Sleep architecture and sleep continuity in healthy twins

There is evidence from previous sleep studies in healthy twins that some features of non-REM sleep and REM sleep are under genetic control. In particular, a high heretability was described for the amount of stage 2 and slow wave sleep and the amount of phasic REM activity (Linkowski et al., 1999). Animal data strongly support that distinct spectral patterns of both non-REM and REM sleep are genetically determined, i.e. a gene locus was found for the theta peak in REM sleep (Tafti und Franken, 2002).

Therefore we investigated the spectral composition of sleep in mono- and dicygotic healthy twins including topographical aspects in the recordings. The twins spent three consecutive nights in our sleep lab and underwent a MSLT (multiple sleep latency test) to assess the degree of daytime sleepiness. In the first step of the data analysis we determined the conventional sleep stages to obtain sleep architecture and sleep continuity parameters. In a second step the digitised EEG data will be submitted to a Fast Fourier Transformation.

Up to now we were able to include n=10 dicygotic twins (mean age 22,7±2,7 yr, range 18–27 yr, 7f: 3m) and n=26 monocygotic twins (mean age 24,3 yr, range 17–38 yr, 16f: 10m) in the statistical analysis.

We determined the intraclass correlation coefficients with respect to all sleep stages, sleep efficiency index, sleep stage latencies, cycle duration and the sleep stage transitions from REM to non-REM, non-REM to wake and light sleep to slow wave sleep. The similarity index in the latter parameter was showed the greatest differences between the mono- and dicygotic twins, which was confirmed by the variance analysis. The data suggest that there is a high genetic determination on the transition from stage 2 sleep to slow wave sleep. However, the sample size of the dicygotic twins is comparably small and has to be increased to confirm these preliminary results.