Contents - Previous - Next


General interpretation

The findings of this study indicate first that patterns of ageing, illness, and the nature of the fallow period of the subsistence base contribute independently to variation in Iban BMI. Differences in these patterns between the sexes confound the functional interpretation of low BMI values.

BMI, illness and ageing

When the data were analysed in relation to numbers of symptoms, the observed sex-differential in response to illness arguably reflected the relative resilience of females compared to males, which has been reported elsewhere (Stint, 1968). Similarly, the pattern of ageing indicated that women tolerated a wider range of BMI values than did men, a finding which resembles that of longitudinal studies of Americans followed from childhood to middle-age (Casey et al., 1992), and of Norwegians monitored for tuberculosis (Waaler, 1984). This is consistent with Henry's review (1990) of limits to survival being a BMI of 13.0 in men and 12.0 in women, and with evidence for behavioural adaptations in low BMI men but not women according to data from India and Ethiopia (Ferro-Luzzi et al., 1992).

This raises the question of whether, in classifications of low BMI, there should be distinct cut-off points for each sex. If only those who were reported to be asymptomatic in the present study are considered, then the mean -2 SD for each sex divided into ages <40 and >40 years suggests lower cut-off points of about 17.0 and 16.5 in men and women <40 years, and 15.0 and 14.5 in men and women above this age. The closely similar values suggest little need for different cut-off points for each sex whatever the case in this respect. It is arguable that there is a need for age-related cut-off points, centiles or Z-scores of BMI in healthy adults for use in the monitoring and surveillance of populations at risk of CED.

On the other hand, by the admittedly weak criterion of reported symptoms of illness, the poor sensitivity of BMI for identifying adults with symptoms suggests that this indicator may be a poor screening tool. A follow-up of these adults, to determine their age-controlled morbidity and mortality risks by BMI class, would be valuable in this respect and enable a test of the claim that energy and nutrient reserves are increasingly important in recuperative capacity as age increases (Stint, 1991).

BMI and the fallow farm

When the analysis was broadened to take the nature of the farm into account, the observed relationships between illness and fallow class in determining nutritional health were shown to be additive rather than interactive. Among men, the response of BMI to fallow farm was marked, illness having no significant independent effect. The pattern among women was the converse. Were short fallow farming to be deleterious to adult health, then negative effects of fallow class on anthropometric indices of short-term nutritional status would be anticipated. Thus, the findings were the opposite of this pattern. They therefore serve to emphasize that conventional health indices are meaningless when the definition of health as a sustainable state is used.

There are various possible explanations for the observed pattern. One is that levels of physical exertion during the harvest season are higher on long fallow farms than elsewhere, for example because such farms may lie a relatively far distance from the longhouse, and physical activity is greater among men than women. Second, 70% of those who were either farming with the shortest fallow plots or not engaged in any farming were in downriver areas within about 2 h travelling time from the District centre. The survey period coincided with the ending of the season for collecting jungle fruits and illipe nut (Shorea spp.): the profitable sale of these required easy access to markets. Thus, the relative commercial advantages for downriver groups may have been significant nutritionally. This indicates that the sustainability of progressively wider areas of economic infrastructure would need to be investigated were King's notion of health (King, 1990) to be subject to exhaustive inquiry.

For these reasons, the nature of the farm partially resembles an index of socio-economic class. If it is interpreted in this way, then these harvest season data on presence/absence of 'illness symptoms' suggest that the BMI of men primarily reflects their economic activities, while that of women reflects their experience of illness. The AMBA was the only measure influenced by illness in both sexes and independently of any effects of the fallow classification of the farm. Without time-series data, it is impossible to judge the relative merits of these two indices as measures of chronic rather than acute low energy nutrition. However, in this population, AMBA may be a more satisfactory health index than the BMI.


1. There is arguably a need for age-related centiles or norms for BMI throughout adulthood.

2. Sensitivity of BMI <18.5 for identifying those reporting symptoms of illness was in the range 20-30%.

3. The BMI may be less useful than AMBA as a functional index for monitoring ill health.

4. There was no support for the practical use of the notion of health as a sustainable state.

Acknowledgements - The authors thank the State Secretary of Sarawak for permission to undertake the survey reported in this paper; Sarawak administrative staff Dr H. Solhee, Dr S. Hordin, J. Langub, A. Tambi, O. Murang, R Rigep, Dr M. Lopez, R Mandy, J. Meling, R Ling, and Dr Khoo Teik Beng for logistical support; C. Uman and R Rosano for survey assistance; the Tuai Tuai Rumah for their cooperation; K. Homewood and V. Tuffrey for helpful advice. S.S.S. gratefully acknowledges a Royal Society study visit award and a University of London Hayter travel grant, S.J.U. a grant from the University of Cambridge Smuts Memorial Fund.


Bailey KV (1961): Rural nutrition studies in Indonesia. II. Clinical studies of hunger oedema in the cassava areas of Java. Trop. Geogr. Med. 13, 234-254.

Bailey KV (1962): Rural nutrition studies in Indonesia. X. Weight and height of Gunung Kidul adults. Trop. Geogr Med. 14, 230-237.

Casey VA, Dwyer JT, Coleman KA & Valadian I (1992): Body mass index from childhood to middle age: a 50-y follow-up. Am. J. Clin. Nutr. 56, 14-18.

Department of Medical Services (1988): Nutrition assessment report 1985/1987. Department of Medical Services, Government of Sarawak, Kuching.

FAO/WHO/UNU (1985): Energy and protein requirements. WHO, Geneva.

Ferro-Luzzi A, Sette S. Franklin M & James WPT (1992): A simplified approach of assessing adult chronic energy deficiency. Eur. J. Clin. Nutr. 46, 173-186.

Freeman DJ (1955): Iban agriculture. London: HMSO.

Habicht JP (1980): Some characteristics of indicators of nutritional status for use in screening and surveillance. Am. J. Clin. Nutr 33, 531-535.

Hatch T & Lim CP (1978): Shifting cultivation in Sarawak: A report based upon the workshop on shifting cultivation held in Kuching on 7-8th December, 1978. Kuching: Department of Agriculture, Government of Sarawak.

Henry CJK (1990): Body mass index and the limits of human survival. Eur J. Clin. Nutr. 44, 329-335.

Hong E (1987): Natives of Sarawak: survival in Sarawak's vanishing forests. Pulau Pinang: Institut Masyarakat.

James WPT, Ferro-Luzzi A & Waterlow JC (1988): Definition of chronic energy deficiency in adults. Eur J. Clin. Nutr 42, 969-981.

King M (1990): Health is a sustainable state. The Lancet ii, 664-667.

Lewis A (1953): Health as a social concept. Br. J. Sociol. 4, 109-124.

Lewis GA (1975): Knowledge of illness in a Sepik society. London: Athlone Press.

Little MA, Galvin K & Mugambi M (1983): Cross-sectional growth of nomadic Turkana pastoralists. Hum. Biol. 55, 811-830.

Norgan NG & Jones PRM (1990): Anthropometry and body composition. In Handbook of methods for the measurement of work performance, physical fitness and energy expenditure in tropical populations, e d. KJ Collins, pp. 95-115. Paris: IUBS.

Richards AJN (1981): An Iban-English dictionary. Oxford: Oxford University Press.

Stini WA (1968): Nutritional stress and growth: sex difference in adaptive response. Am. J. Phys. Anthrop. 31, 417-426.

Stini WA (1991): The biology of human ageing. In Applications of biological anthropology to human affairs, eds CGN Mascie-Taylor and GW Lasker, pp. 207-236. Cambridge: Cambridge University Press.

Strickland SS & Ulijaszek SJ (1993): Body mass index, ageing and differential reported morbidity in rural Sarawak. Eur. J. Clin. Nutr 47, 9-19.

Trowbridge FL & Sommer A (1981): Nutritional anthropometry and mortality risk. Am.J. Clin. Nutr 34, 2591-2592.

de Vasconcellos MTL (1994): Body Mass Index: Its relationship with food consumption and socioeconomic variables in Brazil. Eur J. Clin. Nutr 48, Suppl. 3, S115-S123.

Waaler HT (1984): Height, weight and mortality: the Norwegian experience. Acta Med Scand., Suppl. 679.

Weiner JS & Lourie JA (1981): Practical human biology. London: Academic Press.


Kennedy: Why did you decide to do a 24h recall of self-reported illness. Did you have a chance to observe the illness and then compare this with the recall? Strickland: We saw people only once. Ferro-Luzzi: What is covariance between BMI and arm circumference?

Strickland: It is 0.6 in both sexes. Norgan: Although BMI and MUAC co-vary, MUAC does give other information and could be useful. Strickland: Yes, but in the harvest time the men's arm and calf muscle could increase with all their physical work.

Contents - Previous - Next