Aspects of screening and prevention of type 2 diabetes: The HUNT3 Survey and the DE-PLAN intervention in HUNT

Abstract

Summary in English Aspects of screening and prevention of type 2 diabetes The HUNT3 Survey and the DE-PLAN intervention in HUNT Background Over the last centuries, global progress has been made in economic growth, living standards and health. Some of the disadvantages of this prosperity are increased inactivity and a global dietary shift towards increased intake of energy-efficient foods. This development has led to a significant increase in the incidence and prevalence of type 2 diabetes (T2D) worldwide. In Nord-Trøndelag in 2006-2008, the prevalence of known diabetes was almost 5% among men and 3.5- 4 among women (≥20 years). Globally, the prevalence of diabetes (+18 years) has increased from 4.7% in 1980 to 8.5% in 2014, and in addition, the proportion of undetected diabetes is estimated to be 30-50%. It is important to prevent T2D, since people with poorly regulated diabetes are at risk of developing both acute and chronic complications, which leads to significant strains for both individuals and society. Clinical studies with intensive lifestyle interventions have proven effective in preventing T2D among high risk individuals, with short-term (3 years) relative risk reduction of 50-60% and 30-40% long-term relative risk reduction (7-20 years). These interventions are resource-intensive and difficult to implement on a population level. More feasible, moderately intensive lifestyle interventions in real-life settings have therefore been conducted and these have shown to be effective in reducing weight among participants. This weight reduction is believed to lead to a lower risk of diabetes. Based on these findings, the National Directorate of Health recommends that high risk individuals are offered low-threshold, multifactorial, intensive follow-up in their municipalities. However, these methods are resource-intensive and it's difficult to reach out to everyone; Approximately 10% of the population in Norway is estimated to be at high risk of diabetes indicated by FINDRISC ≥15 while only 60% of the municipalities in Norway offer the recommended intervention, with the result that <10% of high risk individuals participate in preventive measures under the auspices of “ The Healthy Life Centres” (Frisklivssentralene). Probably the situation is comparable in other countries. To prevent T2D it is important to find those at risk. A recommended risk screening tool is the Finnish Diabetes Risk Score (FINDRISC), a questionnaire with eight anthropometric questions (age, body mass index (BMI), waist circumference, physical activity, daily intake of fruit, berries or vegetables, past or present use of antihypertensive medications, previously measured high blood sugar and first or second degree relatives with diabetes mellitus), where the score ranges from 0 to 26, and a score of ≥15 is associated with a high risk of developing T2D over the next 10 years. The first studies that developed and validated FINDRISC are now 20-30 years, and globally, the incidence of diabetes has increased significantly over the years. This increase in T2D also reflects an increase in associated risk factors such as overweight and obesity. Objective Based on this background, our goal in this thesis, with material from HUNT3 (Health Survey in North Trøndelag, 2006-2008), have been to: • Estimate, by sex and age, the prevalence of FINDRISC ≥15 for identifying IFG, IGT, and diabetes, in a general Norwegian population. • Validate FINDRISC as a screening tool for diabetes risk, by estimating the 10-year cumulative diabetes incidence, by age (20-80+ years) and gender. • Inviesigate if a 2-year low-grade lifestyle intervention for prevention of T2D among high risk individuals (FINDRISC ≥15) is effective, and whether the (possible) effect varied by age, gender, educational level and depressive symptoms (This intervention was based on HUNT3 and was called "HUNT DE-PLAN" ). Method We estimated the precalence of FINDRISC ≥15 by sex and age in the HUNT3 population (50 806), and Investigated how well FINDRISC ≥15 identified IGT, IFG and diabetes in this population. Furthermore, we followed the same HUNT3 population for ~ 10 years and validated FINDRISC's ability to find those at increased risk of developing diabetes by looking at the use of diabetes medicine from 2006-08 up to 2017 (we linked our data to the Norwegian Prescription Database), as a measure for diabetes incidence. People at high risk of T2D with FINDRISC ≥15 (5297) in HUNT3 were invited to a 2 -year low-grade diabetes prevention program (HUNT DE-PLAN), consisting of clinical and biochemical measurements, questionnaires and lifestyles advices, after 0, 6, 12 and 24 months, where we examined diabetes, glycemia and obesity, based on age, gender, educational level and depressive symptoms among 2634 (49.7%) participants. Results The prevalence of elevated FINDRISC was 11.0%, 12.1% in women and 9.6% in men. At the same time, we found that FINDRISC had a relatively high PPV to detect those with abnormal glucose tolerance (IFG, IGT, diabetes) and that PPV was higher amongst men and the elderly. Over the course of ~ 10 years, 4.0% developed diabetes (13.5% of those with FINDRISC ≥15, 2,8% with those of FINDRISC <15 ) and sensitivity and specificity were 38% and 90%, respectively. Sensitivity was highest in women, whereas diabetes incidence was highest among men and in the age groups <64 years for women and 45-74 years for men. Glycemia and BMI increased during our 2-year intervention and the diabetes incidence was 10.3% during this period. The increase in glycemia and BMI was consistent across all subgroups. Conclusion Our findings, in an updated risk population, indicate that FINDRISC's sensitivity to capture those with undiscovered T2D and those at high risk of developing T2D is significantly lower than in the reference studies. Our results also show that most people who develop T2D have FINDRISC <15 and that the risk of developing T2D, if FINDRISC ≥15, is much lower than stated in official guidelines, where FINDRISC ≥15 is estimated to give a risk of 30% or more for developing T2D over the next 10 years. This makes FINDRISC less valuable for use in selected screening (as recommended in the national diabetes guidelines in Norway), since the reason to screen is to identify the population who would subsequently develop T2D if left untreated, for the purpose of including them in suitable intervention programs. Furthermore, our 2-year low-grade intervention was not sufficient to achieve a clinically beneficial effect on the risk of developing T2D, and other methods than those we used seem necessary to prevent T2D.