attending two clinics in Tower Hamlets at birth, 6 months and 18 months and compared them to the Tanner-Whitehouse standards (Tanner, Whitehouse and Takaishi, 1966) and a comparable data set based on a sample o f middle-class Indian infants living in Delhi (Ghosh, Hooja, Ahmad, Acharyulur and Bhargava,1974; cited by Davies and Wheeler, 1989). Ghosh's data had already been compared by Rao, Satyanarayna and Sastry (1976) to the weights o f 18 month olds from well to do families living in Hyderabad who had found similar distributions to the data of Ghosh et al. In another Delhi study, Banik, Nay a, Krisna and Raj (1972) compared the weights o f children from rich and poor families, and found that their middle class families were slightly heavier than Ghosh's.
Davies and Wheeler found the Tower Hamlets infants approximated the 25th centile o f the Tanner-Whitehouse standards, but their mean weight was similar to those of the Indian sample which, the authors suggest, would therefore be a more appropriate set o f reference curves. The concem, however, is less to do with Asian infants typically approximating the 25th centile, but the observation that the rate o f their weight gain has been observed to decrease during the first year compared to other infants, including those in other ethnic groups.
Warrington and Storey (1988) undertook a longitudinal study to compare the growth o f children o f Asian parents with those o f indigenous white children in Rochdale during the first two years of life. The sample o f concurrent first births did include a higher proportion of Asian infants with a lower birth weight than the Caucasian sample, as reported in other studies (Brooke and Wood, 1980; Davies, Senior, Cole, Blass and Simpson, 1982). However, in contrast to other studies, there were few differences between the growth of the Asian and Caucasian infants and both were comparable to the UK standards (Tanner et al 1966).
Few studies have reported growth patterns o f infants from minority ethnic groups other than South Asian. The South London Growth and Development Study collected data on a cohort of infants living in a disadvantaged but ethnically heterogeneous area o f South London (Skuse, Wolke and Reilly, 1992). Data on the growth of the infants was monitored from birth, using data collected by the
individual health clinics or family doctors, and the children were visited at home when aged between 12 to 16 months.
The distribution o f the mean scores for weight varied by ethnicity and became more marked with age. At 3 months, when compared to the NCHS 50* centile (1977), the mean West African infants weight was +0.71 SD; the mean White UK was +0.30 SD; and the Indian Asians +0.08 SD. At nine months, the mean weight o f the West African infants was +0.42 SD.; the White UK was-0.06; Indian Asian -0.40 SD. At
12 months the differences were most marked: the mean weight o f West African infants was +0.23 SD; the White UK -.40 SD; and Indian Asian -1.44 SD.
This data set was the first to provide evidence of ethnic differences in patterns of infant weight gain that were above the expected growth of White UK population, as well as documenting a rapid decline in weight gain for the Asian infants during the last three months of the year.
Evidence for an association between feeding practices and growth comes from work by John Fell (1994) who compared a group o f Asian and White infants, matched for age, in Birmingham. The weights for length z-scores of the Asian infants were significantly lower than those o f the White UK infants. Solids were introduced significantly later to the Asian infants than the White. The infants most affected by the timing o f the introduction in solids were those starting at or after 6 months who were significantly more wasted than those starting solids before 6 months. The picture is still not clear though; Warrington and Storey also investigated nutritional
intake in their prospective study described earlier and found a weak relationship between nutrition and growth.
Fenton, Bhat, Davies and West (1989) investigated four Asian infants failing to thrive and concluded that the problem was a result of the mother's social isolation and inability to communicate and the father's refusal to accept that there was a problem in the family. They observed the mother's punitive attempts to force the child to eat. The four cases cited were similar in that the weight loss developed at the time of weaning, but the authors' diagnosis of failure to thrive is not specific, except to say that the infants were originally on or about third centile and had dropped below it. They reported that by treating the mother they were able to
improve the prognosis for the child, but gave few details o f the specific intervention offered. The authors attributed part of the source o f the problem to the shock o f acculturation. This case report is helpful in pointing out the complexity o f treating the problem of infant feeding, but highlights the dangers of stigmatising a health problem as 'ethnic' when social isolation and depression are not characteristics confined to mothers of any particular ethnic origin.
4.5 The use of culture and ethnicity as a variable
Senior and Bhopal (1994) point out that the concept of ethnicity is by no means simple. The psychological literature generally uses the term 'culture' to represent a unit which is bound by behaviours, religion, and customs, which is very similar to the definition used in the health literature for 'ethnicity'. If the concept o f ethnicity is complex, then the assessment of'ethnicity' or 'cultural belonging' will be equally fraught with difficulty
The use of ethnicity as an explanatory variable for variations in infant health is cautioned by Pearson (1991) who suggests that:
"....differentials must, however be interpreted carefully. Terminology and categories have not always been used consistently, and data have been analysed and interpreted with varying degrees of rigour insight and sophistication" (p.88).
Pearson proposed that in England and Wales there is in fact greater variation in infant health within the 'Asian' group than between mothers bom in the UK and the constituent Asian national categories. Black (1991) advises that those working with ethnic minorities should avoid stereotyping and appreciate that an individual's response and adaptation to Britain will be unique.
Ahmad (1993) reviewed the literature on the health of British Asians. He found few authors attempted to interpret their findings against the disadvantaged background in which their sample lived, thus implying that apparent differences in health were due to linguistic and cultural factors alone, rather than poverty. Given that racial, and to a lesser extent ethnic groups, are often viewed as natural groupings, then their use in describing inequalities becomes converted into a model o f explanation. Rather than observed ethnic variation prompting the study for underlying causes, it becomes the explanation. As Ahmad wrote
"effect is converted into cause".
This view is supported by Sheldon and Parker (1986) who report that it is rarely clear in studies whether race or ethnicity is being used as a social factor, indicator of culture, biological variate, or statistical proxy.