalence of risk factor (e)
• Relative risk of risk factor (R)
• Sensitivity, Specificity and PPV
• Sensitivity R / (1 e(R-1)
• Specificity (1-e)/(1-d) 1 - Rd/(1 e(R-1))
• PPV Rd / (1 e(R-1))

15
Effectiveness
Strategy Fx Reduction Sens Spec PPV
__________________________________________________
_____________ Population-based
1 20 0.625 0.467 0.714 High risk 2
9 0.952 0.275 0.476 ____________________________
___________________________________ Assumptions
Lifetime risk 0.4 RR 5 1 Shift the whole
distribution by 10 increase 2 Selecting only
osteoporotic subjects and increase BMD by 10
16

• What about an genetic-environmental approach ?

17
What is Gene-Environment Interaction?

• Effects of high risk genotypes vary depending on
  environmental exposure or restricted to exposed
  subjects
• Effects of environmental risk factor vary
  depending on susceptible genotypes

18
. . . more emphasis has been placed on the
concept of "effect" rather than on
"interaction". There is no reason to
believe that VDR gene would act in isolation
from other genetic and environmental
factors
19
Some Misunderstanding

• A single, simple observation of differential
  effect between genotypes of a genetic marker
  across different environmental milieu is not
  sufficient evidence for genetic-environmental
  interaction
• A statistical interaction is not necessary the
  same with a GxE interaction

20
Detection of GxE Interaction?

• Twin modelling
• Regression analysis
• Sibling interaction analysis

21
Heritability of Bone Density
Age rMZ rDZ H2 LSBMD Slemenda et
al 44 0.85 0.33 0.97 Pocock et
al 47 0.92 0.36 0.92 Nguyen et
al 50 0.74 0.43 0.78 Spector et
al 60 0.68 0.29 0.78 Flicker et
al 69 0.70 0.33 0.74
22
Formulation of G x E Models

• Parameters
• Lifetime risk of fracture (d)
• Prevalence of risk factor (e)
• Relative risk of risk factor (R)
• Prevalence of genotype (g)

Formulation Genotype Risk Prevalence RR Absence
Absence (1-g)(1-e) 1 Absence Presence (1-g)e Re
Presence Absence (1-e)g Rg Presence Presence ge
Rge
23
Models of Interaction

• Model I Re Rg 1
• Model II Re gt 1, Rg 1
• Model III Rg gt 1, Re 1
• Model IV Re gt 1, Rg gt 1

24
Effects of GxE on PPV and PAF

• g No GxE Model 1 Model 2 Model 3
  Multiplicative
• __________________________________________________
  ___________________________________
• 0.1 0.22 1.00 (0.23) 0.75 (0.12) 0.56
  (0.15) 0.82 (0.14)
• 0.15 0.23 0.89 (0.23) 0.56 (0.12) 0.27
  (0.09) 0.72 (0.18)
• 0.20 0.23 0.69 (0.23) 0.46 (0.12) 0.15
  (0.04) 0.64 (0.21)
• 0.30 0.23 0.50 (0.23) 0.37 (0.13) 0.06
  (0.02) 0.52 (0.23)
• 0.40 0.23 0.40 (0.23) 0.32 (0.13) 0.03
  (0.01) 0.44 (0.25)
• d0.15, R 2, e 0.30

25
Summary

• For a RR2 or 3, low PPV and PAF
• Introduction of GxE increases PPV, but decreases
  PAF
• High prevalence of susceptible genotype increases
  PAF, but decreases PPV

26
Future Directions

• Description of osteoporosis/fx in populationgene
  frequencies, prevalence of risk factors
• Determinants of osteo/fx in population risk
  factors, genetic markers, population genetics.
• Determination of osteo/fx in families familial
  aggregation, heritability studies, segregation
  studies
• Gene environmental studies

27
Future Directions

• Natural history of osteporosis
• Intervention clinical trials, genetic
  differences in response to treatments
• Prevention screening, counselling, carrier
  detection
• Impact of osteoporosis mortality, morbidity, QoL