Study design and nature of diabetes may explain findings of Finnish studyBMJ 1997; 314 doi: https://doi.org/10.1136/bmj.314.7076.301a (Published 25 January 1997) Cite this as: BMJ 1997;314:301
- a EURODIAB, Department of Epidemiology and Public Health, University College London, London WC1E 6BT
Editor–Seppo V P Koskinen and colleagues show that the expected social class gradient in mortality is abolished in Finnish people with diabetes.1 They suggest that the most likely explanations are a lack of social class differences in health behaviours and an equitable health service. But it is difficult to see why Finnish people with diabetes adopt healthy lifestyles regardless of social class whereas non-diabetic Finnish people do not. Several studies indicate that the social class gradient in health behaviours persists in people with diabetes2 and that glycaemic control–in part a reflection of equity in access to health care–is poorer in lower social classes, even in Finnish adolescents.3
Other explanations need to be sought. Non-insulin dependent diabetes is more prevalent in people of low socioeconomic status, while there is little evidence for a social class gradient in insulin dependent diabetes.4 Thus the proportion of all diabetes that is insulin dependent will be highest in the most affluent people, and insulin dependent diabetes in this non-elderly cohort is associated with a greater risk of death than non-insulin dependent diabetes. Furthermore, patients who were treated with diet alone could not be identified as having diabetes and were therefore not included in the study population. The proportion of diabetes that is treated by diet alone is highest in the uppermost social classes. These people have a more favourable mortality experience than diabetic patients treated with drugs or insulin, and their exclusion may thus have attenuated the social class gradient in mortality.
Finally, it is clear that the development of diabetes depends on a complex interaction between genetic and environmental factors. While it is reasonable to assume that the genetic burden of diabetes is similar across social classes, the environmental trigger (usually obesity) is more common in more deprived groups. There is evidence that lean people with diabetes have a strong genetic burden of disease and are at an increased risk of complications.5 Again, these people will constitute a greater proportion of the diabetic population in the higher social classes, and this would also attenuate the social class gradient in mortality.
We suggest that there are more plausible explanations for the Finnish findings, which have more to do with study design and the nature of diabetes than with changes in health behaviours or healthcare services.