Ethnicity, infection and sudden infant death syndrome

Abstract

Epidemiological studies found the incidence of SIDS among Indigenous groups such as Aboriginal Australians, New Zealand Maoris and Native Americans were significantly higher than those for non-Indigenous groups within the same countries. Among other groups such as Asian families in Britain, the incidence of SIDS has been lower than among groups of European origin. Cultural and childrearing practices as well as socio-economic factors have been proposed to explain the greater risk of SIDS among Indigenous peoples; however, there are no definitive data to account for the differences observed. We addressed the differences among ethnic groups in relation to susceptibility to infection because there is evidence from studies of populations of European origin that infectious agents, particularly toxigenic bacteria might trigger the events leading to SIDS. The risk factors for SIDS parallel those for susceptibility to infections in infants, particularly respiratory tract infections which are also major health problems among Indigenous groups. Many of the risk factors identified in epidemiological studies of SIDS could affect three stages in the infectious process: (1) frequency or density of colonisation by the toxigenic species implicated in SIDS; (2) induction of temperature-sensitive toxins; (3) modulation of the inflammatory responses to infection or toxins. In this review we compare genetic, developmental and environmental risk factors for SIDS in ethnic groups with different incidences of SIDS: low (Asians in Britain); moderate (European/Caucasian); high (Aboriginal Australian). Our findings indicate: (1) the major difference was high levels of exposure to cigarette smoke among infants in the high risk groups; (2) cigarette smoke significantly reduced the anti-inflammatory cytokine interleukin-10 responses which control pro-inflammatory responses implicated in SIDS; (3) the most significant effect of cigarette smoke on reduction of IL-10 responses was observed for donors with a single nucleotide polymorphism for the IL-10 gene that is predominant among both Asian and Aboriginal populations. If genetic makeup were a major factor for susceptibility to SIDS, the incidence of these deaths should be similar for both populations. They are, however, significantly different and most likely reflect differences in maternal smoking which could affect frequency and density of colonisation of infants by potentially pathogenic bacteria and induction and control of inflammatory responses.

Sudden infant death syndrome, Indigenous populations, Aboriginal australian, Bangladeshi, Cigarette smoke, Risk factors, Inflammatory cytokine genes, Interleukin-10

1 Introduction

Before the introduction of the various campaigns to reduce the risk factors for Sudden Infant Death Syndrome (SIDS) in Australia, New Zealand, Britain and the United States, the incidence of these unexplained deaths among different ethnic groups was striking (Table 1). The incidences among Indigenous groups such as Aboriginal Australians, New Zealand Maoris and Native Americans were high [1,5–7] and have remained so despite the dramatic decline in SIDS among populations of European origin. Cultural and childrearing practices as well as socio-economic factors have been proposed to explain the greater risk of SIDS among Indigenous peoples; however, there are no definitive data to account for the differences observed. A careful review of SIDS cases in Indigenous and non-Indigenous infants found no evidence to support criticisms that the higher rates among Indigenous children were due to bias in diagnosis [8].

Table 1

Variation in the incidence of SIDS among ethnic groups within countries

Country	Ethnic group	SIDS/1000 live births	Refs.
Australia	Aboriginal	6.1	[1]
	Non-Aboriginal	1.7
United Kingdom	European	1.7	[2]
	Bangladeshi	0.3
United States	Total population	2	[3,4]
	Oriental	0.3
	Poor Afro-American	5.0
	Native American	5.9
	Alaskan Natives	6.3
New Zealand	Maori	7.4	[5]
	Non-Maori	3.6

We have addressed the differences among ethnic groups in relation to susceptibility to infection because there is evidence from studies of populations of European origin that infectious agents, particularly toxigenic bacteria might trigger the events leading to SIDS. The risk factors for SIDS parallel those for susceptibility to infections in infants, particularly respiratory tract infections (Table 2). In the United States, the campaigns to reduce the risks of SIDS have been less successful among Afro-American and Native American groups. Among Afro-American infants, it has been noted that the magnitudes of the differences in deaths due to respiratory infections were similar to those for SIDS [9].

Table 2

Identified risk factors for SIDS

Non-modifiable	Modifiable
Peak age range 2–4 months	Prone sleeping
Ethnicity	Overheating
Male gender	Cigarette smoke exposure
Night time deaths	Mild respiratory infections
	Lack of breast feeding
	Poor socio-economic conditions
	No or late immunisation

Children of Indigenous groups also have higher incidences of serious respiratory tract and ear infections. Infants in these groups are colonised earlier and more frequently by respiratory pathogens [10–12]. The genetic, developmental or environmental factors responsible have not yet been identified; however, in the high-risk groups, maternal smoking is much more prevalent than in the low-risk groups [13,14]. Poor ventilation and other factors such as dampness have also been associated with high levels of bacterial flora in the home environment of some Indigenous communities [15].

In Britain, lower socio-economic conditions are reported to be an important factor relating to the risk for SIDS [16]. Although many Asian families living in Britain are classified in lower socio-economic groups, the incidence of SIDS among Indian, Pakistani and Bangladeshi families was lower than in families of European origin. Infant deaths due to respiratory infections are also lower in these Asian groups than in families of European origin [2]. In Hong Kong where many families live in suboptimal circumstances, there is also a very low incidence of SIDS [6] indicating an ethnic or genetically determined protective mechanism against SIDS. Table 3 summarises some of the physiological, environmental and cultural differences among infants from European, Asian and Aboriginal Australian ethnic groups.

Table 3

Risk factors for SIDS among different ethnic groups

Factor	Caucasian European	Bangladeshi	Aboriginal Australian	Refs.
SIDS/1000 live births	2	0.3	6.1	[1,2]
Prone sleeping	+	−	−	[13]
Mothers who smoke (%)	25	3	75	[4,13]
IgG levels at birth	+	++	++	[17]
Bed sharing	+	+++	+++	[18]
Switch to circadian rhythm (age in weeks)	8–16	12–20	?	[19]
Breast feeding	+	+++	+++	[3,18,19]
Bacterial colonisation	+	?	+++	[10,12]

−, +, ++, +++, rare to common; ?, not known.

Virus infection might be an important predisposing co-factor in the series of events leading to death, but there is little evidence that SIDS is due to any one specific viral disease [20]. Toxigenic bacteria and/or their toxins have been identified in SIDS infants in several different countries (Table 4). Many of these bacteria express molecules that act as superantigens. The cytokines they induce help eliminate infection in the non-immune host; however, if these responses are not controlled, they can cause tissue damage or even death. The responses are those underlying the pathology of septic and toxic shock [50,51].

Table 4

Toxigenic bacteria and their toxins implicated in sudden death in infancy

Species	Toxin	Superantigen	Refs.
Staphylococccus aureus	Enterotoxins, TSST	Yes	[21–24]
Bordetella pertussis	Pertussis toxin,	No	[25–28]
	endotoxin	Yes
Haemophilus influenzae	Endotoxin	Yes	[29–31]
Clostridium perfringens	Enterotoxin A	Yes	[32,33]
Clostridium botulinum	Botulism toxin	No	[34–36]
Streptococcus pyogenes	Pyrogenic toxins A & B	Yes	[30]
Escherichia coli	Enterotoxins, verotoxins	?	[37–42]
	curlin	Yes	[43]
Streptococcus mitis	?	Yes	[44]
Helicobacter pylori	Endotoxin, vacuolating toxin, urease	Yes	[45–48]
Pneumocystis carinii	?	?	[49]

?, toxin/antigen unknown.

Among the bacterial species implicated in SIDS, Staphylococcus aureus best fits the mathematical model proposed by the common bacterial toxin hypothesis [52]. Staphylococcal toxins can kill healthy adults or older children [51,53]. Staphylococcal enterotoxin A (SEA), B (SEB), C (SEC) and the toxic shock syndrome toxin (TSST) have been identified in the tissues from over half of SIDS cases from five different countries [23,54].

2 Assessment of the risk factors in relation to susceptibility to infection

Many of the risk factors identified in epidemiological studies of SIDS (Table 5) could affect three stages in the infectious process: (1) frequency or density of colonisation by the toxigenic species implicated in SIDS; (2) induction of temperature sensitive toxins; (3) modulation of the inflammatory responses to infection or toxins.

Table 5

Risk factors for SIDS at major stages of the infection process

Risk factors for SIDS	Bacterial colonisation	Toxin induction	Inflammatory control
Prone position	+	+	?
Age range	+	?	+
Ethnic group	+	?	+
Excess of males	+	?	?
Night time death	−	−	+
Virus infection	+	?	+
Cigarette smoke	+	?	+
Overheating	−	+	+
No breast feeding	+	?	+
No/late immunisation	?	?	+

+, one or more effects; −, no effect; ?, not known.

3 Risk factors affecting colonisation

3.1 Prone sleeping position

Prone sleeping is a major risk factor for SIDS. In Norway, it was demonstrated that the most significant decrease following the campaign to discourage the prone position was among infants between 2–4 months of age who had signs of infection before death [55]. The prone sleeping position results in increased numbers of bacteria and an increase in the variety of species in nasal secretions of infants with respiratory virus infections [56].

3.2 Age range

During the 2–4 month age range in which most SIDS deaths occur, 80–90% of infants express the Lewis^a antigen. This antigen acts as one of the receptors on epithelial cells for three species of bacteria implicated in SIDS: S. aureus[57–59]; Bordetella pertussis[60]; Clostridium perfringens[61]. The proportion of infants expressing Lewis^a decreases with age. By 18–24 months, the antigen is usually found on red cells of 20–25% of children, a proportion similar to that observed in adults [62]. Lewis^a was identified in respiratory secretions from 71% of SIDS infants tested in one study [57].

S. aureus is the most common isolate from healthy infants in the 2–4 month age range [56,63]. Over half of normal infants are colonised by S. aureus during the period in which SIDS is most prevalent [56,63], and over 60% of the isolates from these children produced one or more pyrogenic toxins [64]. While S. aureus was isolated from about 56% of healthy infants 3 months of age or younger, 86% of SIDS infants in the same age range had these bacteria in their respiratory tract [63]. The toxins produced by staphylococcal isolates from SIDS infants vary in different geographic areas. The toxins produced by isolates from Scottish infants were predominantly SEB and SEC. In contrast, isolates from SIDS infants in Hungary predominantly produced SEA, and SEC was produced by only one isolate from a healthy child [24]. TSST was identified in tissues of over half of SIDS infants from Australia [25].

If these toxigenic bacteria are so common in infancy, why is SIDS not more prevalent and by what mechanisms have the prevention campaigns acted to reduce the incidence of SIDS? These questions are addressed in Section 4.

3.3 Ethnic group

Children in some Indigenous groups are colonised earlier and more heavily by respiratory pathogens than children of European origin [10–12].

3.4 Gender

In the studies by Harrison et al. [56], there was an interaction between gender and prone sleeping in that males sleeping prone, with or without infection, had significantly higher counts of Gram-positive cocci (including S. aureus) compared with females.

3.5 “Passive exposure” to cigarette smoke

The term “passive” smoking implies a lower level of exposure to cigarette smoke among infants exposed to this environmental pollutant. There is evidence, however, that some infants have as much cotinine in their body fluids as active smokers [65].

Smokers are more frequently colonized by staphylococci [66]. Buccal epithelial cells (BEC) from smokers bound significantly more S. aureus, B. pertussis, and several Gram-negative bacteria [67]. The enhanced binding was not associated with upregulation of cell surface antigens observed with virus-infected cells. Pre-treatment of cells from non-smokers with a water-soluble cigarette smoke extract (CSE) significantly enhanced bacterial binding. The enhancement was observed with CSE dilutions up to 1 in 300. Coating of mucosal surfaces by passive exposure to cigarette smoke in the child's environment might enhance attachment of a variety of bacterial species [67].

3.6 Mild upper respiratory tract infection

Many SIDS infants were reported to have a mild respiratory tract infection before death. Assessment of medical records for 31 SIDS deaths in a Canadian Aboriginal population indicated the majority had symptoms of colds, virus infections or breathing difficulties [15].

In vitro, infection with respiratory syncytial virus (RSV), influenza A or influenza B virus significantly enhanced binding of S. aureus and B. pertussis to the HEp-2 epithelial cell line [58,60,68,69]. Similar patterns were observed with some Gram-negative species identified in SIDS infants [68,69]. The changes in cell surface antigens that can act as receptors for some bacterial species could contribute to the increased binding observed [69,70]. These findings support the increased isolation of potentially pathogenic bacteria from infants with symptoms of virus infection [56].

3.7 Breast feeding

Glycoconjugates such as the Lewis^a and Lewis^b antigens present in human milk can significantly reduce binding of pathogens such as C. perfringens and S. aureus to epithelial cells. Breast milk also contains IgA which can aggregate bacteria making them easier to expel in mucus. It also contains antibodies specific for some adhesins involved in binding to epithelial cells [59,61] and during endemic periods has protective antibodies against viral infections such as RSV [71].

4 Risk factors and induction of temperature sensitive toxins

Despite common carriage of toxigenic strains of S. aureus among normal healthy infants [64], SIDS is a rare event. This is probably due to the limited temperature range in which the toxins are produced, 37–40 °C. The temperature of the nasopharynx is usually below 37 °C [72]. It is thought that some of the risk factors for SIDS can affect the temperature of the nasopharynx which could result in the permissive range for toxin production being reached. These include mild respiratory infection and prone sleeping position. Blocking a nostril with secretions during a viral infection or by bedding or clothing could impede cooling due to the passage of air over the mucosal surface. The nasal temperature in the prone, but not the upright position, was demonstrated to reach 37 °C in 5/30 (16.7%) children in whom there was no evidence of respiratory tract infection [73]. The recommendations to keep infants cool and to place them in the supine position to sleep make sense in relation to these findings.

5 Risk factors affecting induction or control of inflammation

Autopsy findings among SIDS infants found evidence for mild infection and associated inflammatory responses in SIDS infants [74]. Inflammatory responses occur in all infants in response to new infectious agents against which they have no specific active or passive immunity. If these responses are not controlled, they could contribute to several physiological responses that have been suggested to cause or contribute to the death of SIDS infants: cardiac arrhythmia; poor arousal; hypoxia; hypoglycaemia; hyperthermia; vascular collapse; anaphylaxis [54,75].

Factors that can enhance inflammatory responses include respiratory virus infections [76–80], additive or synergistic effects between bacterial toxins [81], interactions between bacterial toxins and products of cigarette smoke [82,83] and hyperthermia [84].

5.1 Risk factors enhancing inflammatory responses

5.1.1 Respiratory virus infection

Most of the studies on mechanisms involved in interactions between virus infection and bacterial toxins have been carried out in animal models. Induction of pro-inflammatory cytokines that contribute to severity of the host's responses to infectious agents or their products such as endotoxin can be enhanced by co-existing virus infection [76–80].

The paper in this volume by Blood-Siegfried et al. examined the interactions between virus infection and endotoxin in a rat pup model. The virus infection appeared to prime the animals for exaggerated responses when challenged with sublethal levels of endotoxin [80].

5.1.2 Combinations of bacterial toxins

Several toxigenic species have been identified in SIDS infants. In vitro models have demonstrated that there are additive or synergistic effects between bacterial toxins [81]. The findings of early and dense colonisation of Indigenous children by different species of potentially pathogenic bacteria needs to be investigated for the ability of components of these bacteria singly or in combination to induce inflammatory responses.

5.1.3 Cigarette smoke

In an animal model, nicotine significantly enhanced the lethal effect of bacterial toxins [82]. Studies with human monocytes found that cotinine, a metabolite of nicotine enhanced production of some inflammatory mediators. In this model system, it was also demonstrated that a water-soluble cigarette smoke extract enhanced tumor necrosis factor-α (TNF-α) responses of RSV-infected human monocytes, and it also enhanced nitric oxide production from monocytes exposed to TSST [83]. Smokers were found to have lower baseline levels of the anti-inflammatory cytokine interleukin-10 (IL-10) and lower levels of IL-10 in response to stimulation with either TSST or endotoxin [54].

5.1.4 Hyperthermia

In an infant rat model, hyperthermia significantly increased production of interleukin-6 (IL-6) but not interleukin-1β (IL-1β). In response to muramyl dipeptide (MDP) which was used as a surrogate for infection, IL-1β was significantly increased but not IL-6. MDP in combination with hyperthermia significantly increased mortality of the animals [84]. These results suggest one possible mechanism underlying the protective effect noted of the “reduce the risks of SIDS” campaigns might be due to keeping infants cool, thereby reducing some of the pro-inflammatory responses to minor infections.

5.2 The effect of risk factors on control of inflammatory responses

The risk factors for SIDS (Table 5) can also be assessed in relation to control of inflammatory responses. Three important factors to be considered are: (1) antibody levels which are at their lowest during the 2–4 month age range; (2) night time cortisol levels change dramatically during this age range in which most SIDS deaths occur; (3) genetic control of pro- and anti-inflammatory responses. The genetic influences will be addressed in Section 6.

5.2.1 Antibody levels and the protective effects of immunisation

Negligible levels of antibodies to common bacterial toxins have been detected in sera of SIDS infants compared with live healthy control infants of the same age [85]. IgA antibodies to TSST, SEC and the enterotoxin A of Cl. perfringens are present in human milk and might be important defences to neutralise the activities of the toxins before they could cross the mucosal barriers, thus providing passive protection for breast fed infants. Pasteurised cow's milk contains antibodies to the staphylococcal toxins, but infant formula preparations do not [86]. A sub-analysis of samples from a large population study of heart disease among Asian families in Britain found that compared to women of European origin, Asian women in the child bearing age range had higher levels of total IgG and IgG specific for some of the staphylococcal toxins [87,88] (Fig. 1). As a result of transplacental transfer of IgG, Asian infants might start life with higher levels of antibodies against toxigenic bacteria in their environment; however, Aboriginal Australian infants have levels of IgG at birth that are significantly higher than those found in infants of European origin [18], and the incidence of SIDS among Aboriginal infants is much greater than that among non-Aboriginal infants in Australia (Table 1). Antibodies specific for bacterial toxins present in serum of Aboriginal Australian infants have not been assessed.

Figure 1

Serum IgG levels in European and Asian women in the Newcastle (UK) Heart Study (age range 25–45 years).

Immunisation against diphtheria, pertussis and tetanus (DPT) appears to have a protective effect against SIDS [89,90]. From October 1990, immunisation for DPT was initiated at two months rather than three months of age for all British infants. In an animal model, the DPT vaccine induced antibodies to the pertussis toxin and also IgG antibodies cross-reactive with some of the pyrogenic staphylococcal toxins identified in SIDS infants [91]. Following the change in immunisation schedules, there was a significant decrease in SIDS deaths among infants over 2 months of age. The greatest reduction in SIDS deaths in Scotland was noted at 4 months of age, a pattern that might reflect a booster effect following primary immunisation at 2 months of age followed by further inoculations at 3 and 4 months [91]. Similar patterns were observed for reduction in SIDS deaths in England and Wales [92]. The protective effect of early immunisation might be due to an earlier switch to a TH1 T-cell cytokine response from the in utero dominant TH2 pattern of responses [93].

5.2.2 Cortisol levels and the peak age range for SIDS

Cortisol suppresses a broad range of inflammatory responses. During the first two months of life there is a steady decrease in plasma cortisol levels [94]. Significant changes in night-time cortisol levels occur during the period in which infants begin to exhibit adult-like physiological patterns reflecting development of circadian rhythm. Between 7–16 weeks of age, the body temperature of infants falls at night to 36.4 °C, similar to that of sleeping adults [95]. Peterson and Wailoo [19] suggested that the “immature” state prior to the physiological switch is a risk factor for SIDS because infants who remain in this stage longer share many of the risk factors with SIDS infants. In contrast to this hypothesis, Asian infants stay in the “immature” stage significantly longer than infants of European origin [19], and Asian infants in Britain have a lower incidence of SIDS than infants of European origin.

In conjunction with the change in body temperature rhythm, there is a dramatic drop in night time, but not day-time, cortisol levels the week following the temperature switch. Peak responses of TNF-α, IL-1β, IL-6 and interferon γ (IFN-γ) to infectious agents occur during late evening or early morning when cortisol levels are lowest and the time during which most SIDS deaths occur [96,97]. Levels of cortisol commensurate with those present at night time in infants following the developmental switch (<5 µgdl⁻¹) had little or no effect on inflammatory responses (IL-6 and TNF-α) elicited from human leukocytes stimulated with TSST; however, levels >10 µgdl⁻¹ found during the day or at night before the physiological changes reduced induction of the cytokines [98].

Rectal temperature and urinary cortisol excretion were measured in infants before and after immunisation for DPT and Haemophilus influenzae b. Rectal temperature increased significantly the night following immunisation indicating an inflammatory response. Infants in the “immature” developmental state had a significant increase in urinary cortisol excretion at night and the morning after immunisation. Once the mature adult-like circadian rhythm pattern had developed, immunisation no longer caused an increase in cortisol output [99].

The period during which there are low levels of night-time cortisol could be a window of vulnerability for SIDS (Fig. 2). If the drop in night-time cortisol occurs when the infant still has enough maternal antibodies to neutralise viruses or toxins or after it has developed its own antibodies, the probability of uncontrolled inflammatory responses is reduced. If the low levels of cortisol occur when the infant has low levels of protective antibodies, it might increase the risk of inflammatory responses that we postulate contribute to some SIDS deaths. Remaining in the “immature” developmental stage for a longer period would have several advantages in relation to susceptibility to infection. The high levels of cortisol to deal with pro-inflammatory responses to new infections would provide time for the infant to produce active immunity to environmental bacteria and their products or to make antibodies in response to childhood immunisations which commence at 8 weeks of age.

Figure 2

Cortisol levels of infants in relation to development of circadian rhythm.

6 Genetic control of inflammatory responses

Our studies on cytokine gene polymorphisms in three ethnic groups in which there are low (Bangladeshi), medium (European) and high (Aboriginal Australians) incidences of SIDS indicate that there are major differences in the distribution of some polymorphisms between Europeans and the other two groups [100,101].

6.1 IL-10 gene polymorphisms

The anti-inflammatory cytokine IL-10 plays an important role in control of pro-inflammatory responses. In animal models, it reduces the lethality of staphylococcal toxins [102]. Evidence from studies on a small number of SIDS infants suggested there was an excess of IL-10 polymorphisms associated with lower levels of IL-10 [103]. Although a second study by this group found additional data to support the original findings [104], another study on a larger sample from the Scandinavian survey of SIDS found no association with any IL-10 polymorphisms among the SIDS infants tested [105,106]. In contrast to the prediction from the genetic studies, our in vitro studies found baseline levels of IL-10 of SIDS parents were increased compared with those of control parents, and there were no significant differences between IL-10 responses of SIDS and control parents to either TSST or endotoxin. The most important finding in these studies was that smokers had significantly lower levels of IL-10, both baseline levels and those measured in response to toxin stimulation [54].

The (G-1082A) polymorphism in the promoter sequence of IL-10 is associated with decreased IL-10 production [106]. The proportion of individuals with the homozygous genotype (GG) prevalent among Europeans was significantly lower among both Bangladeshis and Aboriginal Australians. The homozygous variant genotype (AA) found in approximately 30% of European populations was predominant in the other two groups (Fig. 3). When the genotypes were assessed in relation to smoking, leukocytes from Europeans who were smokers with the genotypes predominant among both Bangladeshi and Aboriginal Australians (GA and AA) showed significantly lower levels of IL-10 in response to low levels of endotoxin [100]. If these responses are similar to those that occur in vivo, the differences in the lower proportions of Bangladeshi women who smoke (3%) compared with Aboriginal Australian women (75%) could be an important factor in explaining the differences in their respective SIDS rates and susceptibility to severe respiratory tract infections (Table 3). This is particularly important since it has been observed that some infants have cotinine levels equivalent to those found in active smokers [65].

Figure 3

Distribution of IL-10 gene polymorphisms (G-1082A) in European, Bangladeshi and Aboriginal Australian populations.

6.2 IL-1β polymorphisms

In an in vitro system, we found previously that parents of SIDS children had significantly higher levels of IL-1β in response to endotoxin or TSST [54]. Among populations such as Aboriginal Australians, there is a higher incidence of meningococcal disease as well as SIDS. Fatal meningococcal infections have been associated with the IL-1β (C-511T) polymorphism (TT) which results in the over-expression of IL-1β[107]. As with the results for IL-10 above, the Bangladeshi and Aboriginal Australian groups showed a significant difference in the distribution of the IL-1β (C-511T) polymorphism compared with Europeans (Fig. 4). The wild type homozygote (CC) predominant among Europeans was rare among the other two ethnic groups. Leukocytes from European subjects with the TT polymorphism who were smokers produced the highest median IL-1β responses to TSST and endotoxin; however, the numbers were too small for statistical analysis [101].

Figure 4

Distribution of IL-1β gene polymorphisms (C-511T) in European, Bangladeshi and Aboriginal Australian populations.

The soluble IL-1β receptor antagonist, IL-1Ra, is involved in non-functional binding to IL-1β which disables the interaction of IL-1β with functional IL-1 receptors present on the cell surfaces. The polymorphism in IL-1RN (T+2018C) results in increased IL-1β biological activity and enhanced pro-inflammatory responses. The IL-1RN polymorphism is also associated with increased levels of IL-1β secretion [108]; however, the mechanism responsible for this remains unknown. Differences in the distribution of this polymorphism among the three ethnic groups were not as dramatically different as for the IL-1β (C-511T) polymorphism (see Fig. 5).

Figure 5

Distribution of IL-1RN gene polymorphisms (T+2018C) in European, Bangladeshi and Aboriginal Australian populations.

6.3 Risk of inappropriately high IL-1β responses to bacterial toxins

Individuals from different ethnic groups in our studies were assigned to groups predicted to be at low, medium, or high risk of strong IL-1β responses based on their genotype for the three cytokine gene polymorphisms assessed: IL-10 (G-1082A); IL-1β (C-511T); and IL-1RN (T+2018C) [100] (Table 6). Based on assessment of the IL-1β responses of the subjects in the studies, we predicted that individuals with the combined alleles for each cytokine, GG/CC/TT genotype (coded as AA/AA/AA), were less likely to produce abnormally high levels of IL-1β in response to toxins compared to individuals with the homozygous variant genotype with the AA/TT/CC alleles (coded as aa/aa/aa) genotype.

Table 6

Designation of “low”, “medium”, and “high” risk groups based on the cytokine gene polymorphisms for IL-1β (C-511T), IL-1RN (T+2018C), IL-10 (G-1082A)

The distribution of the low, medium and high-risk genotypes between ethnic groups varied significantly (p=0.000) (Table 7). There were no differences between the non-Indigenous Australian and British Caucasian populations (p=0.35), but both were significantly different from either Bangladeshi or Aboriginal Australian populations (p=0.00). The Bangladeshi and Australian Aboriginal populations had a predominance of individuals with a high-risk combined genotype for strong IL-1β responses (~60%), and ~25% of individuals with a medium risk. In both these populations, 5–10% of individuals had genotypes that were of low risk of strong IL-1β responses.

Table 7

Distribution of the subjects into the combined genetic risk categories for high IL-1β responses in Australian (non-Indigenous), British, Bangladeshi, and Aboriginal Australians (Aboriginal)

Ethnicity	Low	Medium	High	P value
Australian	20	21	5	0.07
British	36	24	1	0.07
Bangladeshi	4	10	18	0.24
Aboriginal	6	33	80	0.24
Australian	20	21	5	0.00
Aboriginal	6	33	80	0.00
British	36	24	1	0.00
Aboriginal	6	33	80	0.00
Australian	20	21	5	0.00
Bangladeshi	4	10	18	0.00
British	36	24	1	0.00
Bangladeshi	4	10	18	0.00

In the Caucasian Australian and British populations over 90% of individuals had polymorphism combinations associated with low or moderate pro-inflammatory responses. The Australian population had ~10% of individuals with a high-risk for uncontrolled inflammatory responses, and <1% of the British control individuals had the high-risk cytokine gene profile.

6.4 Ethnic differences in cigarette smoking

If genetic make up were the main factor in susceptibility to SIDS, the incidence of these deaths should be similar for Aboriginal Australians and Bangladeshis. The significantly higher incidence of SIDS among Aboriginal Australian infants compared with Bangladeshis living in the UK is evidence against this. The data indicate that the genetic susceptibility alone is insufficient to explain the risk for SIDS, and other genetic or environmental risk factors are required. Both groups usually place infants to sleep in the supine position. Both have high levels of maternal IgG at birth. Infants in both groups are usually breast-fed and both groups have a high proportion of co-sleeping. The major difference is the proportion of women in these two communities who smoke [13,14]. If there are significant interactions between cytokine gene polymorphisms and components in cigarette smoke that could lead to higher expression of pro-inflammatory and lower levels of anti-inflammatory cytokines, these might help explain the differences in the incidence of SIDS observed for these two groups (Table 3).

Further evidence for the effect of maternal smoking comes from a study of differences in the incidence of SIDS among Native Americans. Risk factors in populations of Native American and Alaskan Native groups in which there were significant differences in incidence of SIDS were compared. Between 1984–1986 the incidence of SIDS was 4.6 per 1000 live births among Native Americans Indians and Alaskan Natives in the northern region of the United States. The incidence among Indigenous groups in the southwestern states was 1.4 per 1000 live births. There was no significant difference in the incidence of SIDS between populations of European origin in the two regions with 2.1 and 1.6 per 1000 live births in the north and southwest regions, respectively. Differences in socio-economic status, maternal age, birth weight or prenatal care were not significant among the Indigenous populations in the two areas. The differences were explained by the high prevalence of maternal smoking during pregnancy among the northern groups and Alaskan Natives but low among the southwest populations [7].

Further studies on how exposure to cigarette smoke affects inflammatory responses in infants in populations with similar genetic and socio-economic backgrounds would provide significant insights into susceptibility to SIDS.

7 Conclusions

The risk of SIDS among infants with an infection and the modifiable risk factors, prone sleeping, head covered or parental smoking, was far greater than the sum of each individual factor. “These risk factors thus modify the dangerousness of infection in infancy”[109].

SIDS is one of the most difficult areas of medical research. There are no animal models that reflect all the combinations of risk factors identified for SIDS. There are no inbred populations to control for genetic background when examining the effects of environmental factors such as cigarette smoke. Over the past 10 years there have been dramatic changes in the incidence of SIDS, but these have not been uniform in all ethnic groups. Due to the decline in SIDS deaths, large study populations are needed to have sufficient power to detect significant differences. Examination of the effects of genetic, developmental and environmental risk factors among different ethnic groups might be the key to future progress in understanding the causes of SIDS, rather than just the risk factors. Research into ethnic differences will require close co-operation among a variety of disciplines and the trust and goodwill of families in different ethnic groups. Understanding the interactions between genetic, environmental and developmental factors in infancy are crucial to solving the mystery of sudden death in an otherwise healthy infant.

Acknowledgements

This work was supported by grants from the Babes in Arms, New Staff Grant from the University of Newcastle (Australia), the Meningitis Association of Scotland, and The Gruss Bequest (UK). We are grateful to colleagues who have worked with us on the various aspects of the projects that provided the background for these studies–R. Bhopal, R.A. Elton, S.D. Essery, C. Fischbacher S.A. Gulliver, V.S. James, J.W. Keeling, D.A.C. Mackenzie, C. Meldrum, N. Molony, M.M. Ogilvie, M.W. Raza, A.T. Saadi, N. Unwin and M. White.