In an aggregation study of 312 probands with primary lymphedema who had onset before the age of 36, R. F. Dale (1985) attempted to quantify the increased risk to family members. The risk to first degree relatives was approximately 10% (relative risk=600), but increased to 20% (relative risk=1200) for children and siblings when there was also an affected parent. The risk to second degree relatives was 2.5% (relative risk=150), and the risk in both first and second degree relatives was 50% higher in both sexes if the proband was male. Dale observed a correlation within families of age of onset, and therefore considered penetrance complete once the family member was five years older than the age at diagnosis of the index patient. These analyses undoubtedly included both recessive and dominant forms of lymphedema, yielding an average risk to family members of any proband with primary lymphedema. Pedigree analysis for individual families would be likely to yield more accurate family-specific risks.
Dale excluded both autosomal recessive and X-linked inheritance because of vertical and male to male transmission in some of these families. In addition, simple autosomal dominant inheritance seemed unlikely because of the sex influence, whereas the incidence was too high in second degree relatives for multifactorial inheritance of a rare condition. The study concluded that because of reduced penetrance, estimated at 50%, and the preponderance of affected females, modified autosomal dominant inheritance provided the most plausible model (Dale, 1985). However, families with different ages of onset and different segregation patterns were grouped together, but no allowances were made for the possibility of genetic heterogeneity (different genes in different families) or different inheritance patterns affecting different families.
Isolated congenital hereditary lymphedema, now known as Nonne-Milroy Disease, was described by Milroy in 1892. The diagnostic criteria consisted of hereditary, congenital, chronic, and permanent lymphedema, which is not present above the waist, is firm but pitting to the touch, and does not affect longevity. However, exceptions such as one case of lymphedema praecox and one with spontaneous recovery were noted. The original paper described a family of 97 members with 22 affected individuals in six generations (Milroy, 1892).
Milroy attempted a follow-up investigation of this family thirty-five years later, at which time he identified thirty additional descendants in the fifth, sixth, and seventh generations. Since only two of these individuals were affected, he concluded that the disease was disappearing from the family, quite possibly by reason of attenuation of the trait through marriage with normal persons." However, a truly autosomal dominant disease would not be expected to show attenuation, and Milroy conceded that because many family members were lost to follow-up, this data was neither complete nor dependable (Milroy, 1928).
The first and only linkage study of congenital hereditary lymphedema was reported by John R. Esterly in 1965. The family he reported consisted of fifteen congenitally affected individuals in three generations, which included a reportedly affected stillborn female. Esterly's group examined all living affected family members, and a diagnosis of Milroy's Disease was confirmed. No linkage was demonstrated to red blood cell antigens, haptoglobin, phenylthiocarbamide tasting, or rate of isoniazid inactivation: some of the few polymorphic markers which were available at that time (Esterly, 1965).
In addition Esterly analyzed the 22 previously documented pedigrees for inheritance pattern, sex ratio of affected individuals, and ratio of affected to unaffected family members. Including Esterly's family, there were 152 affected people, among an unknown total number, in 23 families. Although lymphedema has been shown to affect three times more females than males in the general population, the sex ratio of all affected individuals in these families was 66 males to 81 females (5 were of unknown sex). Among the offspring of affected individuals, the ratio of affected to unaffected was 82:91, which is compatible with a simple autosomal dominant trait (Esterly, 1965). However, there were several families with reduced penetrance, and much of the data was incomplete or relied entirely on family reports. In addition, mild lymphedema is often difficult to detect, even when it is familial and congenital.
If only the families without skipped generations were included, the ratio of affected to unaffected individuals in at risk sibships changed to 50:46, and the sex ratio became 24:25. Esterly concluded that although all pedigrees were consistent with simple autosomal dominant inheritance, a second mode of inheritance may explain the excess of females in some of the families (Esterly, 1965).
Several investigators have addressed the preponderance of affected females with primary lymphedema. Some researchers have suggested the possibility of a sex-linked (X chromosome) modifier gene. Hormonal factors such as pregnancy and menstruation may also result in a greater penetrance of lymphedema praecox and tarda in females or increased expression of congenital lymphedema after menarche. However, the most plausible explanation for the excess of females with lymphedema is ascertainment bias because females may be more likely to report symptoms and seek treatment. Although three times more females are affected with primary lymphedema in the general population, this ratio drops to below two in sibships of many families with Milroy's or Meige's Disease.