Are elevated cerebrospinal fluid levels of IL-6 in sudden unexplained deaths, infectious deaths and deaths due to heart/lung disease in infants and children due to hypoxia? A ˚ Vege 1 , TO Rognum 1 , AO Aasen 2 , OD Saugstad 3 Institute of Forensic Medicine 1 , Institute of Surgical Research 2 and Institute of Paediatric Research 3 , Rikshospitalet, Oslo, Norway Vege A ˚ , Rognum TO, Aasen AO, Saugstad OD. Are elevated cerebrospinal fluid levels of IL-6 in sudden unexplained deaths, infectious deaths and deaths due to heart/lung disease in infants and children due to hypoxia? Acta Pædiatr 1998; 87: 819–24. Stockholm. ISSN 0803–5253 Many SIDS cases probably die after periods of hypoxia and it has been shown that hypoxia may stimulate IL- 6 production. The purpose of this paper was to examine if there were any correlations between hypoxanthine in vitreous humour and Il-6 in CSF. The concentration of IL-6, IL-1b and TNFa in cerebrospinal fluid of 50 Sudden Infant Death syndrome (SIDS) cases, 9 borderline SIDS cases, 18 infectious deaths, 8 violent deaths and 22 cases with heart/lung disease were measured by ELISA. The hypoxanthine (Hx) vitreous humour concentrations in the same groups were determined by high performance liquid chromatography. The IL-6 levels in cases of infectious death, heart/lung disease and borderline cases were significantly higher than in the SIDS cases ( p  0:01). The Hx levels were in the same range in cases of SIDS, borderline SIDS and infectious death, and they were significantly higher than the levels in cases of violent death and heart/lung disease ( p  0:01). There was no correlation between hypoxanthine and IL-6 in any of the groups. In the cases studied IL-6 elevation is probably not induced by hypoxia, but is rather a result of immunological stimulation.  Hypoxanthine, hypoxia, IL-6, IL-1b, TNFa, SIDS A ˚ Vege, Institute of Forensic Medicine, National Hospital, 0027 Oslo, Norway Many Sudden Infant Death Syndrome (SIDS) victims have signs of infection prior to death (1). A significant propor- tion of SIDS victims also shows immune stimulation in the respiratory and gastrointestinal tract, which resembles observations in infectious death in infancy (2–4). Recently, we have also demonstrated that half of the SIDS victims have Interleukin-6 (IL-6) levels in the cerebrospinal fluid (CSF) similar to those found in meningitis and septicaemia (3). In parallel with the increase in knowledge about SIDS, the concept of borderline SIDS has been introduced (5). Borderline SIDS cases are cases in which there is positive information either in the clinical history, or about the circumstances of death, or pathological findings at the autopsy, but it is insufficient to explain the cause of death (5). In most of the borderline cases, morphological and/or microbiological signs of slight infections are the reason for not using the term SIDS. Elevated levels of the cytokines IL-6, interleukin-1b (IL-1b) and Tumour Necrosis Factor-a (TNFa) have been measured in blood and CSF in bacterial and non-bacterial meningitis and sepsis (6–8) and IL-6 production is inde- pendent of age and sex (3, 9). IL-6, IL-1b and TNFa are small proteins that are produced by a variety of cells, such as lymphocytes, macrophages, monocytes, endothelial cells, keratinocytes, astrocytes and microglia (10, 11). The cytokines are released as a response to bacteria and viruses, as well as many other factors (10–12). Depend- ing on the clinical condition they are released locally in the tissues, in the blood or CSF. IL-1 can induce increased slow- wave sleep (SWS) when injected into the cerebral ventricles of animals (13). Increased SWS can also be due to increased body temperature (13), for instance induced by IL-6, which is reported to be an endogenous pyrogen (10). In addition to the above mentioned stimuli to IL-6 liberation it has also been claimed that hypoxia can induce IL-6 production (14, 15). It has previously been shown that many SIDS victims die after periods of hypoxia (16). We therefore wanted to examine whether there was a relationship between IL-6 and hypoxia, as estimated by vitreous humour hypox- anthine (Hx) levels. The immunomediators IL-1b and TNFa were also included in the present study and special interest was taken in examining the borderline SIDS group, to see if this could bring us closer to understanding the death mechanism in SIDS. A minor part of the IL-6 values, and less than half of the hypoxanthine levels, has been reported in previous studies (3, 17). However, in the present paper the results are used in a different approach, as the aim of this study was to look for a possible relationship between hypoxia and IL-6 release in the central nervous system (CNS). Materials and methods Subjects Cerebrospinal fluid and vitreous humour from 46 SIDS Acta Pædiatr 87: 819–24. 1998  Scandinavian University Press 1998. ISSN 0803-5253