Cardiovascular diseases (CVD) and diabetes mellitus (DM) are major public health concerns in modern society. Although mortality from CVD has decreased substantially in many European countries in the past 30 years due to better prevention (e.g. encouraging people to stop smoking and discouraging people from starting) and better medicinal treatment, CVD remains the category of diseases with the highest number of life-years lost. Due to sedentary lifestyles and unfavourable nutrition habits, obesity has developed into a new ‘epidemic’ with adverse consequences for CVD and DM. Type 2 diabetes is associated with a high risk of cardiovascular complications and mortality, with a 4-fold increased risk of dying from CVD compared to non-diabetic counterparts.1 Type 2 diabetes is now estimated to cause 1 million deaths per year worldwide.
Randomized and controlled treatment trials have shown that reducing weight and treating glucose levels in diabetic patients is effective in reducing the complications of disease.2 An important question, however, is whether we can and, if so, whether we should take more effective measures at an earlier stage of the natural progression of disease.
A screening programme for high-risk persons would allow us to detect abnormal glucose levels long before signs or symptoms would lead those persons to consult a doctor (and long before they would be diagnosed with disease). Findings of the Hoorn study demonstrated that the HbA1c in screen-detected diabetic patients was on average 2% points lower than in clinically diagnosed cases.3 This observation might suggest that with further control of CVD factors, screening for diabetes (and CVD factors) could have a beneficial effect. However, there is no evidence, e.g. from RCTs, that such an important and relatively easy public health approach would help reduce CVD complications and death. In addition, it has been shown that glucose levels may vary, and that even without treatment glucose levels may be found to have come down 1 year later (regression). Earlier detection in such individuals would actually mean overtreatment. In the Netherlands, an RCT (pilot study) on screening for diabetes has recently been executed (trial number ISCRCTN 75983009).
In addition to the question of whether we should launch screening programmes, there is the question of how these should be implemented. Recently, the Department of Health in the United Kingdom has suggested and planned to screen for and manage vascular risk in people aged 40–74 years.4,5 The journal presents a unique pilot study, which involved sending a risk factor questionnaire and a waist measurement tape to 2856 persons in the rural town of Harjavalta in Finland.6 In 1469 persons, waist circumference was also measured by professional nurses: there was only a mean 3 cm difference with the home measurements. This proves that simple self-measurements can in fact be accurate screening methods, requiring limited resources and manpower.
Persons at high risk were then invited for further screening tests involving an oral glucose tolerance test, blood pressure measurements, lipids analysis and anthropometric variables. In the results it is difficult to see the positivity overlap between the tests per person, but the most striking fact is the remarkably high percentage of persons at risk. Of all initial respondents (response rate of 73%), 84% had at least one pre-specified risk factor. Impaired glucose tolerance was detected in 193 (13%) and newly diagnosed diabetes in 65 (4.5%) of persons examined. This strategy clearly identifies a large proportion of yet undetected high-risk persons who may benefit from consecutive early treatment/intervention.
The authors conclude that the criteria for metabolic syndrome (from the International Diabetes Federation or the National Cholesterol Education Program Adult Treatment Panel III) or the Finnish Diabetes Risk Score (FINDRISC) questionnaire are more efficient tools for the selection of persons for further risk stratification in general practice. The FINDRISC lead to 697 positives (‘referrals’) with 41 diabetes patients detected.
What the authors do not say, however, is that waist circumference measurement alone (women ≥88 cm/men ≥102 cm) as the initial screen test is even more efficient: it resulted in 50 detected diabetes patients out of only 681 positive referrals.
In conclusion, this study shows there is great potential in reaching high-risk populations who are in a screen-detectable preclinical phase, not yet known to the health care system, and who may benefit from earlier intervention if it is followed up by high-quality and effective measures. This is a much more attractive strategy than case-finding in people consulting a doctor, as it enables a large population to take the first steps themselves, e.g. by measuring waist circumference.
The challenge of course is to reach that large population across a broad spectrum of socio-economic and ethnic groups, and to unambiguously show that such approaches are effective. Twenty years ago, Geoffrey Rose seemed, in retrospect, to divide public health approaches between primary prevention (‘good for all’) and secondary prevention (‘good for some’).7 With hindsight, this view can be ascribed to the technology, the health care system and possibly the lack of focus on primary prevention that characterized the sector in those days. However, we have progressed since then and based on current insights, a large part of the future generation might benefit from approaches targeting high-risk groups, as reported in this journal, provided that the effectiveness of such approaches is relatively high for a substantial part of the population. Annual home testing may become part of our everyday life.