Original articles (peer reviewed) Archives - Page 3 of 4

Evaluation of the SMARTACOP score for stratifying pneumonia risk in a Torres Strait Islander population

Issue: Vol 14 No 1, January 2014 – March 2014

McIver L¹, Berghofer H¹, Tongpao L¹, Kippin A¹, Buettner P² (2013)
Evaluation of the SMARTACOP score for stratifying pneumonia risk in a Torres Strait Islander population. Australian Indigenous HealthBulletin 14(1). Retrieved [access date] from http://healthbulletin.org.au/articles/evaluation-of-the-smartacop-score-for-stratifying-pneumonia-risk-in-a-torres-strait-islander-population
Continue reading →

Community stores influence the health of Aboriginal people living in the Fitzroy Valley region of the Kimberley

Issue: Vol 12 No 1, January 2012 – March 2012

Bussey C (2012)
Community stores influence the health of Aboriginal people living in the Fitzroy Valley region of the Kimberley. Australian Indigenous HealthBulletin 12(1). Retrieved [access date] from http://healthbulletin.org.au/articles/community-stores-influence-the-health-of-aboriginal-people-living-in-the-fitzroy-valley-region-of-the-kimberley/
Continue reading →

Aboriginal and Torres Strait Islander kids and bikes: socio-cultural factors and safety

Issue: Vol 7 No 2, April 2007 - June 2007

Malcolm Vick

View PDF

Abstract

Objective: The study sought to document young Indigenous people’s bike riding practices, explore cultural factors shaping those practices and consider ways those practices might put them at risk.
Methods: Ninety five Townsville Aboriginal and Torres Strait Islander school students participated in face to face interviews or focus groups, and completed quizzes and questionnaires on their bike riding behaviour, knowledge of road rules, and bike safety issues; the study was conducted in 2000.

Results: Indigenous young people are frequent bike users in a wide range of contexts. They demonstrate sound knowledge of road rules, but largely ignore many of them. Different aspects of their riding are shaped by mixes of poverty, sense of obligation, pragmatism, and style. They identify and act on safety issues only in relation to other road users’ behaviour, including normal conditions of road use (e.g., closeness of cars on busy roads), racist driver aggression, and ‘social’ danger from police.

Conclusions: Young Indigenous cyclists’ riding practices appear broadly consistent with practices documented among other social groups, and put them at risk of injury, although they widely adopt practices to minimise danger from other road users. Racism, and their visibility, makes them a target for harassment that also puts them at risk. They do not, however, understand road use in terms of risk and the study highlights complex cultural factors shaping their decision making as cyclists.

Implications: Young Indigenous cyclists’ own behaviour might be addressed by a combination of education and enforcement regarding safer riding practices, within and by Indigenous communities and families. Securing their safety also requires racist aggression by non-Indigenous road users, and police relationships with Indigenous people to be addressed.

Suggested citation: Vick M (2007) Aboriginal and Torres Strait Islander kids and bikes: socio-cultural factors and safety. Australian Indigenous HealthBulletin;7(2): Original article 1. Retrieved [access date] from http://www.healthinfonet.ecu.edu.au/html/html_bulletin/bull72/original_articles/bulletin_original_articles_vick.htm

Introduction

Aboriginal and Torres Strait Islander people are high-risk road users as demonstrated in their heavy over-representation in Australian road trauma statistics [1,2]. Further, their involvement is almost certainly under-reported [3,4]. Young people can also be seen to be especially at risk for essentially developmental reasons. There is almost no published research on safety issues for young Indigenous bike riders, although there is data on Aboriginal children’s helmet wearing [5], on levels of exposure and involvement in crashes of young cyclists more generally [6], and on the significance of Indigenous young people’s bike riding as a safety issue [4]. This paper addresses the lack of research on this issue through a small scale study in Townsville, where Indigenous people are a significant and visible minority (5.1% of total population) [7].

Much bicycle safety research focuses on the incidence, nature and severity of crashes and injuries [8,9,10,11], on contributing factors ranging from environmental and other ‘external’ factors [12,13], and the relations between the introduction, and encouragement or enforcement of protective measures (e.g., helmet legislation), cyclists’ responses to such protective measures (e.g., greater risk-taking), and the safety outcomes of those measures [14,15,16,17,18,19,20,21,22,23]. Some studies seek to explain as well as to document cyclists’ behaviour. Some focus, especially in the case of younger children, on developmental factors, such as the ways limited cognitive and motor skills and experience affect their awareness and their capacities to judge, make decisions and execute those decisions [24,25]. Others focus on social factors such as socioeconomic background, gender, parental intervention, and early experiences with alcohol [26,27,28,29,30], and on social-psychological factors such as susceptibility to peer pressure, identity and image [15]. Characteristically, reasons for behaviour remain relatively unexplored, and are attributed to presumed universal psychological characteristics of adolescents or treated very superficially [26,27,28,29]; thus, children’s explanation of their aversion to helmets because they are ‘silly, uncomfortable, or inconvenient’ [27, p.220] is accepted at face value without exploring further the array of meanings and values – the cultural contexts – which inform such judgments.

In contrast with these approaches, and to address these limitations in existing research, this study takes a sociological approach, in line with the Australian Transport Safety Board’s recommendations [3] concerning the need to ‘research historical and cultural factors influencing beliefs and perceptions about health and injury; and develop protocols for undertaking research in indigenous communities’ (p.233). In this case, the social-contextual factors include the contexts and conditions of bike riding, while more cultural factors include the meanings and values that Indigenous young people draw on in relation to bike riding and the situations that accompany it. In general terms, it builds on approaches to socio-cultural factors shaping road use developed in work by Connell [31,32], Natalier [33], Redshaw [34], Vick [35] and Walker and colleagues [36] and applies these in the particular context of Indigenous young people’s bike use.

Methods and approach

The study involved 95 Aboriginal and Torres Strait Islander students in Townsville. Participants were aged between 9 and 16 years, and comprised 36 male and 42 female primary students; and 9 male and 18 female secondary students. The study, while initiated by the author as a non-Indigenous academic researcher, drew on support from local Indigenous organizations, in keeping with established principles for research involving Indigenous communities [37,38]. The project was primarily concerned with providing a basis for more effective responses to what representatives of some key Indigenous community organisations saw as high levels of risk to their young people associated with their use of bikes. However, theorisation of the data formed an important part of generating the ‘useful knowledge’ we sought.

The project relied heavily on input from Indigenous research assistants and advisers. Broad parameters for the research were established through consultation and negotiation between the researcher, research team and community advisers. The specific methods were largely developed and wholly executed by a small team of Indigenous researchers, under leadership of an Indigenous project officer, to ensure that the research was culturally sensitive and appropriate.

The students from whom data were gathered were informed of the project through Indigenous support programs and staff in Townsville primary and secondary schools, and volunteered to take part in it. In order to protect them from any possible adverse consequences of what they had to say, and with a view to using this security to encourage them to speak freely, they were never identified by their own names, but participated under pseudonyms of their own choosing. Research ‘sessions’ were held at school during lunch break, and students’ loss of free time was compensated by provision of lunch (pizza, garlic bread and soft drink), and ‘sample bags’ of stickers and other ephemera donated by the local Department of Transport office, businesses, and professional sporting clubs.

Focus groups and semi-structured individual interviews provided the major means for eliciting qualitative data. The decision to use both forms of interview was dictated by the need to interview all students who volunteered, with limited time available in each school, but also reflected the recognition that focus groups and individual interviews are generate somewhat different data. On the one hand, focus group data may suffer from a degree of distortion of some individuals’ views as dominant voices shape a ‘consensus’ view. On the other, they, offer both a sense of the complexity of views, especially on contentious issues, and of the strength of consensus on others [39].

Both individual and group interviews sought to elicit free-ranging discussion of participants’ experiences, attitudes and values, and matters of concern. Quizzes about knowledge of road rules and their application were conducted orally as the first part of the interviews. Students were also asked to complete questionnaires, which sought details of bike use, including frequency, social context, condition of bike most usually ridden, access to helmet, and safety precautions. A quantitative account of the findings and related methodological issues has been presented elsewhere [40]. Here, I am concerned with the qualitative data for the light it sheds on broadly cultural factors shaping their bike use.

Results

Why they ride

Bike riding figured prominently in participants’ lives. They reported using their bikes for a variety of purposes including going to school, other ‘functional’ purposes (going to the shop); family related activities (‘errands’/‘doing a favour for family members’), recreational and social activities (doing things with, or visiting, friends, or simply as a recreation in itself) and for exercise. Around half claimed that bikes were their main mode of transport.

Knowing and doing

Most demonstrated a sound basic knowledge of road rules, including road signs, traffic lights, and laws governing helmet use and doubling. There were some important gaps in their knowledge, however, and most were unsure of proper procedure at roundabouts. Whatever their detailed knowledge of the road rules, they virtually all recognised that the rules applied to all road users, including themselves as bike riders.
‘Knowing’ often did not translate into ‘doing’. A substantial proportion either stated straight out that they knew but did not follow the rules, or that they did so only sometimes; others demonstrated in their answers to a range of questions about their actual road use that they did not comply. Younger children claimed to regularly obey traffic lights and stop signs, while older students said they did so ‘sometimes’ and ‘never’.

‘Doubling’ and pragmatism

An overwhelming proportion of the students said that they knew doubling was illegal. Yet almost all of them admitted to participating in doubling (two or more children on one bike). Younger ones were more often passengers than ‘doublers’; older ones did both. Both sexes were involved, although the proportion of males was higher. Many of the younger students said they did not double because their parents did not allow them to, or because they did not have the skill.

Most recognized that doubling could be dangerous, but about half of them said that it was only dangerous ‘if you don’t do it properly’. In other words, safety depended on skill.

Given established research findings about young people and risk-taking, and the widely admitted hostility between Indigenous young people and police, authority and the law, we had anticipated that both these would come into play as explanations for their behaviour. We were wrong. Instead, what came through in the students’ comments was a straightforward practicality. The following were typical of their responses to the question ‘Why do you or your friends double?’: ‘To get somebody to place to place’; ‘When I wear no shoes I get doubled’; ‘So we get to places faster’; ‘Cos they haven’t got a bike’; ‘If we don’t have transport’; and ‘Because it’s faster than walking together’. Their practicality seems to reflect both the depressed financial position of large numbers of Indigenous families, and the high value placed on personal loyalties and obligations (not to leave a friend or relative to walk, for instance).

Helmets and ‘cool’

Many of the students, especially those at primary school, claimed that they consistently wore a helmet. Most of the older ones stated that while they owned or had access to a helmet, they wore it sometimes but preferred not to. When asked why they, or others, might wear a helmet, the most common response across all ages was ‘because I have to’. Fewer, but a still substantial proportion, said they did so because of the risk of injury, or ‘to protect my head’, while the next most common response was that it was legally required, and that not to wear one was to risk drawing police attention. Very few said that peer pressure (‘my friends tell me to’) and peer practice (‘because my friends do’) influenced their decision to wear helmets.

While many said they wore their helmets more or less regularly, they clearly did not like wearing them. Some expressed this in strong but general terms, with comments such as ‘well it’s just f***ed’. Most explained it more specifically in terms of style: ‘they’re ugly’, ‘you look like a loser’, and ‘it’s not cool’. A high proportion of girls (no boys!) added comments such as ‘it messes up my hair’. Others offered practical explanations, such as that that ‘they’re hot’, or that they didn’t have access to one. Many also stated that they did not wear helmets because they did not think they were necessary: ‘I don’t expect to get hurt’.

Whereas doubling seems to be a fairly simple, practical matter for them, helmet-wearing appears to be a more complex issue. Their dislike of helmets because they are hot can be seen as pragmatic: for much of the year temperatures and humidity are oppressively and relentlessly high, and even in winter daytime temperatures are often in the high 20s. The unavailability of helmets for some can be seen as a socio-economically imposed practical factor, quite independent of any personal values related to helmet-wearing choice. Relations with authority are also important, and combine pragmatism (a practical submission to family discipline, and a fear of getting fined) with more abstract symbolic values (respect for both parents and to a significantly lesser degree, the law), although both of these lose force with the transition from childhood to adolescence, and have less effect with males than with females. As with doubling, physical safety barely enters as an issue. The crucial issue, however, is style, and their references to ‘cool’ resonate with Dawes’s [41] observations about Indigenous young people’s take up of ‘American’ youth-cultural styles. This is a reminder of the complexity of cultural factors, blending distinctive practices from within the Indigenous communities with influences made available through the mass media.

Safe riders vs good riders

Most claimed to be safe riders. The majority of those who did not, said that they simply had not thought about the issue. Almost none said they considered themselves an unsafe rider. Some aspects of their knowledge and self-reported behaviour (e.g., their knowledge of and claimed compliance with traffic signs and lights) support their claims. Other considerations, such as the incidence of doubling, and their privileging of style over safety considerations in relation to helmets, point in the opposite direction.

Their responses to questions about what constitutes a ‘good’ rider made few direct connections to safety. Many of the female interviewees talked about ‘good riding’ in terms of what might be described as ‘functional skill’: being able to control a bike, to react to unexpected circumstances, and the like. While not couched in terms of safety, these accounts of good riding can be seen as compatible with, if not sufficient for ensuring safety. But for most of the boys, good riding appears to be quite disconnected from considerations of safety. Rather, they described a range of high-level skills, coupled to notions of risk, bravado and being ‘deadly’. Indeed, the stories they told in relation to ‘good riding’ involved running red lights, jumping gutters and other manoeuvres, or performing other feats, or successfully managing unexpected gravel patches. It was not so much that safety considerations were dismissed, or subordinated to other concerns; rather, they simply did not enter the picture. At the same time, and in contrast to some accepted views about risk-taking as something valued in itself by adolescents [42], there was no sense that those interviewed valued risk for its own sake; rather, the risk was simply a by-product of pushing skill to its limits.

Police and the law

We asked about their experiences and attitudes in relation to the law and the police, carefully framing questions in neutral terms, and starting with simple factual matters such as whether they had ever been stopped, and why, before asking about their thoughts and feelings about the incident. Well over half, including a substantial minority of younger participants, reported having been stopped, mainly for not wearing helmets.
The older boys, especially, expressed a strong sense of hostility to, and contempt for, the police. It is noteworthy that with only a handful of exceptions, this was the only context where the students used ‘strong language’: for several, police were ‘F***’n cops’; others commented that they ‘hated’ police while one young boy said, ‘they are stupid and they suck’.

Their comments on particular experiences with police ranged from the emotionally charged ‘I was a bit afraid’, through ‘I felt bad’, ‘I thought I was in a alien world’, and ‘I was shame’, to the more pragmatic and strategic ‘I should of taken the back streets’. Here, consistently with the reasons they had given earlier for wearing helmets, almost half the younger ones said that being stopped had been effective in encouraging them to take up wearing their helmets (‘don’t want to be caught again’). Most of the older ones, however, suggested that the effect had been, rather, to encourage them to take less conspicuous routes or, in a few case, to take flight more quickly. A handful simply expressed contempt for the whole process: ‘who gives a f***’.

Particularly noteworthy, here, is the disconnection of the attitudes expressed in the context of immediate, face-to-face interaction with police, from those expressed elsewhere – intensity and hostility in one instance, low-key and unemotional in the other. For the most part, then, police and the laws they represent are a relatively insignificant, marginal consideration shaping their bike riding, but when they enter the picture more directly, the interactions appear highly flammable.

Respondents raised two issues which we had not included in our list of matters to raise, but which have a bearing on the connection between road safety and wider social concerns. Some of the older participants, including some of the girls, claimed that as ‘black kids’, they felt singled out for attention by police, either because they were ‘black’ per se, or because they were already ‘known’ (which some of them saw as a further result of ‘being black’). Some also related occasions when, having attracted police attention for some bike-related matter, they were then ‘asked all these dumb questions’: ‘Aren’t you so and so’s brother…’, ‘Weren’t you at…?’, and ‘What can you tell me about…?’ They placed such attention in the context of continuing tensions between ‘black’ youths (and the Indigenous community generally) and police, a view supported by a range of public comments and reports by Indigenous leaders and outside observers [43,44].

Dangerously significant others

While their responses across the various forms of data show little connection between their behaviour and considerations of safety, many of them did indicate a sense of danger from others. A very high proportion of younger respondents, for instance, said that they habitually rode on the footpath because of the danger from cars and other vehicles. Most had stories of incidents – some relatively serious – which entailed physical danger or harm: encounters with car doors, or vehicles reversing from driveways. From their accounts, it seems highly likely that such incidents are very substantially under-reported.

Among older participants there was a thread of commentary, supported by stories of particular incidents, which suggested that roads are dangerous for Aboriginal and Torres Strait Islander youth for social – racist – reasons, as distinct from what might be thought of as road-, traffic- or driving-related factors. Many spoke of being harassed verbally by drivers or other occupants of passing cars. One wrote, ‘I’ve seen drivers swerve at black kids’, and others related similar experiences.

Their comments appear to be borne out by the flurry of public discussion which followed an incident in which a young man deliberately drove his car into a 15 year old Aboriginal boy riding with a group of friends [45,46]. The incident produced further stories from members of local Indigenous communities about other incidents in which ‘white’ youths used motor cycles and motor cars to threaten and attack Indigenous young people, and claims that they formed part of ongoing racist harassment in the city [44].

Conclusions

The data suggest that for the young people who contributed to this study, bike riding involves an array of risks, from their own behaviour, the normal conditions of road use, and from the hostile behaviour of others. Equally, safety appears to be a relatively marginal concern for these young people. Even where they do recognize that there might be dangers, they see them as arising from the actions of others, overlooking the ways in which their own riding practices are unsafe. Importantly, they associate ‘good’ riding with potentially high-risk demonstrations of ‘deadly’ skill rather than with ‘safe’ riding.

The findings are broadly consistent with those of other relevant studies. In particular, the interview data are consistent with findings that crashes involving Indigenous road users in general, and Indigenous cyclists in particular, are under-reported [4,6, pp.6-7,47], that Indigenous children and adolescents are highly exposed to risk (6, p.8). They are also broadly consistent with findings of many studies in non-Australian and non-Indigenous Australian contexts that there is a widespread aversion to wearing helmets and that wearing them is related to gender, age, economic circumstance and parental pressure [15,29].

There findings here differ from existing studies in two significant ways. First, while many studies of helmet wearing identifies calculations of risk as a significant factor the data here reveal almost nothing that could be interpreted as an awareness, much less a calculation, of risk [20]. Rather, behaviour appears to be shaped either by purely pragmatic considerations, framed within a culturally informed understanding of social relationship and obligation. This attention to reasons and their cultural significance constitutes the second major difference from the findings of existing studies. Where existing literature attributes behaviour to presumed universal psychological characteristics of adolescents or accepts reasons given for behaviour at face value [27,28], the findings here indicate that behaviour is shaped through the ways situations are interpreted and meanings attached to actions. The fact that those interviewed drew on values and language which appear distinctively Indigenous (e.g., ‘shame’) and also on values and language which are clearly shared with non-Indigenous young people (‘uncool’) points to the complexity of the cultural factors involved.

Implications

The tendency to attribute danger to the action of others, to separate routine behaviours of convenience and obligation, such as doubling, from questions of safety, and their association of good riding with risk taking could usefully be taken into account in planning road safety interventions. The social contexts, including the clearly expressed sense of hostility between many of our adolescent participants and figures of ‘white’ authority suggest that it is important that such interventions be seen to emanate from within the Indigenous community itself, and that careful consideration be given to methods of policing. Their accounts of racist aggression on the roads (and other evidence of this) suggest the importance of addressing this issue at the level of both education and policing. Finally, the almost complete absence of detailed research about Indigenous young people’s bicycle riding, the incidence and severity of crashes involving young Indigenous cyclists and the significance of cultural factors in shaping the understandings underlying decision-making on the road identified here all point to the need for further research in this area.

Acknowledgements

This study was funded by a grant from the Centre for Accident Research and Road Safety, Queensland (CARRS-Q) at the Queensland University of Technology.

Indigenous people who contributed to the development and carrying out of the project included Rachel Atkinson, Deb Alvoen, Andrea Geary, Jeannie Herbert, Winnunga Koorine, Leah Morris, Jeremy Pau, Ralph Rigby and Val Stanley.

Contact details

Associate Professor Malcolm Vick, Director of Research, School of Education, James Cook University, Townsville QLD 4811, Australia, ph: (07) 4781 4229, email: Malcolm.vick@jcu.edu.au

References

1 Cercarelli L R (1994) Road Crashes Involving Aboriginal People in Western Australia. Accident Analysis and Prevention; 26(3):361-369
2 Cercarelli R, Knuiman M (2002) Trends in road injury hospitalisation rates for Aboriginal and non-Aboriginal People in Western Australia. Injury Prevention; 8(3):211-215
3 Australian Transport Safety Board (2004) Road safety in Australia: a publication commemorating World Health Day 2004. Canberra: ATSB
4 Styles T, Edmonston C (2005) Australian Indigenous Road Safety 2005 Update. Road Safety Research Report CR225. Canberra: Australian Transport Safety Bureau
5 Thomas H, Somers R L, Anderson R W G (2004) South Australian primary schools bicycle helmet usage survey. CASR Report Series CASR06. Adelaide: Centre for Automotive Safety Research.
6 Legge M, Williams H A, Cercarelli R (2006) Reported Road Crashes in Western Australia 2003. Perth: Road Safety Council of Western Australia
7 Townsville City Council (2003) The Townsville region: a social atlas 5th edition. Retrieved 19 September 2004 from http://www.townsville.qld.gov.au/about/atlas/analysis_02.asp
8 Attewell, R, Glase K, McFadden M (2000) Bicycle helmets and injury prevention. A formal review. Canberra: ATSB
9 Cook A, Sheikh A (2003) Trends in serious head injuries among English cyclists and pedestrians. Injury Prevention; 9(3):266-267
10 Thompson M, Rivara P (2001) Bicycle-related injuries. American Family Physician; 63(10):2007-14, 2017-18
11 N. Meuleners L B, Gavin A L, Cercarellia L R, Hendrie D (2003) Bicycle Crashes and Injuries in Western Australia, 1987-2000. Perth: Injury Research Centre, School of Population Health, University of Western Australia.
12 Reading R (2004) Bicycle helmet campaigns and head injuries among children: Does poverty matter? Child: Care, Health and Development; 30(1):89
13 Legg S, Laurs E, Hedderley D (2003) How safe is cycling with a schoolbag? Ergonomics; 46(8):859-869
14 Adams J, Hilman M (2001) Bicycle helmets. Risk taking is influenced by people’s perception of safety and danger. British Medical Journal; 322(7293):1063
15 Finch C F (1996) Teenagers’ attitudes towards bicycle helmets three years after the introduction of mandatory wearing. Injury Prevention; 2(2):126-130
16 Grant D, Rutner S (2004) The effect of bicycle helmet legislation on bicycling fatalities. Journal of Policy Analysis and Management; 23(3):595-612
17 Ivett L (2001) The role of parents/carers in the road safety education of children and youth. Retrieved 3 October 2002 from http://www.Officeofroadsafety.wa.gov.au/Facts/papers_2001/paper_8/paper8.pdf
18 Karkhaneh M, Kalenga J-C, Hagel B E, Rowe B H (2006) Effectiveness of bicycle helmet legislation to increase helmet use: a systematic review. Injury Prevention; 12(2):76-82
19 Kendrick D, Royal, S (2004) Cycle helmet ownership and use; a cluster randomised controlled trial in primary school children in deprived areas. Archives of Disease in Childhood; 89(4):330-335
20 Lardelli-Claret P, de Dios Luna-del-Castillo J, Jiménez-Moleón J, García-Martín M, et al. (2003) Risk compensation theory and voluntary helmet use by cyclists in Spain. Injury Prevention; 9(2):128–132
21 Parkinson G, Hike K (2003) Bicycle helmet assessment during well visits reveals severe shortcomings in condition and fit. Pediatrics; 122(2):320-3
22 Reading R (2003) The efficacy of bicycle helmets against brain injury. Child: Care, Health and Development; 29(3):226
23 Thompson D, Thompson R, Rivara, F (2001) Risk compensation theory should be subject to systematic reviews of the scientific evidence. Injury Prevention; 7(2):86-88
24 Ampofo-Boateng K, Thomson J (1991) Children’s perception of safety and danger on the road. British Journal of Psychology; 82(4):487-505
25 Maring W, van Schagen I (1990) Age dependence of attitudes and knowledge in cyclists. Accident Analysis and Prevention; 22(2):127-136
26 Berg P, Westerling R (2001) Bicycle helmet use among schoolchildren—the influence of parental involvement and children’s attitudes. Injury Prevention; 7(3):218-222
27 Khambalia A, MacArthur C, Parkin P C (2005) Peer and adult companion helmet use is associated with bicycle helmet use by children. Pediatrics; 116(4):939-942
28 Klein K S, Thompson D, Scheidt P C, Overpeck M D, Gross L A (2005) Factors associated with bicycle helmet use among young adolescents in a multinational sample. Injury Prevention; 11(5):288-293
29 Macpherson A K, Macarthur C, To T M, Chipman M L, Wright J G, Parkin P C (2006) Economic disparity in bicycle helmet use by children six years after the introduction of legislation. Injury Prevention; 12(4):231-235.
30 Parliamentary Travelsafe Committee (2003) Reducing the road toll for young Queenslanders – is education enough? Report No. 40. Brisbane: Legislative Assembly of Queensland
31 Connell R (1995) Masculinities. St Leonards, NSW, Allen and Unwin
32 Connell R (2001) Studying men and masculinity. Resources for Feminist Research; 28(1-2): 43-57
33 Natalier K (2001) Motorcyclists’ interpretations of risk and hazard. Journal of Sociology; 37(1):65-80
34 Redshaw S (2001) Changing driving behaviour a cultural approach. Australian Journal of Social Issues;36(4):315-331
35 Vick M (2003) Danger on the roads! Masculinity, the car, and safety. Youth Studies Australia; 21(1):32-37
36 Walker L, Butland D, Connell R (2000) Boys on the road: masculinities, car culture, and road safety education. Journal of Men’s Studies; 8(2):153-169
37 National Medical and Health Research Council (2003) Values and ethics: guidelines for ethical conduct in Aboriginal and Torres Strait Islander health research.Canberra: Commonwealth of Australia
38 Australian Institute of Aboriginal and Torres Strait Islander Studies (2000) Guidelines for ethical research in Indigenous studies. Canberra: AIATSIS.
39 Litosseliti L (2003). Using Focus Groups in Research. Continuum: New York
40 Vick M (2005) Indigenous youth, bikes and safety. Journal of the Australasian College of Road Safety; 16(1):29-35
41 Dawes G (1998) Break on through: Indigenous youth subcultures and education. Townsville: Centre for Social Research, James Cook University
42 Bradley G, Matsukis P (2000) Identity status and youth recklessness. Youth Studies Australia; 19(4):48-53
43 Ingram K (2003) Indigenous people ‘shot with flare guns, bashed’. Townsville Bulletin; 27 June:6
44 Seeney H (2003) Racial tensions rising. Townsville Bulletin; 13 June:1
45 Seeney H (2003) Cyclist, 15, killed. Man to front court on driving charge. Townsville Bulletin; 9 June:3
46 Stockdale K (2004) Lives ruined. The Wyles death case – what the judge said. Townsville Bulletin; 27 March:1-3

Are boys more vulnerable to nutritional stress? Child weight growth in a Queensland Aboriginal community (1950-1982) in comparison with the new WHO references

Issue: Vol 6 No 3, July 2006 - September 2006

Hilary Bambrick

View PDF

Abstract

Background: Previous studies have universally found that Aboriginal children exhibit poor growth in relation to international references.

Objective: To determine how weight-for-age growth of children born 1950-1982 in a large Queensland Aboriginal community compare with the recent World Health Organization (WHO) international reference data, and whether girls and boys show similar patterns of growth.

Methods: Weights were obtained from clinic records for 109 children (birth to 60 months, mean=135.4 measurements). Percentiles were fitted and smoothed using cubic B-splines. Growth of girls and boys were compared with the WHO references.

Results: Girls’ growth approximated WHO references, while boys’ growth was generally reduced. The heaviest boys were significantly heavier than the comparison data.

Conclusions: Generalisability of the data is limited, but they suggest that young boys may be more vulnerable to sub-optimal environmental circumstances than young girls. Community growth percentiles, while not necessarily appropriate for clinical diagnosis, provide useful comparison with international data and can illustrate variation at population level.

Implications: Previous studies have found that Aboriginal children exhibit ‘poor’ growth overall, but this may not be the case among girls when more recent references are used for comparison. Poor growth may be apparent only among boys, perhaps reflecting greater vulnerability to nutritional and other stress. The greater variability in boys’ growth and the sex differences in growth potential should further be explored in light of adult mortality differences. These findings may be cautiously generalisable to similar communities, and perhaps useful as a descriptive baseline against which to assess future improvements in Aboriginal child health.

Suggested citation: Bambrick H (2006) Are boys more vulnerable to nutritional stress? Child weight growth in a Queensland Aboriginal community (1950-1982) in comparison with the new WHO references. Australian Indigenous HealthBulletin;6(3): Original article 1. Retrieved [access date] from
http://www.healthinfonet.ecu.edu.au/html/html_bulletin/bull_63/original_articles/bulletin_original_articles_bambrick.htm

Introduction

There has long been debate about the appropriate use of growth reference data, whether at the individual or population level. Previous ‘international’ growth references, in particular, have been criticised for being unrepresentative. The once widely-used National Center for Health Statistics (NCHS) 1977 reference, for example, relied on data gathered from children in ‘white’, middle class families living in privileged areas of the United States (US), who tended to be formula-fed rather than breastfed, and were unlikely to represent ‘healthy’ growth [1]. References used more recently were those from the US Centers for Disease Control and Prevention, which were based on the growth of both breast-fed and formula-fed infants, in proportion to feeding practices in the US [2]. However, the questionable value of using growth data from a country with the highest – and rising – rates of obesity remained.

In April 2006, following the recognition that exclusive breast-feeding promotes satisfactory growth in infants until six months, the World Health Organization (WHO) released updated growth references [3]. These are based on an international sample of infants exclusively breast-fed for six months, with breastfeeding continuing as solids are introduced. This is considered to represent the healthy biological norm for human babies [3]. These references specifically incorporate data from ethnically diverse groups, and are intended to be representative of growth patterns for many different populations.

Charting Aboriginal child growth

There are currently no detailed Australian growth reference data, and none for Indigenous Australians (but see Smith et al. [4], who have created a small set of data).

The NCHS 1977 references (and earlier, similarly biased data) were used for many years to assess growth patterns among Australia’s Indigenous children. Jose and Welch [5] and Cox [6] were among the first to document significant faltering in growth among Aboriginal children. Such a flattening of the growth trajectory can occur as supplementary feeding and weaning introduce new challenges to infant growth, and is commonly observed in poorer countries and in disadvantaged communities in rich countries [7, 8]. The period of slow growth is usually partially compensated for by catch-up growth in the following one to two years. This pattern of faltering in both height and weight growth, usually between three and 12 months of age, was subsequently observed to occur in a number of different Aboriginal populations throughout Australia [4, 9–13].

There is no doubt that the health of Australia’s Indigenous children remains much worse than that of the rest of the population, with infant mortality still between two and four times higher [14]. Is this health differential reflected in patterns of growth? The earlier observations of typical faltering were made in relation to now superseded growth references. This paper examines whether such faltering remains apparent when child growth is compared with the more recent and more representative WHO references, and whether it occurs equally for girls and boys.

Methods

Study population

Between 1950 and 1982, an infant health clinic operated in a large, urbanised Aboriginal community located in rural south east Queensland (population currently about 1,200). All infants and children resident in the community attended the clinic for regular weighing and health checks by clinic nurses until they reached ‘school age’ (about 5 years old).

The data used in this study were limited to individuals still resident on the community in 2000. (Potential biases within the sample are discussed later.) Infant growth records were accessed as part of a study into type 2 diabetes. Recruitment into the study was by two methods: diabetes diagnosis recorded at the community hospital (‘cases’) and by random household sampling of people who had never been diagnosed with diabetes (‘controls’). This produced a total of 111 women and 93 men; 43% of whom had been diagnosed with diabetes). Details of recruitment methods are provided elsewhere [15].

Data were fortnightly weight measurements from birth to five years. These data have been validated elsewhere [16]. Other measures of infant health and growth, such as length and head circumference, were not routinely measured.

The study was approved by the Community Council, the Elders and the Human Research Ethics Committee at the Australian National University, Canberra. Participants gave their informed consent for their infant health records to be accessed.

Sampling

Adult diabetes status is not necessarily independent of child growth patterns (see discussion below). It was therefore important that the sample used to calculate the percentiles was as representative of the community as possible in terms of subsequent diabetes status, rather than biased towards children who subsequently developed diabetes as adults. To avoid over-representation of those children who were later diagnosed with diabetes, only a random sample of the diagnosed participants (43% of the original sample) were included in the percentile calculations to constitute approximately 20% of the study sample, equivalent to the proportion of adults within this community with diagnosed diabetes [15]. Child weight growth records were available for 40 women and 47 men participating in the study who had never been diagnosed with diabetes. Added to these were 10 women with diagnosed diabetes and 12 men with diagnosed diabetes (randomly selected from the original sample) to comprise approximately 20% of the final study sample. A total of 50 females and 59 males were included in the analyses.

Analyses

The date of measurement was transformed to months from birth (where a month had 30 days). Percentiles for each month were calculated separately for boys and girls from birth to age 60 months using observations lying within the month +/-0.5.

Construction of the community growth curves was informed by the methods used by the WHO: their reference curves were generated using cubic B-splines to closely represent the empirical data, along with LMS methods to account for any skewing of data [3].

For the community data presented here, cubic B-splines were chosen to generate the percentile curves as they were found to demonstrate adequate goodness-of-fit with the data, while varying the degrees of freedom was the method employed to deal with any skewness in the raw data. The community data were smoothed using cubic B-splines with 13, 15, 17, 20, 17, 15, 15, and 13 degrees of freedom respectively. This method produced curves that were visually smooth, yet captured periods of more rapid or slower growth. In order to produce a visually smooth curve, fewer degrees of freedom were used for the more extreme percentiles as there is more uncertainty (higher standard error) associated with these than with more central percentiles. The differences between the results produced by the WHO method and the one used here are likely to be negligible.

To quantify differences between the weight growth of girls and boys in this community, an index was calculated by expressing the community percentile at each age as a ratio of its corresponding WHO reference percentile. For example, for the 5 th percentile at 0 months the index for girls was 1.08, or 8% above the WHO 5th percentile (2.7kg/2.5kg), and for boys it was 1.04, or 4% (2.7kg/2.6kg). The mean differences in the resulting indices between girls and boys for each of the 5th, 50th and 95th percentiles were analysed so that the relative growth of girls and boys could be compared.

Results

On average there were 135.4 weights for each child, measured over the five years (that is, at approximately fortnightly intervals). There were between two and 245 observations per girl and between one and 251 observations per boy. The number of measurements available for girls and boys were fairly evenly distributed across the ages. Tables A1 and A2 (view Appendix) show the calculated growth curve percentiles (5th, 10th, 25th, 50th 75th, 90th 95th), from birth to 60 months for girls and boys respectively.

Comparisons with the WHO reference

Figures 1 and 2 illustrate the community 5th, 50th and 95th percentiles in relation to the new WHO 2006 references.

Figure 1. Calculated community growth curve (girls) from birth to 60 months in relation to the 2006 WHO reference, showing 5th, 50th, and 95th percentiles.

Figure 2. Calculated community growth curve (boys) from birth to 60 months in relation to the 2006 WHO reference, showing 5th, 50th, and 95th percentiles.

The community percentiles for girls match very well the WHO references with only slightly reduced growth for the few months, followed by some slightly more rapid growth from 12 months. Overall, girls tracked the WHO reference fairly closely, with some relative decline among the heaviest girls from about three years of age.

Weight growth among boys appears more variable. The heaviest boys (95th percentile) tracked the WHO reference well for the first 12 months, and then exceeded the reference until about three years. The lightest boys (5th percentile) had slower growth than the 5th percentile of the WHO reference for the first 18 months, and then followed the reference closely. The community median for boys was lower than the WHO median until about 36 months.

On average for the whole period from birth to five years, girls at the 5th percentile were 255g heavier than the WHO reference (95% CI =173-337g, p<0.001) while those at the 95th percentile weighed on average 479g less (95% CI = 340-617g, p<0.001). There were no differences for girls between the two medians (95% CI = -78-57g, p=0.759). For boys, those at the 5th percentile weighed an average of 166g less than the WHO reference (95% CI = 81-251g, p<0.001) while the heaviest boys were on average 319g heavier (95% CI = 208-430g, p<0.001). The community median for boys was on average 234g lighter than the median WHO reference (95% CI =157-310g, p<0.001).

Comparisons between girls and boys in the community

The means and standard deviations of the average indices of community growth relative to the WHO reference percentiles are shown in Table 1. These differences in weight growth between girls and boys in the community in relation to their respective WHO references were highly significant. At the 5 th percentile, the relative growth of girls was on average 4.5% greater than boys (95% CI = 3.8-5.2%, p<0.001) and at the median it was 2.6% greater than boys (95% CI = 1.7-3.5%, p<0.001). At the 95th percentile, boys’ growth was 4.7% greater than girls (95% CI = 4.1-5.4%, p<0.001). In summary, the lightest boys weighed less than girls relative to their WHO references, as did boys at the median compared with girls at the median, while the heaviest boys were significantly heavier than the heaviest girls relative to the WHO references.

Table 1. Means and standard deviations for the 5th, 50th and 95th community percentile relative to the WHO references

Percentile		Mean ratio relative to WHO reference	Standard deviation
5th	Girls	1.02	0.04
5th	Boys	0.97	0.05
50th	Girls	1.00	0.02
50th	Boys	0.98	0.03
95th	Girls	0.97	0.03
95th	Boys	1.02	0.03

Discussion

Growth faltering and overall poor growth in Aboriginal children has long been considered near universal. However, past findings were based on comparisons with growth references that have now been superseded. Using 32 years of infant health records from a large Queensland Aboriginal community, this study compared the weight growth of girls and boys aged from birth to five years with the latest WHO international reference data, and compared the patterns of girls’ and boys’ growth.

The growth of girls generally tracked the new reference data, while boys in this sample showed greater variability in postnatal growth patterns. In general, boys grew relatively more slowly than girls, but the heaviest boys were much heavier than the international reference and gained weight relatively more rapidly than the heaviest girls. Previous studies in Aboriginal communities have also found some sex differences in faltering patterns; for example, boys showed a greater deficit in weight-for-age than girls in a study of Western Australian Aboriginal children [10], but not in overall variability as found here, where boys’ growth tended to be at the two extremes.

The faltering that occurred among boys may simply be a continuation of a prenatal characteristic: because of the more rapid early fetal growth generally among males, they may be more susceptible to prenatal nutritional stress. There may be postnatal sex differences in energy storage: for example, McCowan and colleagues [17] found that girls were more likely to exhibit catch-up growth than boys, which may be protective in times of subsequent nutritional stress [18].

It is, of course, possible that the observed differences in weight growth between boys and girls could be due to fewer poorly growing girls surviving into adulthood (and thus being included in the study) giving the illusion that girls had ‘healthier’ growth. However, the demographic profile of the cohort suggests that this is unlikely. There were nearly 20% more adult women than men from this cohort (born 1950-1982, ages 19-51 years at the 2001 census) in the community. This, along with the demographic structure of the community at the time of the study (Figure 3), suggests that the apparent better growth of young girls compared with boys among the survivors is not an artefact of greater mortality among poorly growing girls. Sex differences in mortality and survivor bias appear instead to be in the other direction, with fewer surviving males, even at very young ages before sex differences in behaviour (particularly relating to trauma and risk-taking) would become evident. The demographic structure of the population could also reflect more men than women moving away from the community (and thus not being available to be included in the sample), but such migration is unlikely to be related to patterns of child growth. Demographic data lend support to the conclusion drawn here that boys may be more vulnerable to nutritional and other environmental stresses than girls, and that the poorer growth exhibited, even by those surviving to adulthood, could actually indicate poorer child growth overall. If, however, there had been fewer girls than boys surviving childhood, it would suggest that those with less healthy patterns of growth had died young and were therefore missed in the study.

Figure 3. Community age distribution, by sex and age group. Data from ABS Census of Population and Housing, 2001.

That boys may be more susceptible to nutritional stress may not explain satisfactorily the greater variability in growth observed among boys in this study, where the heaviest boys appear to have gained weight very rapidly. There are extremes in the community data in boys’ growth: the lightest males were significantly lighter than the reference, as were those at the median, but the heaviest boys were much heavier. This was not the case for girls, where the heaviest girls were found to be significantly lighter than the heaviest WHO reference. This sex difference is particularly interesting.

Maternal diabetes during pregnancy tends to produce high birthweight babies (>4,500g) and heavier infants, but this would be expected to affect male infants and female infants equally. It is possible, however, that there was higher mortality (either pre- or postnatally) among girls of diabetic mothers.

Slow and rapid infant growth have both been implicated in the subsequent development of type 2 diabetes. The ‘programming hypothesis’ suggests that babies who are subject to nutritional deprivation prenatally or in early postnatal life are ‘programmed’ to develop insulin resistance (related to diabetes) if there is subsequent over-nutrition [19, 20], while large babies resulting from maternal diabetes during pregnancy are also at increased risk of developing diabetes as adults [21, 22].

Men and women in this community exhibit similar levels of diagnosed diabetes [15], and, as similar proportions of women and men with diagnosed diabetes were included in the sample, the differences between boys’ and girls’ growth are unlikely to be due to an over-representation of men with diabetes. It is expected that if there had been any diabetes-related bias in the sampling – either subsequent development of diabetes or over representation of children of mothers with diabetes – this would have occurred equally for both groups.

That faltering seems to occur even among the heaviest children suggests that a number of children are born large (perhaps due to maternal diabetes), but that postnatal environmental circumstances are less than optimal for growth. Again, there is no reason to suspect that these environmental circumstances might be different for boys and girls.

Weight-for-age is not a perfect indicator of overall health and growth. The same weight could theoretically be observed for a stunted but relatively heavy child and a lean but tall child. Weight used in conjunction with linear measurements would provide a better overall indication of growth, but, unfortunately, length measurements were not routinely taken at the infant clinics.

Growth is not a reflection only of nutritional adequacy. Infections can play a significant role in reducing growth (and also poorer growing children tend to be more susceptible to infection). Other environmental factors which either relate to infant and child growth, through nutrition, or interact with it include socioeconomic status, emotional stress, and, in some regions, season and climate [9, 23]. Again, it is unlikely that exposure to these environmental factors differed systematically by sex, so they cannot explain the greater variability among weights for boys unless the response to these factors differs by sex.

This study did not take into account any secular trends in child growth. Health outcomes for children in the community did improve over the study period: infant mortality declined from approximately 250 per thousand in 1952, through 150 in 1960s, 40 in the 1970s (still twice the rate for the rest of Queensland at the time) to approximately 16 per thousand in the 1980s [24]. It could be expected that the most vulnerable children were also the lightest, and, as infant mortality declined, more of these children would have survived and would form part of this sample. Or conversely, gains in growth may have contributed to the decline in mortality.

Indeed, postnatal growth patterns in the study community have changed slightly over the study period, but the average increase may reflect an increase in the proportion of heavier children, rather than an overall increase [25, 26]. This may be what is reflected in the weights of the heaviest boys in the sample. Changed feeding practices (Figure 4) probably had little influence on growth patterns in this study population: differences in growth between breast-fed and formula-fed infants in this community were found in a previous study to be negligible [27], probably reflecting the much stronger influences of other community environmental factors (such as exposure to infection).

Throughout the 30-year study period the living conditions in this community were generally those of substantial deprivation relative to Australia generally. State government administration of the community continued into the 1980s. Many traditional practices had been lost, children were often institutionalised in dormitories, and rations (white flour, sugar, tea and some poor quality meat) were relied on until the 1970s. The population in this study was unlikely to be well-nourished by today’s standards, but severe malnutrition was probably rare [25]. All the children included in the sample, however, had survived into adulthood, suggesting that the growth characteristics of the sample may be better than one containing children who did not survive into adulthood.

Figure 4. Changing patterns of breastfeeding in the study community, 1953-1972. Data from [27]

Adequate weight gain is not always equivalent to good health, but the WHO references form a useful point of comparison to assess overall growth patterns within a community. The community growth percentiles presented here are not intended to be considered optimal for Aboriginal children, but go some way to describe what may have been ‘normal’ under the particular circumstances of a socioeconomic disadvantaged, urbanised community.

The size of the study data set was limited by the size of the community, so there is some uncertainty, especially at the more extreme percentiles, that delivers some ‘lumpiness’ into the derived curves. These curves could have been smoothed further, but this would have compromised the accuracy of the curves as a reflection of the observed data.

Another limitation of this analysis is that comparisons between the observed and reference percentiles do not take into account the imprecision of the two sets of estimates. The WHO reference data, being based on many more measurements, are comparatively precise, while the community percentiles would have much greater variability.

Taking these limitations into consideration, comparison with the latest international weight growth reference data suggest that significant faltering was not necessarily universal among Aboriginal children, but that boys may have been more vulnerable to nutritional (and perhaps infection) stress than girls. Boys appear to have followed two extremes of growth, being either very heavy or very light in relation to the international data, while girls tracked the reference percentiles fairly closely. It is unknown what may be behind these sex differences in variability. Given that life expectancy of Australia’s Indigenous people is far lower than for the rest of its population and the relative difference between Aboriginal men and women is also substantial [14], factors leading to these sex differences in mortality may extend back into sex differences in responses to early childhood environment. Possible differences between girls and boys in achieving ‘healthy’ growth therefore need to be assessed further and addressed. These findings may be generalisable to similar communities, and may be used cautiously as a baseline against which to assess future improvements in Aboriginal child health.

Acknowledgements

The author wishes to thank the study community for their warm enthusiasm and participation, Dr Alan Dugdale for enabling the use of the infant health clinic data, and Dr Ann Cowling and Dr Mark Clements for statistical contributions. The research was partially funded by an Australian Postgraduate Award (APA) PhD Scholarship. Additional funding was provided by the Australian Institute of Aboriginal and Torres Strait Islander Studies (AIATSIS grant S6116076) and the Faculty of Arts at The Australian National University.

Further information

Contact details:
Dr Hilary J Bambrick, Research Fellow, National Centre for Epidemiology and Population Health, The Australian National University, Canberra ACT 0200, Australia, email: hilary.bambrick@anu.edu.au

Ethical approval:
The study was approved by the Community Council, the Elders and the Human Research Ethics Committee at the Australian National University, Canberra. Participants gave their informed consent for their infant health records to be accessed.

Conflict of interest:
None

Keywords: weight-for-age, child growth, nutrition, Aboriginal health, WHO growth references, sex differences

References

1 Kuczmarski RJ, Ogden CL, Guo SS, Grummer-Strawn LM, Flegal KM, et al. (2002) 2000 CDC Growth Charts for the United States: methods and development. National Center for Health Statistics
2 National Center for Health Statistics (2000) CDC Growth Charts. Retrieved Last accessed 2 May 2005 from http://www.cdc.gov/nchs/about/major/nhanes/growthcharts/datafiles.htm
3 WHO Multicentre Growth Reference Study Group (2006) WHO child growth standards: length/height-for-age, weight-for-age, weight-for-length, weight-for height and body mass index-for-age: methods and development. World Health Organization, Department of Nutrition for Health and Development
4 Smith RM, Smith PA, McKinnon M, Gracey M (2000) Birthweights and growth of infants on five Aboriginal communities. Australian and New Zealand Journal of Public Health;24(2):124-135
5 Jose DG, Welch JS (1970) Growth retardation, anaemia and infection, with malabsorption and infestation of the bowel: the syndrome of protein-calorie malnutrition in Australian Aboriginal children. Medical Journal of Australia;1:349-356
6 Cox JW (1979) Growth characteristics of preschool Aboriginal children. Australian Paediatric Journal;15:10-15
7 Allen LH (1994) Nutritional influences on linear growth. European Journal of Clinical Nutrition;48(Supplement 1):S75-S89
8 Shrimpton R, Victora CG, de Onis M, Lima RC, Blössner M, et al. (2001) Worldwide timing of growth faltering: implications for nutritional interventions. Pediatrics;107(5):61-67
9 Gracey M (1991) Nutrition of Australian Aboriginal infants and children. Journal of Paediatrics and Child Health;27:259-271
10 Gracey M, Andersen C, Brooks B (1989) Low birthweight and impaired growth to 5 years in Australian Aborigines. Australian Paediatric Journal;25:279-283
11 Henneberg M, Schilitz A, Lambert KM (2001) Assessment of the growth of children and physical status of adults in two Aboriginal communities in South Australia. American Journal of Human Biology;13:603-611
12 Paterson BA, McKinnon CP, Edmond KM (2001) A review of annual growth screening in Aboriginal schoolchildren in Australia. Journal of Paediatric and Child Health;37:18-23
13 Rousham E, Gracey M (1997) Persistent growth faltering among Aboriginal infants and young children in north-west Australia: a retrospective study from 1969 to 1993. Acta Paediatrica;86:46-50
14 Australian Institute of Health and Welfare (2004) Australia’s Health 2004. (Cat No. AUS 44) Canberra: AIHW
15 Bambrick HJ (2001) Prevalence of diagnosed and undiagnosed diabetes on a Queensland Aboriginal community. Australian Indigenous Health Bulletin;1(1)
16 Dugdale AE, Musgrave IA, Streatfield K (1990) The changing growth of Aboriginal children. Journal of Paediatrics and Child Health;26:192-196
17 McCowan L, Harding J, Barker S, Ford C (1999) Perinatal predictors of growth at six months in small for gestational age babies. Early Human Development;56:205-216
18 Kuzawa CW (1998) Adipose tissue in human infancy and childhood: an evolutionary perspective. Yearbook of Physical Anthropology;41:177-209
19 Barker D, Hales C, Fall C, Osmond C, Phipps K, et al. (1993) Type 2 (non-insulin dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to reduced fetal growth. Diabetologia;36:62-67
20 Barker DJP (1999) The fetal origins of Type 2 diabetes mellitus. Annals of Internal Medicine;130:322-324
21 Bennett PH (1999) Type 2 Diabetes among the Pima Indians of Arizona: an epidemic attributable to environmental change? Nutrition Reviews;57(5):S51-S54
22 Pettitt DJ, Knowler WC (1998) Long-term effects of the intrauterine environment, birth weight, and breast-feeding in Pima Indians. Diabetes Care;21(Supplement 2):B138-B141
23 Boddy J, Skuse D, Andrews B (2000) The developmental sequelae of nonorganic failure to thrive. Journal of Child Psychology and Psychiatry and Allied Disciplines;41(8):1003-1014
24 Dugdale AE (1996) Patterns of disease among children: a simple and versatile measure of child health in communities. Journal of Paediatric and Child Health;32:400-404
25 Dugdale AE, Muller M, Alsop-Shields L (1994) Patterns of weight growth in Aboriginal children on Queensland communities. Journal of Paediatrics and Child Health;30(1):55-58
26 Muller M, Eaton-Evans J, Dugdale AE (1984) Growth of Aboriginal infants. Medical Journal of Australia;141:228-230
27 Dugdale AE (1980) Infant feeding, growth and mortality: A 20-year study of an Australian Aboriginal community. Medical Journal of Australia;2:380-385

Relationships between BMI, waist circumference, hypertension and fasting glucose: rethinking risk factors in Indigenous diabetes

Issue: Vol 5 No 4, October 2005 - December 2005

Hilary Bambrick

View PDF

Abstract

Objective: To determine whether the body mass index (BMI) threshold defined for obesity (30kg/m²) adequately reflects risk in an Aboriginal community with a high rate of Type 2 diabetes.

Methods: Data about five diabetes risk factors (age, BMI, waist circumference (WC), hypertension and family history) and fasting glucose (FG) were obtained from a random sample of 117 Aboriginal adults (62 women and 55 men) never diagnosed with diabetes. Linear regression between BMI, WC and FG, and sensitivity and specificity analyses in predicting elevated FG and hypertension were conducted.

Results: BMI ≥30kg/m² and central obesity assessed by WC (women≥88cm; men≥102cm) were strongly and positively associated. Among women, central obesity was near universal, occurring at BMIs below the ‘healthy’ range of 20-25. WC was linearly associated with other diabetes risk factors. WC≥88cm was more sensitive but less specific than BMI≥30 in predicting elevated FG and hypertension among women, while BMI≥25 among men tended to be both more sensitive and more specific than both BMI≥30 and WC≥102cm.

Conclusions: In women, central obesity is a better predictor of diabetes and CVD risk than BMI≥30, which is not a reliable indicator. BMI≥25 was a good predictor in men.

Implications: BMI is a useful clinical tool to identify individuals at risk, but to be relevant the guidelines defining risk may need to be reduced for the Aboriginal population. For women, a BMI≥25 could more adequately reflect risk, while the current WC of 88cm remains appropriate. For men, a reduction in both BMI to ≥25 and WC to 90cm may better reflect diabetes and CVD risk.

Suggested citation: Bambrick H (2005) Relationships between BMI, waist circumference, hypertension and fasting glucose:
Rethinking risk factors in Indigenous diabetes. Australian Indigenous HealthBulletin;5(4): Original article 1. Retrieved [access date] from http://www.healthinfonet.ecu.edu.au/html/html_bulletin/bull_54/original_articles/bulletin_original_articles_bambrick.htm

Introduction

Type 2 diabetes is a serious health burden for many Australian Aboriginal communities; prevalence estimates are as high as 40% in some more urbanised populations [1, 2], more than five times the estimated rate for Australia as a whole [3]. In addition to the morbidity and higher mortality arising directly from diabetes, diabetes also contributes substantially to risk of cardiovascular disease (CVD) and renal disease, which are both significant causes of excess Aboriginal mortality [4]. Diabetes in Indigenous communities causes widespread disability, disrupts family and social life and reduces capacity for work.

There are five risk factors which are currently recognised by Australia’s National Health and Medical Research Council (NHMRC) as contributing significantly and independently to diabetes risk (Table 1) [5]. These variables are in reality continuous, but determining certain cut-off points – or thresholds – can be useful in a clinical setting for identifying individuals at risk. With the exception of age, the NHMRC risk criteria are applied uniformly across population groups in Australia.

Table 1. NHMRC guidelines for independent risk factors relating to Type 2 diabetes

Risk	Assessment criteria
Obesity	Body mass index (BMI) ≥30kg/m²
Central obesity	Waist circumference Women: waist ≥88cm; men:≥102cm
Hypertension	Systolic pressure >140mmHg and Diastolic Pressure >90mmHg
Age	≥35 years (≥55 years for non-Aboriginal Australians)
Family history	At least one first-degree relative has been diagnosed with Type 2 diabetes

Source: NHMRC (2001) [5]

There is growing evidence, however, that thresholds such as these may not confer uniform risk across different population groups, and therefore applying such definitions across different populations is inappropriate. The appropriate application of body mass index (BMI), in particular, has been questioned in different populations internationally [6–10].

Despite links between obesity and diabetes, BMI does not relate to diabetes in all ethnic groups in the same way it does in people of primarily European descent; the threshold where BMI becomes a significant health risk may be lower or higher. For example, if a BMI of 30kg/m² is applied in Japan only 3% of the population are obese, which fails to reflect the rapid adoption of a more Western, energy-dense diet and more sedentary lifestyle in the last 60 years [10]. To adequately reflect the change in BMI that has occurred and the accompanying increase in diseases, Kanazawa and colleagues have recommended that the BMI threshold for obesity in Japan be reduced from 30 to 25 [10]. Similarly in China, the relationship between diabetes and BMI becomes apparent at much lower values than those widely accepted as indicating obesity [8, 9]. In Asian populations, risk of diabetes at any given BMI may be greater than in Europeans [6].

Conversely, using BMI≥30 as a cut-off may also overestimate associated diabetes risk in some other populations. BMI has been found to be poorly predictive of cardiovascular and diabetes risk in Polynesians, suggesting either that BMI is not an adequate measure of adiposity or that adiposity is not as strongly associated with CVD as in other populations [7]. These apparent contradictions are because BMI does not necessarily reflect the proportion of the body that is fat, and, in particular, average body composition may vary by ethnic group. Lanham and colleagues, for example, found that for lower than national average BMIs, Chinese Australian women carried a higher percentage of fat [11].

Excess abdominal fat is a particular risk for diabetes, indicating insulin resistance arising from the release of free fatty acids [12, 13]. Fat distribution differs by ethnicity, carrying implications for the risk for diabetes and other chronic diseases. Some population groups, such as Indians, Japanese and Australian Aborigines, appear to have a greater propensity than others to store fat abdominally [10, 14–18] and may be at greater risk of disease than others with comparable BMI. In Australia, such a variation in typical adipose distribution has been noted for some years, especially among Aboriginal women [16, 19–22], and also more recently among Aboriginal men [23].

Relationships between ‘obesity’ (BMI≥30) as defined by current NHMRC guidelines and risk for diabetes and cardiovascular (CVD) should therefore be reassessed in Australia’s Indigenous population, as these definitions may be inappropriate for this population. Examining the relationships between diabetes risk factors would illuminate three highly relevant clinical issues. First, whether excess central adiposity is associated with increases in diabetes and CVD risk factors (elevated fasting glucose and hypertension); second, whether the current definitions of general and central obesity are in line with each other in this setting; and, hence, third, whether there are perhaps more appropriate obesity criteria to use in this population to identify individuals at risk. Any assumptions about diabetes risk that use inappropriate thresholds for risk factors may be masking the real risk of diabetes and related morbidities in some communities; the current ‘healthy’ weight guidelines as recommended by NHMRC (BMI between 20 and 25 kg/m²) may need to be revised for Indigenous Australians.

Methods

Data were collected between September and December 2000, in a large urbanised Aboriginal community (population approximately 1,100) in southeast Queensland, Australia. This community has a high prevalence of diagnosed diabetes (20% among people aged over 18 years) [2].

Participants were selected through random household sampling using a Kish grid. One adult per household who had never been diagnosed with diabetes was asked to take part (pregnant women were excluded). Data were collected on each of the five NHMRC risk factors for Type 2 diabetes [5]. Standing height was measured to the nearest half-centimetre. Weight was measured using digital bathroom-type scales to the nearest 100g. Waist circumference was obtained using a tailor’s tape measure at the level of the umbilicus and measured to the nearest centimetre. Blood pressure was taken after five minutes resting, using an automatic monitor on the participant’s non-dominant side. Age in years was determined by the age the participant turned in the calendar year of the study. Positive family history was identified via questionnaire. Twelve-hour fasting blood glucose level was measured, to the nearest 0.1mmol/l using a personal glucose monitor that was calibrated regularly.

Relationships between modifiable risk factors were analysed using linear regression (SPSS for Windows v11.0). Sensitivity and specificity analyses of each modifiable risk factor in predicting elevated fasting glucose and hypertension were also conducted. The results were assessed to determine whether current NHMRC clinical guidelines are appropriate in this population.

Results

Sixty-two women (aged 18-66, mean = 31.8 years) and 55 men (aged 19-65, mean = 34.5 years) who had never been diagnosed with diabetes took part in the study (response rate 74%). The prevalence of diabetes risk factors among this random sample was extremely high (Table 2). Forty per cent of women and 25% of men had three or more diabetes risk factors. Fasting blood glucose was elevated above the ‘normal’ level of 5.5mmol in 35% of women and 51% of men (Table 3).

Table 2. Prevalence of independent diabetes risk factors among adults never diagnosed with diabetes

Independent risk factor	Women % (n=62)	Men % (n=55)
Age ≥35 years	35.5 (22)	45.5 (25)
Overweight or obese: BMI ≥25 kg/m²	59.7 (37)	54.5 (30)
Obesity: BMI ≥30 kg/m²	38.7 (24)	23.6 (13)
Central Obesity ^a	72.6 (45)	29.1 (16)
Systolic blood pressure>140mmHg	11.3 (7)	27.3 (15)
Diastolic blood pressure >90mmHg	22.6 (14)	34.5 (19)
Systolic and diastolic hypertension ^b	8.1 (5)	18.2 (10)
Family history ^c	71.9 (44)	45.5 (25)

^a For women, waist ≥88cm; for men, ≥102cm.
^b Systolic >140mmHg and diastolic >90mmHg.
^c Reported that one or more members of their immediate family (first degree relatives: includes parents, offspring, siblings – both full-siblings and half-siblings) had been diagnosed with diabetes.

Table 3. Fasting blood glucose levels. Categories from NHMRC

Fasting glucose	Women %(n=59)^a	Men %(n=48)^a
≥7mmol ^b	4.8 (3)	12.7 (7)
5.5-6.9mmol ^c	30.6 (19)	38.2 (21)
<5.5mmol ^d	59.7 (37)	36.4(20)

Source: NHMRC (2001) [5]
^a Fasting glucose was not obtained for three women and seven men.
^b Diabetes likely.
^c Diabetes status uncertain.
^d Diabetes unlikely.

Waist circumference increased linearly with BMI among both women and men (Figure 1), as might be expected. Most women surveyed (73%) had a waist circumference that placed them in the centrally obese category (WC≥88cm), including three out of ten women with BMIs below the ‘healthy’ range (<20). Overall, nearly one-half of the women with BMIs under 25 were centrally obese under current guidelines. This indicates a dominant pattern of centrally distributed fat even among those whose BMI indicated normal or less than normal bodyweight under current guidelines.

Above a BMI of 25 – the current threshold defining overweight – all women in the present study were centrally obese and might therefore be at increased risk of diabetes and cardiovascular disease. This was not the case among men, as central obesity (WC≥102cm) commenced only in the overweight range of 25 to 30, and all men with a BMI over 32 were centrally obese (Figure 1).

Click to enlarge

Figure 1. Scatter plot of waist circumference and BMI showing linear regression line (mean and 95% confidence interval) for women and men. Horizontal reference line (dotted) is set at 88cm for female waist circumference and 102cm for men, the accepted thresholds for central obesity; filled markers indicate the presence of central obesity, unfilled markers indicate its absence. The vertical reference lines (dotted) enclose the range of BMIs where central obesity was found to be both present and absent (17.5 to 24 for women, 28 to 32 for men).

Linear relationships between risk factors

Waist circumference was positively and significantly associated with increasing blood pressure and fasting blood glucose level among women (linear regression: systolic p=0.018, diastolic p=0.002, fasting glucose p=0.003) (Figure 2). These relationships were not as strong for BMI, so that central fat distribution was a stronger predictor of other risk factors than was overall body mass. The waist circumference threshold of 88cm for women appears to be appropriate given that (with one exception for fasting glucose) the thresholds for blood pressure and fasting glucose level were surpassed only by those who had a waist measurement above 88cm.

The relationships between risk factors appear to be more complex among men than among women. Despite an overall increase in both systolic and diastolic blood pressure and fasting blood glucose with waist circumference (Figure 2), these relationships were not significant (p=0.137 and 0.275 respectively). The only significant association was found for fasting glucose (p=0.017), and even this association was fairly weak (Rsq=0.1169). Hypertension and elevated fasting glucose were also apparent at low waist circumferences, which may suggest that 102cm is too high a definition of central obesity for this group of men.

Click to enlarge

Figure 2. Scatter plot of systolic and diastolic blood pressure and fasting blood glucose for waist circumference showing linear regression line (mean and 95% confidence interval). Reference lines (dotted) are at 88cm for female and 102cm for male waist circumference, and at 140mmHg for systolic pressure, 90mmHg for diastolic pressure, and 5.5mmol/l for fasting glucose. Filled markers indicate the presence of central obesity, unfilled markers indicate its absence.

Sensitivity and specificity analyses

Central obesity among women was a highly sensitive predictor for elevated fasting blood glucose (both at diabetes diagnostic level of ≥7mmol and elevated ‘at risk’ level ≥5mmol) (Figure 3). It was, however, not a very specific predictor, leading to a high proportion of ‘false positives’. Obesity was more specific (fewer false positives) but less sensitive. Obesity was a poor predictor of hypertension in women (low sensitivity, moderate specificity), while central obesity was highly sensitive but again yielded a high false positive rate. As a predictor of hypertension, overweight approached central obesity for sensitivity, and was more specific than both central obesity and general obesity. All measures had fairly low positive predictive values, but high negative predictive values, with overweight and central obesity performing better than obesity. A BMI of 25 or greater was a better predictor of central obesity than a BMI of 30 or more (Figure 4), having both greater sensitivity and negative predictive value.

Among men, overweight was more sensitive in predicting elevated fasting glucose and hypertension, but less specific than both obesity and central obesity. Overweight also had higher negative predictive value than obesity and central obesity, but lower positive predictive value than central obesity for hypertension and both obesity and central obesity for elevated fasting glucose. A reduced waist circumference threshold (≥90cm) in men was much more sensitive and generally had higher negative predictive value than the current central obesity threshold in predicting elevated blood glucose and hypertension, but was less specific and had slightly lower positive predictive value (Table 4).

Click to enlarge

Figure 3. Sensitivity and specificity analyses comparing current guidelines marking obesity (BMI ≥30), overweight (BMI ≥25) and central obesity (waist circumference ≥88cm for women and ≥102cm for men) in predicting elevated blood glucose and hypertension.

Figure 4. Sensitivity and specificity analyses comparing current thresholds marking obesity (BMI ≥30) and overweight (BMI ≥25) in predicting central obesity (waist circumference (≥88cm) in women.

Table 4. Sensitivity analyses for waist circumference =90cm in predicting elevated fasting glucose and hypertension among men

	Fasting glucose≥7mmol	Fasting glucose≥5.5mol	Systolic hypertension	Diastolic hypertension
Sensitivity	1.00	0.93	0.85	0.89
Specificity	0.24	0.40	0.21	0.26
Positive predictive value	0.18	0.68	0.30	0.46
Negative predictive value	1.00	0.80	0.78	0.78

The unmodifiable risks, age ≥35 years and positive family history of diabetes, were not significantly associated with either waist circumference or BMI among women or men (Table 5), and were not influencing the observed relationships between these and other diabetes risks. Younger women had lower BMIs than the older women, but their waist circumferences were greater. These differences were not significant, but raise questions about a possible secular trend of increasing central obesity, producing a more ‘at risk’ body shape in younger women.

Table 5. Means (SD) and significance levels (independent t-test) for waist circumference and BMI for age and family history

	Women				Men
	Age (years)		Family history		Age (years)		Family history
	≥35	<35	Yes	No	≥35	<35	Yes	No
Waist circumference (cm) (SD)	106.90 (13.1)	110.17 (18.8)	103.79 (17.7)	99.76 (16.0)	101.48 (17.2)	98.76 (14.3)	101.55 (15.4)	98.81 (16.1)
	p=0.162		p=0.424		p=0.554		p=0.552
BMI (kg/m²) (SD)	30.48 (6.6)	28.51 (9.2)	29.90 (8.5)	27.40 (7.9)	27.69 (6.4)	27.13 (5.0)	27.40 (5.5)	27.39 (5.9)
	p=0.396		p=0.302		p=0.736		p=0.997

Conclusions and implications

The positive relationship between BMI and waist circumference was linear and strong among both women and men. These variables are in reality continuous, but thresholds are useful in a clinical setting in identifying individuals at risk. Many women with BMIs conventionally considered normal, or even underweight, were found to be centrally obese by current NHMRC thresholds, and all women with a BMI of 25 or greater were centrally obese. Both BMI≥25 and WC≥88cm were found to be fairly good predictors of elevated fasting glucose and hypertension in women. Among men, BMI ≥25 and WC ≥90cm were fairly good predictors of elevated FG and hypertension.

Given the relationship between central obesity and BMI, either the risk threshold for central obesity for women in this population could be raised, or the threshold for obesity could be lowered to more adequately reflect diabetes and CVD risk. As WC≥88cm was more sensitive than BMI≥30 in predicting elevated fasting glucose and hypertension, reducing the BMI threshold rather than increasing the waist circumference threshold is a consideration that should be investigated further. The threshold for overweight, BMI≥25, was not only strongly predictive of central obesity but was both more sensitive than BMI≥30 and more specific than central obesity in predicting elevated fasting glucose and hypertension.

Among men, the relationship between BMI and waist circumference suggests that BMI≥30 appears to adequately reflect central obesity when the waist circumference threshold is ≥102cm. BMIs currently regarded as ‘overweight’ included both those with and without central obesity, while ‘obese’ adequately predicted those with central obesity. However, the current waist circumference threshold may be set too high, with ≥90cm being a better predictor of elevated fasting glucose and hypertension, and this was associated with BMIs ≥25.

The relationships between risk factors are complex, and no single one is sufficient to assess diabetes risk. Linear regressions and sensitivity analyses suggest, however, that both WC≥88cm and BMI≥25 may be appropriate first stage screening tools for women, and BMI≥25 and possibly WC≥90cm might be more useful for men in this population. While they do yield a high proportion of false positives in elevated fasting glucose and hypertension, these measures are more sensitive than BMI≥30, and will identify more readily those individuals who may be at risk.

The current NHMRC criterion defining central obesity is a very good predictor of the presence or absence of other independent diabetes risk factors among women, including fasting glucose. It is therefore recommended that, if BMI is to be used, then the ‘at risk’ threshold should be reduced for Australian Aboriginal women. Current NHMRC definitions of overweight and obesity appear to be inappropriate for women in this Aboriginal community, given the relationships between waist circumference and other diabetes risk factors. This finding is likely to be generalisable to other Aboriginal groups. Similar relationships have been observed between BMI and central obesity elsewhere, and waist circumference has been found in other communities to be associated with CVD and diabetes risk [23, 24]. Central obesity is clearly correlated with diabetes risk, and the BMI guidelines for Indigenous people should reflect this risk. This increased risk at lower BMIs is not reflected by the current guidelines for obesity.

A reduced BMI threshold for women by 5kg/m² to 25kg/m² could be a more appropriate indicator of individual risk. An even lower threshold for Aboriginal populations has been advocated by Daniel and colleagues, who suggest that a BMI ≥22kg/m² might be appropriate [22]. Among men, reducing waist circumference threshold to =90cm and BMI to 25kg/m² would appear to better reflect risk.

Internationally too, there are precedents for defining population-specific obesity thresholds; lowering the threshold has been proposed for Japanese [10] and Chinese [8, 9, 25] populations, for example. Alternatively, BMI should perhaps not be regarded as a risk factor at all; waist circumference alone is perhaps a better predictor of diabetes and CVD risk in both Indigenous and non-Indigenous populations. Wang and Hoy found waist circumference better than BMI at predicting CVD risk in a remote Aboriginal community [24, 26], as did Welborn and colleagues in a national survey of the Australian population [27]. Dalton and colleagues suggest that waist-hip ratio might be better in a clinical setting than BMI at identifying at risk individuals [28]. BMI is not intrusive to measure and is easy to calculate, and it therefore remains a useful clinical tool. It may also be perceived (and may actually be) more achievable for individuals to reduce their overall body weight, rather than target fat reduction in specific areas. For BMI to be relevant in Indigenous communities, however, the classifications of healthy and at risk need to be reconsidered, particularly among women.

Further, the data presented here demonstrate that even women who are ‘underweight’ by NHMRC guidelines may still be centrally obese, and therefore at risk of elevated fasting glucose and hypertension. Because of their low BMI, this group of women may be ignored in initial clinical assessments of diabetes risk. Women with low BMIs may need to be targeted specifically for diabetes screening. There may also be a secular trend of increased waist circumference, producing an ‘at risk’ body shape among younger women.

Waist circumference is much more strongly and significantly associated with other risk factors among women than among men. Given the strong positive linear relationship between waist circumference and BMI, reducing overall obesity among women, especially, could have a substantial impact on diabetes risk by reducing central adiposity.
This paper highlights the problems associated with developing guidelines for one population and applying them to another. Indigenous Australians are now recognised to be at particular risk of diabetes, but the only specific risk factor in the current NHMRC clinical guidelines that recognises this is age.

Current national guidelines on waist circumference in women appear to be appropriate, but the relevance of these guidelines for BMI thresholds should be investigated further. In men, both waist circumference and BMI thresholds could be reduced to better reflect diabetes and CVD risk.

Acknowledgements

The author wishes to thank the community involved for their warm enthusiasm and participation in the study, especially Toni Kirk and Maureen Weazel for assisting with the surveys. Thanks also to two anonymous reviewers for their comments on the manuscript. The research was partially funded by an Australian Postgraduate Award (APA) PhD Scholarship. Additional funding was provided by the Australian Institute of Aboriginal and Torres Strait Islander Studies (AIATSIS grant S6116076) and the Faculty of Arts at the Australian National University.

Further information

Contact details:
Dr Hilary Jane Bambrick, Research Fellow, National Centre for Epidemiology and Population Health, Building 62
The Australian National University, Canberra ACT 0200, Australia, Email: hilary.bambrick@anu.edu.au, Ph: +61 2 6125 8595, Fax: +61 2 6125 0740

Ethical Approval:
The study was approved by the Community Council, the community’s Health Action Committee, and the Human Research Ethics Committee at the Australian National University.

Conflict of interest:
None

Key words:
Type 2 diabetes, BMI, waist circumference, central obesity, fasting glucose, hypertension, risk factor, Indigenous health, clinical guidelines.

References

1. Carter JN, Bartley PC (1996) The diabetes epidemic in Australia: prevalence, patterns and the public health (letter). Medical Journal of Australia;164:51
2.Bambrick H (2001) Prevalence of diagnosed and undiagnosed Type 2 diabetes in a Queensland Aboriginal community. Australian Indigenous HealthBulletin, 1(1), Article 1. Retrieved [access date] from http://www.healthinfonet.ecu.edu.au/html/html_bulletin/bull_11/bulletin_original_articles.htm
3.de Courten M, Hodge A, Dowse G, King I, Vickery J, et al. (1998) Review of the epidemiology, aetiology, pathogenesis and preventability of diabetes in Aboriginal and Torres Strait Islander populations. Canberra: Commonwealth Department of Health and Family Services
4. Australian Institute of Health and Welfare (2002) Australia’s health 2002: the eighth biennial report of the Australian Institute of Health and Welfare. Canberra: Australian Institute of Health and Welfare
5. National Health and Medical Research Council (NHMRC) (2001) National Evidence Based Guidelines for the Management of Type 2 Diabetes Mellitus. Sydney: Australian Centre for Diabetes Strategies
View report
6. Seidell JC (2000) Obesity, insulin resistance and diabetes–a worldwide epidemic. British Journal of Nutrition;83 Suppl 1:S5-8
View abstract
7. McAnulty J, Scragg R (1996) Body mass index and cardiovascular risk factors in Pacific Island Polynesians in New Zealand. Ethnicity and Health;1(3):187-195
8. Bei-Fan Z (2002) Predictive values of body mass index and waist circumference and risk factors of certain related diseases in Chinese adults: study on optimal cut-off points of body mass index and waist circumference in Chinese adults. Asia Pacific Journal of Clinical Nutrition;11(Supplement 8):S685-S693
9. Cockram CS (2000) The epidemiology of diabetes mellitus in the Asia-Pacific region. Hong Kong Medical Journal;6(1):43-52
10. Kanazawa M, Yoshiike N, Osaka T, Numba Y, Zimmet P, et al. (2002) Criteria and classification of obesity in Japan and Asia-Oceania. Asia Pacific Journal of Clinical Nutrition;11(Supplement 1):S732-S737
11. Lanham DA, Stead MA, Tsang K, Davies PS (2001) The prediction of body composition in Chinese Australian females. International Journal of Obesity and Related Metabolic Disorders;25(2):286-91
View abstract
12. Bjorntorp P (1991) Metabolic implications of body fat distribution. Diabetes Care;14(12):1132-43
13. Bergman RN, Ader M (2000) Free fatty acids and pathogenesis of type 2 diabetes mellitus. Trends in Endocrinology and Metabolism;11(9):351-6
View abstract
14. Rush EC, Plank LD, Mitchelson E, Laulu MS (2002) Central obesity and risk for type 2 diabetes in Maori, Pacific and European young men in New Zealand. Food and Nutrition Bulletin;23(Supplement 3):82-6
15. Yajnik CJ (2002) The lifecycle effects of nutrition and body size on adult adiposity, diabetes and cardiovascular disease. Obesity Reviews;3:217-224
16. O’Dea K (1987) Body fat distribution and health outcome in Australian Aborigines. Proceedings of the Nutrition Society of Australia;12:56-65
17. Yudkin JS (1996) Non-insulin-dependent diabetes mellitus (NIDDM) in Asians in the UK. Diabetic Medicine ;13(9 Supplement 6):S16-S18
18. McNeely MJ, Boyko EJ, Shofer JB, Newell-Morris L, Leonetti DL, et al. (2001) Standard definitions of overweight and central adiposity for determining diabetes risk in Japanese Americans. American Journal of Clinical Nutrition;74(1):101-107
19. O’Dea K, Patel M, Kubisch D, Hopper JL, Traianedes K (1993) Obesity, diabetes and hyperlipidemia in a Central Australian Aboriginal community with a long history of acculturation. Diabetes Care;16(7):1004-1010
20. O’Dea K, Lion RJ, Lee A, Traianedes K, Hopper JL, et al. (1990) Diabetes, hyperinsulinemia, and hyperlipidemia in a small Aboriginal community in northern Australia. Diabetes Care;13(8):830-835
21. Rowley KG, Best JD, McDermott R, Green EA, Piers LS, et al. (1997) Insulin resistance syndrome in Australian Aboriginal people. Clinical and Experimental Pharmacology and Physiology;24:776-781
22. Daniel M, Rowley KG, McDermott R, O’Dea K (2002) Diabetes and impaired glucose tolerance in Aboriginal Australians: prevalence and risk. Diabetes Research and Clinical Practice;57:23-33
23. Piers LS, Rowley KG, Soares MJ, O’Dea K (2003) Relation of adiposity and body fat distribution to body mass index in Australians of Aboriginal and European ancestry. European Journal of Clinical Nutrition;57(8):956-63
View abstract
24. Wang Z, Hoy WE (2004) Body size measurements as predictors of type 2 diabetes in Aboriginal people. International Journal of Obesity and Related Metabolic Disorders;28(12):1580-4
View abstract
25. Lee ZS, Critchley JA, Ko GT, Anderson PJ, et al. (2002) Obesity and cardiovascular risk factors in Hong Kong Chinese. Obesity Reviews ;3(3):173-182
26. Wang Z, Hoy WE (2004) Waist circumference, body mass index, hip circumference and waist-to-hip ratio as predictors of cardiovascular disease in Aboriginal people. European Journal of Clinical Nutrition;58(6):888-93
View abstract
27. Welborn TA, Dhaliwal SS, Bennett SA (2003) Waist-hip ratio is the dominant risk factor predicting cardiovascular death in Australia. Medical Journal of Australia;179(11-12):580-5
View abstract
28. Dalton M, Cameron AJ, Zimmet PZ, Shaw JE, et al. (2003) Waist circumference, waist-hip ratio and body mass index and their correlation with cardiovascular disease risk factors in Australian adults. Journal of Internal Medicine;254(6):555-63
View abstract

Type 2 diabetes and patterns of alcohol use in a Queensland Aboriginal community

Issue: Vol 4 No 3, July 2004 - September 2004

Hilary Bambrick

View PDF

Abstract

Objectives
To assess patterns of alcohol consumption in people with and without diabetes.
Methods
Location was a large Aboriginal community in southeast Queensland. Participants with diagnosed diabetes were identified through the hospital database (49 women, 38 men), and never-diagnosed participants recruited through random household sampling (62 women, 55 men). Alcohol consumption patterns were ascertained by questionnaire.
Results
Overall, 56% of participants consumed alcohol. On average alcohol was consumed only twice per week, but the number of drinks consumed per drinking day was high (17; range: 3.5-20). Compared with never-diagnosed participants, participants with diabetes were less likely to drink (women RR=0.3, 95%CI 0.2-0.5; men RR=0.7, 95%CI 0.5-0.9), drank less per week (ANOVA: women 9.4 versus 34.4, p<0.001; men 10.6 versus 31.2, p=0.004), had fewer drinking days (women 0.5 versus 1.9, p<0.001; men 0.7 versus 1.7, p=0.011), and consumed less per drinking day (ANOVA: women 15.5 versus 18.4, p=0.003; men 16.0 versus 16.5, p=0.006).
Conclusions
People with diabetes are less likely than others to drink, and those who do drink consume fewer drinks on a day when they drink, suggesting they have modified their behaviour based on health advice. However, the overall quantity consumed in all groups remains high, at levels considered risky and high-risk in both the short- and long-term.
Implications
Given the diabetes and cardiovascular implications of heavy short-term and long-term alcohol use, specific patterns of alcohol use by people with diabetes should be assessed further to develop strategies to reduce the amount of alcohol consumed on a drinking day.

Suggested citation: Bambrick H (2004) Type 2 diabetes and patterns of alcohol use in a Queensland Aboriginal community. Australian Indigenous HealthBulletin;4(3): Original article. Retrieved [access date] from
http://www.healthinfonet.ecu.edu.au/html/html_bulletin/bull_43/original_articles/bulletin_original_articles_bambrick.htm

Introduction

Type 2 diabetes is a significant cause of excess Indigenous morbidity and mortality, and cardiovascular disease is the leading cause of Indigenous death [2]. Moderate alcohol consumption may confer some cardiovascular benefits [3], reduce the incidence of diabetes [4], and reduce cardiovascular risk among those who have diabetes [5], but the heavy use of alcohol can contribute to poorer cardiovascular health and may increase serious complications from diabetes, through increasing triglyceride levels [6], and can cause liver and pancreas damage [7].

Fewer Indigenous people consume alcohol than the non-Indigenous people, but when alcohol is consumed, it tends to be at much higher levels [8]. These broad trends do not consider the impact that being diagnosed with diabetes (which is very common among Indigenous people) might have on alcohol use or differences between women and men. Determining such patterns of alcohol use at a community level is especially important given its influence on diabetes management and contribution to overall cardiovascular health.

Methods

This study took place in a large, urbanised Aboriginal community (population approximately 1,200) in southeast Queensland between September and December 2000. Participants with diagnosed type 2 diabetes (49 women, 38 men) were identified through the community hospital database. At the time of the study, the hospital served as an outpatient clinic providing near universal coverage of the community. Participants without diagnosed diabetes (‘never-diagnosed’) were recruited through household sampling (62 women, 55 men) where one person aged over 18 years from each household was randomly selected using a Kish grid.

Response rates were high (approximately 90% diagnosed and 70% never-diagnosed eligible people), but men were slightly less likely than women to participate. Participants with diagnosed diabetes were on average 15 years older than never-diagnosed participants (women: diagnosed range 19-71, mean 47 years, never-diagnosed range 18-66, mean 32 years; men: diagnosed range 27-79, mean 49 years; never-diagnosed range 19-65, mean 35 years). Further details on participant selection methods are provided elsewhere [9].

Participants were asked about the frequency of their alcohol consumption (the usual number of days per week on which alcohol was used) and the amount of alcohol they consumed (number of drinks usually consumed on a drinking day). The total number of drinks usually consumed per week was then estimated.

The study was approved by the Human Research Ethics Committee at the Australian National University.

Results

Overall, 48% of women and 67% of men in the study (56% of all participants) reported that they consume alcohol. Participants with diagnosed diabetes were less likely to drink than never-diagnosed participants (women RR=0.3, 95%CI 0.2-0.5; men RR=0.7, 95%CI 0.5-0.9).

Most of those who did consume alcohol did so on fewer than two days per week. Overall, only 21% of women and 19% of men consumed alcohol on three or more days each week. On a day when alcohol was used, however, the number of drinks consumed tended to be very high (mean=17, range 3.5 to >20). No participant reported consuming just one or two drinks on a day when they were drinking. Figure 1 shows estimated weekly alcohol consumption, usual number of drinking days per week and usual number of drinks consumed on a drinking day for women and men by diabetes diagnosis.

Figure 1. Usual number of drinks per week (A) (calculated by midpoint of number of days * midpoint number of drinks per drinking day), usual number of drinking days per week (B) and usual number of drinks consumed per drinking day (C) for diagnosed (solid line) and healthy (dotted line) women and men. ANOVA was used to test significance. Age was controlled for. NB. ‘Number of drinks’ was not standardised, for example, as one can of beer contains approximately 1.4 ‘standard drinks’, the number of standard drinks consumed may be much higher than the number reported. In Australia, one standard drink contains 10g of alcohol (the amount of alcohol in a ‘standard’ drink varies internationally) [1] .

After controlling for age, differences between diagnosed and never-diagnosed participants were significant for number of drinks per week (ANOVA: women 9.4 versus 34.4, p<0.001; men 10.6 versus 31.2, p=0.004), and number of drinking days per week (women 0.5 versus 1.9, p<0.001; men 0.7 versus 1.7, p=0.011). Women and men with diabetes who did drink were drinking less on a drinking day than others (ANOVA: women 15.5 versus 18.4, p=0.003; men 16.0 versus 16.5, p=0.006). This might be partially related to age, at least among women, as the significance of the differences was reduced when age was included in the model as a covariate (women p=0.48; men p=0.012).

The differences between these groups may not be fully explained by the smaller proportion of people with diagnosed diabetes who drink. When only those who did drink were considered, diagnosed men drank significantly less per week than others (14.6 versus 42.4, p=0.039) but were not drinking significantly less often (1.0 versus 2.3 times per week, p=0.079). Diagnosed women who drank were not drinking significantly less per week than other women who drank (26.54 versus 54.2, p=0.156), nor did they drink significantly less often (1.3 versus 2.9, p=0.075).

Discussion

Of all study participants, 56% (20% diagnosed women, 69% never-diagnosed women, 53% diagnosed men, 76% never-diagnosed men) reported that they consumed alcohol. This contrasts with the 83% reported for Australia as a whole [2]. People in this community who have had diabetes diagnosed were less likely to consume alcohol than those who had never been diagnosed with diabetes. They also drink less per week, had fewer drinking days per week and consumed less per drinking day. Much of the difference was due to the greater proportion of those with diabetes abstaining from alcohol altogether. Diagnosed women and men consumed less alcohol than never-diagnosed women and men, all those who drank consumed large quantities of alcohol on a drinking day.

Low levels of alcohol consumption (such as 1-2 drinks on several days per week) is protective against cardiovascular disease and diabetes [3, 5, 10]. In this study, participants were either non-drinkers or consumed large quantities on a few drinking days per week.

According to the Australian alcohol guidelines, the risk of alcohol-related harm can be considered in terms of short-term and long-term risks (Table 1) [11]. Short-term risky or high-risk drinking refers to the quantity consumed on a given drinking day, while long-term risk refers to the usual quantity consumed each week. In the short term, risks include increased likelihood of accidents, physical and sexual assault, and a reduced capacity to protect against sexually transmitted infections and pregnancy. In the long-term, regular heavy drinking by women increases their risk of cardiovascular disease and other chronic degenerative disorders, such as liver damage and diabetes resulting from pancreatitis.

Table 1. Australian risk classifications for alcohol-related harm in the short- and long-term [11]

Risk type	Women		Men
Risk type	Risky	High-risk	Risky	High-risk
Short-term risks (number of drinks per day) Increased risk of alcohol-related violence, accidents and risky behaviours	5-6	≥7	7-10	≥11
Long-term risks (number of drinks per week) Increased risk of chronic degenerative diseases such as diabetes and cardiovascular disease	15-28	≥29	29-42	≥43

In the Australian population as a whole, 12% of women and 15% of men drink one or more times per week at levels considered either risky or high-risk in the short-term [12]. Prevalence of short-term risky and high risk drinking differed in the current study according to diabetes status, with never-diagnosed women reporting the highest levels and diagnosed women the lowest (20% of diagnosed women, 29% of diagnosed men, 69% of never-diagnosed women, 62% of never-diagnosed men). Many of those with diagnosed diabetes did not drink, but those that did still consumed quantities of alcohol that were risky or high-risk in the short-term. Given the ready availability of simple sugars in alcohol, heavy alcohol consumption among diabetics, even if infrequent, is a serious concern in terms of short-term blood glucose control and long-term contribution to cardiovascular damage.

Long-term risky and high-risk drinking has been estimated to occur in approximately 20% of Indigenous Australians, while for Australian women and men the overall prevalences are 9% and 10% respectively [13]. Results from this study indicate that among those who have never had diabetes diagnosed, up to 60% of women and 49% of men are drinking at risky and high-risk levels. The prevalence of long-term risky drinking is, however, much lower among people with diagnosed diabetes, at 6% for women and 16% for men.

The current study found that not only is the prevalence of short-term risky drinking higher among never-diagnosed women than never-diagnosed men, the long-term drinking behaviour of women is at least as risky as that of men. In the past, the heavy consumption of alcohol has been frequently associated with notions of masculinity [14, 15]. A propensity for more general risk-taking is also commonly associated with men rather than with women, especially young men, and particularly those who are disaffected and feel that they have little to lose [16]. The combination of being both male and socially disadvantaged is generally considered to produce the worst outcomes for health [17]. This no doubt remains a very important aspect of alcohol use, but this study demonstrates patterns of consumption by women also need to be addressed more comprehensively. Some of women’s excess risky drinking in comparison to men’s arises from the lower thresholds defining risk for women, but many women reported drinking greater quantities of alcohol than men – amounts that are two and three times those defined as high risk.

The use of self-reported alcohol consumption is a limitation of this study. ‘Number of drinks’ consumed was not based on Australian standard drinks, but, given that standard drinks containing 10g of alcohol are frequently smaller than individual items consumed (such as a can of beer which is approximately 1.4 standard drinks), the reporting is likely to underestimate rather than overestimate the number of standard drinks consumed by participants. As total weekly consumption was derived from the midpoints of both alcohol frequency categories (number of days per week) and number of drinks usually consumed, the study may overestimate consumption for some but underestimate for others. Further, the survey questions were framed in the present, and may not indicate longer term drinking patterns. For example, some may have considered themselves to be non-drinkers because it had been several months since they had consumed alcohol. As questions were not asked about past drinking behaviours it cannot be confirmed that being diagnosed with diabetes causes changes to drinking behaviours.

Major strengths of the study include the very high response rates (70%-90%), and the willingness of participants to answer questions relating to their alcohol consumption behaviour. Participants’ willingness to take part in the study and awareness that their answers would be kept confidential suggests there is no reason to doubt the integrity of the responses.

The age difference between those diagnosed with diabetes and never-diagnosed participants was not a significant factor in the different drinking behaviours of the two groups. Age was found to have an effect only in the amount consumed per drinking day. That fewer people with diagnosed diabetes drank and that those who did drink did so less often than never-diagnosed people is probably primarily due to efforts made to follow diabetes management advice to ‘drink less alcohol’. If so, this success should be built upon, with further efforts to reduce the amount consumed on a drinking day.

Changing drinking behaviours of individuals means changing the dominant drinking culture. The pattern of either abstaining from alcohol or consuming risky quantities is probably the result of interlinking factors with deep historical roots. Polarising morality around alcohol use widens the divide between non-drinkers and heavy drinkers, and fails to allow room for moderation, which could confer some cardiovascular benefits.

Alcohol may contribute to obesity through supplying excess calories [18], but a review of national data on Indigenous obesity found that those who did not drink alcohol had, on average, a higher body mass index (BMI) than those who drank [19]. This could be because heavy drinking often correlates with heavy smoking, and also because heavy drinking itself may limit the intake of foods. For example, participants in the present study who regularly missed meals sometimes reported that this was because they had been drinking or were ‘grog sick’. The lower BMIs among drinkers found in the review of national data could also be because the behaviour-modifying potential of diabetes diagnosis was not considered. People with diagnosed diabetes may have both a higher BMI and be less likely to consume alcohol than those without.

An environment where there are low levels of physical activity combined with a diet that is typically energy-dense promotes both cardiovascular disease and diabetes. The effects of physical inactivity and an energy-dense diet may be exacerbated by the patterns of drinking that are taking place among never-diagnosed people, increasing their future risk of diabetes and cardiovascular disease.

Overall, there appears to have been some success in reducing alcohol intake among those who have been diagnosed with diabetes, but considerable emphasis should be placed also on finding strategies to reduce the amount that people drink ‘on a drinking day’.

Acknowledgements
The data used here were gathered during PhD research into diabetes. The author thanks the Aboriginal community involved for its warmth, enthusiasm and participation – in particular the study participants and the members of the Community Health Team. Many thanks are extended also to the Australian Institute of Aboriginal and Torres Strait Islander Studies for financing some of the research (AIATSIS grant number S6116076), and to Antonia Kirk and Maureen Weazel for their invaluable assistance. Thanks to Emily Banks and two anonymous reviewers for their helpful comments.

References
1. Food Standards Australia New Zealand (2002) Food Standards Code. Retrieved 20 January 2003 from http://www.foodstandards.gov.au/foodstandardscode/index.cfm#_FSCchapter2
2. Australian Institute of Health and Welfare (2004) Australia’s health 2004: the ninth biennial report of the Australian Institute of Health and Welfare. Canberra: Australian Institute of Health and Welfare
3. Mukamal KJ, Conigrave KM, Mittleman MA, et al. (2003) Roles of drinking pattern and type of alcohol consumed in coronary heart disease in men. New England Journal of Medicine;348(2):109-118
4. Nakanishi N, Suzuki K, Tatara K (2003) Alcohol consumption and risk for development of impaired fasting glucose or type 2 diabetes in middle-aged Japanese men. Diabetes Care;26(1):48-54
5. Solomon CG, Hu FB, Stampfer MJ, et al. (2000) Moderate alcohol consumption and risk of coronary heart disease among women with type 2 diabetes mellitus. Circulation;102(5):487-488
6. McGrath M, Collins V, Zimmet P, Dowse G (1991) Lifestyle disorders in Australian Aborigines. Diabetes and cardiovascular disease risk factors – a review. Canberra: Brolga Press
7. Khot UN, Khot MB, Bajzer CT, et al. (2003) Prevalence of conventional risk factors in patients with coronary heart disease. Journal of the American Medical Association;290(7):898-904
8. Australian Bureau of Statistics and Australian Institute of Health and Welfare (2001) The health and welfare of Australia’s Aboriginal and Torres Strait Islander peoples. Canberra: Australian Bureau of Statistics and Australian Institute of Health and Welfare
9. Bambrick HJ (2001) Prevalence of diagnosed and undiagnosed diabetes on a Queensland Aboriginal community. Australian Indigenous HealthBulletin;1(1)
10. Nakanishi N, Suzuki K, Tatara K (2003) Alcohol consumption and risk for development of impaired fasting glucose or type 2 diabetes in middle-aged Japanese men. Diabetes Care;26(1):48-54
11. National Health and Medical Research Council (2001) Australian alcohol guidelines: health risks and benefits. Canberra, ACT: National Health and Medical Research Council
12. Australian Institute of Health and Welfare (2002) Australia’s health 2002: the eighth biennial report of the Australian Institute of Health and Welfare. Canberra: Australian Institute of Health and Welfare
13. Australian Institute of Health and Welfare (2002) 2001 National Drug Strategy Household Survey: first results. (AIHW Cat. No. PHE 35) Canberra: Australian Institute of Health and Welfare
14. Lemle R, Mishkind ME (1989) Alcohol and masculinity. Journal of Substance Abuse Treatment;6(4):213-222
15. Williams RJ, Ricciardell LA (1999) Gender congruence in confirmatory and compensatory drinking. Journal of Psychology;133(3):323-331
16. Chisholm JS (1996) The young male syndrome: the evolutionary psychology of risk-taking in young men. In: Robinson G, ed. Aboriginal health: social and cultural transitions: proceedings of a conference at the Northern Territory University, Darwin, 29-31 September, 1995. Darwin: Centre for Social Research and the Faculty of Aboriginal and Torres Strait Islander Studies, NTU:124-128
17. Connell RW, Schofield T, Walker L, et al. (1998) Men’s health: a research agenda and background report. Canberra: Commonwealth Department of Health and Aged Care
18. Hoy WE, Pugsley DJ, Mathews JD, Hargreave JC (1996) Diabetes and impaired glucose tolerance in an Aboriginal community on the Northern Territory of Australia. In: Robinson G, ed. Aboriginal Health: Social and Cultural Transitions. Proceedings of a conference at the Northern Territory University, Darwin, 29-31 September, 1995. Darwin: Centre for Social Research and the Faculty of Aboriginal and Torres Strait Islander Studies, NTU:44-48
19. Cunningham J, Mackerras D (1994) Overweight and obesity – Indigenous Australians. (Occasional Paper. ABS Catalogue No. 4702.0) Canberra, ACT: Australian Bureau of Statistics

Dr Hilary Bambrick
National Centre for Epidemiology and Population Health
The Australian National University
Canberra ACT 0200
Australia
Phone: +61 2 6125 8595
Email: hilary.bambrick@anu.edu.au

Introduction

Methods and approach

Results

Why they ride

Knowing and doing

‘Doubling’ and pragmatism

Helmets and ‘cool’

Safe riders vs good riders

Police and the law

Dangerously significant others

Conclusions

Implications

References

Introduction

Charting Aboriginal child growth

Methods

Study population

Sampling

Analyses

Results

Comparisons with the WHO reference

Comparisons between girls and boys in the community

Discussion

Acknowledgements

Further information

Introduction

Methods

Results

Linear relationships between risk factors

Sensitivity and specificity analyses

Conclusions and implications

Acknowledgements

Further information

References

Introduction

Methods

Results

Discussion

Contact us