2 7 . 2 2 6.2 pg/ml, r,=0.402, p<0.05). SPLI 38.327.7, r,=0.229) this correlation could not be found (Fig. 3).Thus prematurity must be considered as a decisive factor for the altered plasma levels of P-EDLI and SPLI also if the age is corrected for a gestational age of 40 weeks. This may indicate the immature state of development of preterms in comparison with full-terms of the same age.

The prospective study presented conducted to prevent SIDS (sudden infant death syndrome). One of the proposed hypotheses on SIDS postulates a brainstem abnormality in the neuroregulation of cardiorespiratory processes. Therefore we characterized cardiorespiratory control mechanisms by examining the neurotransmitter substance P in plasma and polysomnographic investigations. With respect to the probable multifactorial origin of SIDS we selected children firstly anamnestically by means of an epidemiologically evaluated pre-, peri- and postnatal risk score. We reported the results of 208 polysomnographically and biochemically examined children anamnestically selected from a group of 2500 neonates. Examinations were performed on infants aged 2-4 weeks up to 1 year. To characterize respiratory control, length and frequency of apnoeas were separately estimated by means of polysomnography in the sleep states active and quiet sleep. If there were polygraphic risk factors representing a disturbance of respiratory control, the children were prophylactically treated with aminophylline 3 x 3 mg/kg b.w. for 4 weeks. We found a significant age dependence both of the mean apnoea duration in active sleep and the substance P level in plasma in the SIDS-risk group but not in the controls. High mean apnoea duration was correlated with low substance P level in the first months of age in SIDS risk infants selected anamnestically. This may reflect a delayed maturation of respiratory control mechanisms. In this way the polysomnography and the investigation of the neuropeptide substance P may be useful for a screening method indicating wether the respiratory control mechanisms are mature or not.

Groups of children with a statistically enhanced risk for SIDS: siblings of SIDS-victims and other full-term and preterm infants with anamnestic signs for an enhanced risk of SIDS and a control group of healthy infants without anamnestic signs of risk were polysomnographically investigated in the first year of life. The mean apnoea duration (MA-value estimated by frequency and length of apnoeas greater than or equal to 3s) in the sleep states active and quiet sleep was calculated. During active sleep there is a significant inverse correlation between MA and postnatal age in the full-term and preterm SIDS-risk infants but not in the controls and siblings of SIDS-victims. In all age groups there are some children in the full-term and preterm infants with extremely enhanced MA-values. For 6 age groups of the first year of life the 90% percentile of the MA-value was calculated separately in the sleep states active and quiet sleep to select infants with an abnormal breathing pattern. This limit could be a help for the decision on therapy.

ZusammenfassungPeriodische Atmung (pA) wird als auffiilliges Atemmuster im S~iuglingsalter beschrieben. Die Studie berticksichtigt neben der reinen Musteranalyse funktionelle Aspekte der Atmungsregulation durch eine schlafphasenbezogene Bewertung. 122 S~iuglinge in 4 A1-tersgruppen (korrigiertes Alter < 0, 0 bis < 4, 4 bis < 8 und 8 bis 17 Wochen) wurden einmalig w~ihrend des Tagesschlafes polygraphisch untersucht und die H~iufigkeit dieses Atemmusters sowie dessen Struktur getrennt for den aktiven (AS) und ruhigen Schlaf (RS) bestimmt. Es wurden drei Gruppen untersucht: I Kontrollkinder, II Reifgeborene mit anamnestischen Risikofaktoren flit den pl6tzlichen Kindstod, III Frtihgeborene mit einem statistisch erh6hten Risiko ftir den pl6tzlichen Kindstod. Beztiglich der H~iufigkeit von pA bezogen auf die Anzahl der Kinder ergaben sich keine signifikanten Gruppenunterschiede, ebenso keine Altersunterschiede. Von 122 batten 38 Kinder im AS und 18 im RS pA, wobei sich beide Gruppen tiberlappten. In allen 4 Altersstufen atmeten mehr Kinder im AS periodisch. Betrachtet man den mittleren Atemausfall (MA), so war insbesondere im AS die altersabh~ingige Abnahme der pA gleichl~iufig mit der Abnahme des Gesamt-MA. Die altersbezogene Reduktion der pA war weniger deutlich, betrachtete man nur die Kinder mit pA. Kinder mit pA im RS hatten in diesem Schlafstadium fast ausschlieBlich periodisch auftretende Atempausen, im AS war der Anteil der pA am Gesamt-MA geringer. Im RS waren Einzelphasen von pA seltener als im AS, sie dauerten l~inger und setzten sich aus durchschnittlich Ringeren Einzelapnoen zusammen. Die Anzahl der Atemztige zwischen den Apnoen war weder von den Schlafstadien beeinfluBt noch lieBen sich altersabh~ingig Unterschiede erkennen. Die Ergebnisse weisen auf eine m6gliche Heterogenit~it des Musters pA im AS und RS hin.SchliisselwOrter aktiver Schlaf-ruhiger Schlaf-mittlerer Atemausfall -ApnoeAtmungsregulation -Risiko ftir den pl6tzlichen Kindstod -SID. SummaryPeriodic breathing (pb) is a special breathing pattern in infants. In 122 infants in the age groups corrected age < 0 weeks (n = 17), 0 to. < 4 weeks (n = 24), 4 to < 8 weeks (n = 48), 8 to 17 weeks (n = 33) daytime polysomnographic recordings were performed to describe the pattern of pb in the sleep states active (AS) and quiet sleep (QS). 3 groups of children were investigated: I: controls (n = 17), II: fullterm infants with SIDS risk factors (n = 71), III: preterm infants (n = 34). There was no difference of percentual frequency of pb in correlation to these groups and also in dependence on age. 38 of all 122 children showed the pattern of pb in AS, only 18 in QS. In all 4 age groups more children had pb in AS. The decrease of the pb in AS with age was similar with the decrease of the sum of all apneas (mean apnea duration-MA). Children with pb had predominantly periodic apneas in this sleep state, in AS the percentage of periodic apneas of all apneas was smaller. In quiet sleep single intervals of periodic breathing were rather seldom, they were longer a...

436 out of about 9000 infants who were born between April 1987 and November 1990 in Jena were selected by means of a risk score for a polygraphy monitoring. This score contains anamnestic factors of the mother and the infant which are characteristic of an enhanced risk for the sudden infant death syndrome (SIDS). We performed a day-time polysomnography (ECG, heart rate, respiration, EEG, EOG, EMG) with a sleep duration of at least one hour. 23 of the infants investigated demonstrated abnormal breathing patterns (increased mean apnea duration, apneas longer than 15 s, apneas with bradycardia) and were treated with aminophylline to prevent a possible SIDS event. Now we investigated the statistical relation between 26 anamnestic factors of our risk score and the need for treatment. In univariate statistical comparisons treated infants differed from untreated infants concerning the occurrence of intraventricular haemorrhages (IVH), the gestational age (GA) (preterm of fullterm) and the occurrence of hyperbilirubinemia (HB) in their anamnesis. In a multivariate analysis with all 26 factors the birth weight was the third important factor beside IVH and GA, but only factor IVH was statistically significant. A cluster analysis of the variables with respect to the separation of the two groups of infants indicates that the variables IVH, GA and HB strongly correlate. In the light of these results we can underline that abnormal breathing patterns can be expected especially in infants with IVH, short GA and low birth weight.