Twin study: genes matter little after age 50

The research record

This is the project's file for this landmark, as written and checked. Sources are listed at the end; anything not yet confirmed is marked "Not yet verified".

Who

Caroline Tanner, Ruth Ottman, Samuel Goldman, J. Ellenberg, P. Chan, R. Mayeux and J. William Langston (corresponding author at The Parkinson's Institute, Sunnyvale, California). JAMA 281(4):341–346, 1999. Funders were not read.

What happened

Twins share their upbringing. Identical (monozygotic) twins share all their genes. Fraternal (dizygotic) twins share about half. If genes drive a disease, identical twins should both get it more often than fraternal twins.

The team screened 19,842 white male twins in the National Academy of Sciences/National Research Council World War II Veteran Twins Registry (zygosity set by PCR or questionnaire). They found 193 twins with Parkinson's (prevalence 8.67 per 1,000 after adjusting for concordance). In 71 identical and 90 fraternal pairs with complete diagnoses, both twins were affected in 15.5% of identical pairs and 11.1% of fraternal pairs (0.129 overall; relative risk 1.39; 95% confidence interval 0.63 to 3.1, so not significant). When one twin was diagnosed at age 50 or younger (16 pairs), all 4 identical pairs were both affected, against 16.7% of fraternal pairs (relative risk 6.0; 1.69 to 21.26).

Why it moved us toward a cure

The result split Parkinson's into two stories. Early-onset disease looked strongly genetic, which fit the families with SNCA (landmark 4) and parkin mutations. Typical late-onset disease looked mostly non-genetic in this sample, which fed interest in toxins such as MPTP (landmark 2) and pesticides. Later work found real but modest genetic risk in ordinary cases, as the GWAS landmarks show (landmark 237). A longitudinal study of the Swedish Twin Registry (46,436 people) later put heritability at 34%, and its authors blamed the earlier null results mainly on too little follow-up time to catch both twins of a pair.

Outcome

Advanced as a framing study. Its "no genes after 50" conclusion has been revised: the Swedish longitudinal twin study estimated heritability at 34% (11% concordance in identical versus 4% in same-sex fraternal pairs), and common variants together explain a share of risk, though most single variants are small.

Sources

• Tanner et al., JAMA 1999, DOI 10.1001/jama.281.4.341, PubMed 9929087
• Wirdefeldt et al., Neurobiol Aging 2011;32:1923.e1–8, DOI 10.1016/j.neurobiolaging.2011.02.017 (PMC4452894)
• Not yet verified: funders; the SNP-based heritability range that landmark 19 gives as 16–36% (not re-checked here).