Citation

Andersson R, Moniruzzaman S. Public Health 2005; 119(1): 72-73.

Affiliation

Division of Public Health Sciences, Department of Social Sciences, Karlstad University, Karlstad SE 651 88, Sweden.

Copyright

(Copyright © 2005, Elsevier Publishing)

DOI

10.1016/j.puhe.2004.10.003

PMID

15560906

Abstract

We thank Dr O.B. Coles for his interest in our study on 'economic development and suicide mortality'.

Health transition theory offers a framework for the understanding of major changes in disease magnitudes and patterns over time. These shifts are propelled by a number of interrelated determinants, such as economic development, demography, culture, socioeconomic and gender inequalities, health care, and so on. Recent research suggests that economic development plays a fundamental role as a driver of general health improvements globally.

As pointed out by Beaglehole and Bonita in their criticism of health transition theory, little is said about injury in health transition terms. Most analysis and attention are focused on historical shifts from communicable to non-communicable disease patterns. They also criticise the theory, in its current shape, for being less useful in predicting future health patterns. An interesting contribution to that end, however, suggested declining rates among all degenerative and man-made 'diseases of civilisation' (cancer, cardiovascular disease and injury) already before this became evident from statistics. Moreover, the same source projected a more psychosocially determined health pattern among richer countries in the future.

Our study on suicide as a function of economic development is part of a broader research program on global determinants of injury, so far with special (and separate) attention to the roles of economic development and culture. Cross-sectional relationships between injury and economic development have been analyzed in seven papers, covering all types of injury (intentional and nonintentional) and all age/gender groups. In summary, our results point at clearly declining injury mortality rates by economic development from middle-income to high-income levels, while increasing rates are often seen from low-income to middle-income levels. Striking deviations, however, were found in unintentional injuries among the young and the elderly, in homicide among small children, and in suicide, especially among women. These findings also show some consistency with longitudinal data we have come across from various countries. This is something we are planning to investigate closer and more systematic in the near future. Our results are also partly consistent with the above-mentioned projection insofar that most injury categories follow trends of cancer and cardiovascular disease, while categories which are more social by nature (linked to risk behaviour in adolescence, aging populations, mental illness, and so on) seem unaffected or positively correlated to economic development. This implies that injury is not a homogeneous phenomenon from a health transition perspective.

Of course, cross-sectional analyses of mortality data from various countries derived from different collection, classification and recording systems must be interpreted with great caution, which is strongly underlined in our papers. Another difficulty is that many middle-income countries in our studies belong to the newly independent states in eastern and central Europe,which all shared extreme health and economic conditions during the 1990s. Therefore, our conclusions are also generally conservative in our papers. Nevertheless, the striking differences between categories we have seen clearly call for deeper analysis and clarification. That is why we have labelled our approach explorative.

Unfortunately, in this particular case, the Editors of Public Health felt that the longer paper we had prepared was not suited to the Journalâ??s style, but that our findings could be better suited to a short communication. When reediting our paper to the shorter format, we obviously lost important parts of our detailed results and discussion, which may have caused confusion. We apologize for that and hope to get a chance to publish an improved full version elsewhere. We have got similar comments related to other papers, and we do acknowledge that the theoretical implications of our aggregate results need serious and comprehensive discussion. We are also planning to come back with a separate paper on these theoretical considerations.