implement in resource-poor countries. Moreover, it excludes people with low-to-moderate cardiovascular disease risk who will ultimately comprise about 80% of future strokes and cardiovascular events, and thereby might have been falsely reassured that they are protected from developing these diseases. Hence, evidence is lacking for the effectiveness of the high-risk approach in preventing stroke and acute cardiovascular events at the population level (appendix p1). For more on potentially modifiable risk factors for stroke see Stroke 2020; 51: 719–28 For more on preventing dementia by preventing stroke see Alz Dem 2019; 15: 961–84 For more on Cut Stroke in Half see Int J Stroke 2018; 13: 633–47 For more on the Stroke Riskometer app see http:// www.strokeriskometer.com/ For more on combined high-risk and population-wide prevention strategies see J Am Heart Assoc 2020; 9: e014494 For more on the high-risk approach for prevention see Cochrane Dat Syst Rev 2019; 1: CD009009 See Online for appendix Figure: Key principles of the WSO Declaration on the primary prevention of stroke and dementia globally Emphasis on population-wide strategies (reduction of exposure to risk factors among the whole population and motivational population-wide prevention strategy with the control of risk factors in all people with increased stroke risk regardless of the level of risk) Abandoning categorisation of people into low, moderate, and high risk; advocating an holistic prevention approach Ideally combining community interventions (eg, health workers in low-income and middle-income countries, nurse educators in high-income countries), pharmacological (eg, polypill), and non-pharmacological (eg, lifestyle modification via the Stroke Riskometer app) interventions for people at risk of stroke 488 www.thelancet.com/neurology Vol 19 June 2020 In Context The WSO advocates for the high-risk approach to be complemented by the population-wide approach to prevention, with the emphasis on the mass approach aimed to lower the level of exposure of the entire population to environmental and lifestyle risk factors for stroke and dementia across the life course and across the continuum of stroke, cardiovascular disease, and dementia risk. There is evidence from a large cohort study that controlling just five lifestyle risk factors (smoking, physical activity, diet, alcohol consumption, weight) could reduce the risk of stroke by 47% (95% CI 18–69) in women and by 35% (95% CI 7–58) in men. The worldwide use of mobile technologies, with very high penetration even in low-income countries, offers a new farreaching interface for lifestyle modification comparable (by the coverage of the population and potential efficacy) with population-wide strategies. Internet-based interventions to modify cardiovascular disease risk through individual-level interventions (eg, the HATICE trial) have shown a significant improvement in a combined end point of systolic blood pressure, LDL cholesterol, and body-mass index (mean difference –0·05, 95% CI –0·08 to –0·01, p=0·008), but no evidence of effect on individual measures of risk. A pilot randomised controlled trial (RCT) of the use of the Stroke Riskometer app showed significant motivational value for the use of relative risk estimates for communicating stroke risk to users, high acceptability (80%), and potential efficacy of the lifestyle modification, although the lifestyle modification effect was not statistically significant. The Stroke Riskometer app is free, validated, and internationally endorsed by the WSO, World Federation of Neurology, World Heart Federation, and European Stroke Organisation. There is also evidence from large RCTs that a healthy diet and exercise coupled with cognitive training can improve or maintain cognitive function in elderly people in the general population, and that blood pressure reduction reduces the risk of incident dementia (hazard ratio [HR] 0·88, 95% CI 0·79–0·98, p=0·019) and Alzheimer’s disease (HR 0·84, 95% CI 0·73–0·97, p=0·021). The use of a combination of blood pressure and lipid lowering medications is proven to be generally safe, even in people with average or belowaverage systolic blood pressure (SBP) and cholesterol levels and, in low dosages such as in the polypill, as an adjunct therapy to other blood pressure and lipid lowering medications. A meta-analysis of RCTs that compared a polypill (including at least one anti-hypertensive and one lipid-lowering medication) with a placebo (or one active component) showed clinically significant reductions in SBP of 9·2 mm Hg (95% CI 5·0–13·4) and LDL cholesterol of 39·1 mg/dL (95% CI 25·9–5·0). Two large polypill primary prevention RCTs (appendix p1) showed a significant positive effect on either SBP (9 mm Hg reduction in the polypill group vs 2 mm Hg reduction in the usual care group) and cholesterol (15 mg/dL reduction in the polypill group vs 4 mg/dL reduction in the usual care group) or a 2·9% absolute risk reduction in cardiovascular disease events (PolyIran trial). In the HOPE-3 trial, a combination of rosuvastatin (10 mg per day), candesartan (16 mg per day), and hydrochlorothiazide (12·5 mg per day) versus usual care reduced cardiovascular disease events by 29% over about 5·6 years in adults at moderate risk of cardiovascular disease. A meta-analysis of 16 RCTs of non-communicable disease