## Measurement of Differences:

Difference can be measured and expressed in a number of ways:

1. Trait Measurement:

Differences can be measured and expressed in a number of ways. For traits which can be categorized in discrete qualitative ways twins can be compared of whether they both show the particular form of a character, in which case they are said to be ‘concordant’ or whether they possess alternative forms and are ‘discordant’.

Such comparisons are particularly useful in looking for heritable elements in disease in which one examines concordance rates for the disease in monozygotic twins and compares them with concordance rates in dizygotic twins.

It is the magnitude of this difference which indicates the extent of genetic susceptibility. One finds high concordance rates for an infectious disease like measles in monozygotic twins, but practically as high ones for dizygotic twins. All concordance means here is that if one twin is infected the second is likely to be infected too.

However, in the case of rickets or poliomyelitis whilst the concordance rate of monozygotic twins is substantially lower than for measles it is very much higher than the concordance rate for the same diseases in dizygotic twins.

This indicates a strong genetic susceptibility to these diseases even though one is essentially of nutritional origin and the other due to an infective agent.

2. Character Measurement:

For characters which are measured on a continuous quantitative scale the differences between pairs of twins can be expressed in such ways as the correlation coefficient (the magnitude of the correlation of the dimension of one twin with the other over all twin pairs of a particular type) or as the within-pair variance averaged over all twin pairs of a particular type.

The usual approach for estimating heritability is to express the difference in these measures between dizygotic and monozygotic twins and express it as a proportion of the variation between dizygotic twins, i.e.

The within-pair variance of dizygotic (fraternal) twins and is the within-pair variance of monozygotic (identical) twins. This is formally equivalent to the concordance rate comparison for discrete traits. Clearly, if V; is low compared to the ratio approaches 1 (or 100 per cent) and one would conclude that variation the trait being considered is largely under genetic causation, i.e. is strongly heritable. Conversely, if V; is practically as large as the ratio approaches 0 and trait variation is mainly under environmental control, i.e. has low heritability.

More formally heritability

VT VG + VE where VG equals the genetic component to the variance; VE the environmental component and VT the total phenotypic variability as measured V} is the estimate of VE and VF of VG + VE.

The whole analysis is obviously a within-family one, comparing the magnitude of within- family genetic variation to the kinds of diversity of environmental factors that occur within the confines of a single family (such as heterogeneity within homes).

The genetic variation between dizygotic twins (as sibs) also bears a relationship to the genetic variation in the population as a whole. Sibs share one-half of their genes in common, but the remaining half is no more alike than unrelated people. Genetically, therefore, one might consider dizygotic twins as half less variable than unrelated pairs in which case If = Vic + VE. With this viewpoint Vf– V1 = (V2VG + VE) – VE = View. With heritability equaling VG/VT we need to double the difference in the variance of fraternal and identical twins to obtain our estimate, i.e. 2(Vf – V1)/VT and make a direct measure of the overall variability of the trait in the population VT.

3. Measurement of Environmental Differences:

Occasionally, monozygotic twins are separate at birth or shortly afterwards and are brought up in different homes and therefore often via different environments. Comparing the magnitude of the difference between such twin 4 reared apart (A) and monozygotic twins reared together (T) can provide striking insight into the environmental liability of a character over a population.

Thus, if using the formal (V1A – V1T)/V1A one obtains a proportion close to 1, the within-pair variance of monozygotic ‘ twins reared together is very small compared with that of those read together is very string compared with that of those reared apart, and we conclude that between-family environment factors play an important part in determining character variability.

Conversely, a proportion y- approaching 0 signifies high heritability in the population. V1A is, of course, also a use parameter for indicating environmental contributions in the dizygotic twin/monozygotf twin formulae.