Pictorial presentation of subclinical atherosclerosis : a measure to reduce the risk for cardiovascular disease

(1)

Pictorial presentation of subclinical

atherosclerosis

A measure to reduce the risk for cardiovascular disease

Anna Bengtsson

Institution for Epidemiology and Global Health and the Institution for Public Health and Clinical Medicine

(2)

Detta verk är skyddat av svensk upphovsrätt (Lag 1960:729) Avhandling för filosofie doktorsexamen

ISBN print: 978-91-7855-592-5 (print) ISBN PDF: 978-91-7855-593-2 (PDF) ISSN: 0346-6612

Information om layout / omslagsfoto / sättning: Inhousebyrån Umeå universitet Elektronisk version tillgänglig på: http://umu.diva-portal.org/

Tryck: Cityprint i Norr AB Umeå, Sweden 2021

(3)

(4)

(5)

Abstract

The overall aim was improved cardiovascular disease (CVD) prevention through the VIPVIZA intervention. This includes the provision of pictorial information of subclinical atherosclerosis to participants and their general practitioners (GPs), follow-up phone call, and written information to the participant. VIPVIZA is a Prospective Randomized Open Blinded End-point (PROBE) trial nested within VIP, a CVD prevention program in primary health care in Västerbotten county, Sweden. Middle-aged individuals at low/intermediate CVD risk were enrolled to VIPVIZA and randomized 1:1 to an intervention (n=1749) or control group (n=1783 who received no pictorial or other information). Preventive measures were managed within primary health care.

At baseline, clinical risk factors were measured and carotid ultrasound examination was performed. The prevalence of subclinical atherosclerotic disease was assessed as intima media thickness and presence of plaque. The association between clinical risk factors and measures of subclinical atherosclerosis was investigated. In addition to conventional risk factor-based risk evaluation, the impact of the VIPVIZA intervention on CVD risk, traditional risk factors and pharmacological treatment was evaluated after 1 and 3 years. Individual interviews were conducted with 15 GPs to explore how a pictorial representation of subclinical atherosclerosis affects physicians in their perception and communication of CVD risk. The interviews were analyzed by qualitative content analysis.

The plaque prevalence was 44.7% in this population. Clinical risk factors explained more of the variation in a combined ultrasound measurement than single measurements. The results up to three years showed a VIPVIZA intervention effect, with lower and sustained CVD risk in the intervention as compared to the control group. The effect was partly mediated by differences in intake of lipid-lowering medication and partly by lifestyle behaviour. The GPs described their risk assessment and patients’ risk perception as more accurate with the VIPVIZA intervention. Informing patients about examination results prior to a consultation can facilitate shared decision-making and enhance adherence to preventive measures.

The results show that the VIPVIZA intervention reduces CVD risk over three years. In the long run this has the potential to reduce the incidence of CVD events.

(8)

Original Papers

This thesis is based on the following papers:

I. Anna Bengtsson, Emma Nyman, Christer Grönlund, Per Wester, Ulf

Näslund, Eva Fhärm, Margareta Norberg. Association of cardiovascular disease risk factors with combined carotid ultrasound measurements in subclinical atherosclerosis. In manuscript

II. Anna Bengtsson, Kristina Lindvall, Margareta Norberg, Eva Fhärm.

Increased knowledge makes a difference! – general practitioners’ experiences of pictorial information about subclinical atherosclerosis for primary prevention: an interview study from the VIPVIZA trial.

Scandinavian Journal of Primary Health Care 2021;39:77-84

III. Ulf Näslund, Nawi Ng, Anna Lundgren*, Eva Fhärm, Christer Grönlund,

Helene Johansson, Bernt Lindahl, Bertil Lindahl, Kristina Lindvall, Stefan K Nilsson, Maria Nordin, Steven Nordin, Emma Nyman, Joacim Rocklöv, Davide Vanoli, Lars Weinehall, Patrik Wennberg, Per Wester, Margareta Norberg. Visualization of asymptomatic atherosclerotic disease for optimum cardiovascular prevention (VIPVIZA): a pragmatic, open-label, randomised controlled trial. Lancet 2019;393:133-142.

IV. Anna Bengtsson, Margareta Norberg, Nawi Ng, Bo Carlberg, Christer

Grönlund, Johan Hultdin, Bernt Lindahl, Bertil Lindahl, Steven Nordin, Emma Nyman, Patrik Wennberg, Per Wester, Ulf Näslund. The beneficial effect over 3 years by pictorial information to patients and their physician about subclinical atherosclerosis and cardiovascular risk: Results from the VIPVIZA randomized clinical trial. Accepted by American Journal of Preventive Cardiology

*Anna Bengtsson was named Anna Lundgren when paper III was published.

(9)

Abbreviations

BP Blood pressure

CAC Coronary artery calcium

CCA Common carotid artery

cIMT Carotid intima media thickness

CRC Clinical research centre

CT Computed tomography

CVD Cardiovascular disease

DBP Diastolic Blood Pressure

DM Diabetes Mellitus

ESC European society of Cardiology

ESH European society of Hypertension

FRS Framingham risk score

GP General practitioner

HD Health Dialogues

HDL-C High density lipoprotein-cholesterol

ITT Intention-to-treat

LDL-C Low density lipoprotein-cholesterol

PHC Primary health care centres

PLS Partial Least Square

(10)

RCT Randomised control trial

SBP Systolic Blood pressure

SCORE Systematic Coronary Risk Estimation

TG Triglycerides

US Ultrasound

VIP Västerbotten Intervention Programme

VIPVIZA VIsualiZation of asymptomatic Atherosclerotic disease for optimum cardiovascular prevention – a randomized controlled trial nested in the Västerbotten Intervention Program

(11)

Enkel sammanfattning på svenska

Bakgrund

Hjärtkärlsjukdom, vilket inkluderar både stroke och hjärtinfarkt är den vanligaste dödsorsaken i Sverige och i världen. Den huvudsakliga orsaken till hjärtkärlsjukdom är åderförkalkning. Åderförkalkningsprocessen börjar tidigt, redan i barndomen, och fortsätter sedan genom livet. Studier har visat att upp till 95 % av all hjärtkärlsjukdom orsakas av påverkningsbara faktorer. För att förhindra hjärtkärlsjukdom är det viktigt att både individer och läkare följer de förebyggande riktlinjer som finns. Dessvärre misslyckas detta arbete ofta eftersom följsamheten till råd om förändrad livsstil och förskrivning av läkemedel inte följs. Det är också en stor utmaning att hitta individer med risk för hjärtkärlsjukdom för att i tid kunna sätta in insatser för att förebygga insjuknande. Detta är viktigt eftersom i ca 50 % av fallen är första symtomet av åderförkalkningssjukdom ett livshotande tillstånd så som hjärtinfarkt eller stroke. Den metod som rekommenderas i dag (SCORE) är en uppskattning av risken att dö inom 10 år av hjärtkärlsjukdom baserat på riskfaktorer. Dessvärre missar denna metod många individer, endast 1 av 3 som insjuknar med hjärtkärlsjukdom hade klassificerats som hög risk innan insjuknandet. Det är många som inte heller känner till det starka samband som finns mellan riskfaktorer och hjärtkärlsjukdom. Om kunskapen kring dels sin egen risk men även sambandet mellan risk och riskfaktorer ökade kanske individer skulle bli mer motiverade till beteendeförändringar. Att visa en bild av individens åderförkalkningssjukdom skulle kunna öka kunskapen och på så sätt kunna leda till förändringar som skulle kunna minska riskfaktorbelastningen.

Metod

Vi har med en enkel och ofarlig ultraljudsundersökning av kärlen på halsen, undersökt om tidiga tecken till åderförkalkning föreligger. Deltagarna till VIPVIZA (som denna studie kallas) rekryterades från Västerbottens Hälsoundersökningar (VIP-Västerbottens Intervention programme), ett program för att förebygga hjärtkärlsjukdom, som erbjuds västerbottningarna det år de fyller 40, 50 och 60 år. Från VIP rekryterades 3532 deltagare som antingen var 40 år gamla och hade ett syskon eller förälder som drabbats av hjärtkärlsjukdom innan 60 års ålder, 50 år gamla och en riskfaktor för hjärtkärlsjukdom eller 60 år gamla. För att undersöka om vetskapen om förekomst av åderförkalkning påverkade deltagarna slumpades hälften av deltagarna till att få beskedet från undersökningen och andra hälften fick inte reda på beskedet. De som fick resultatet fick det hemskickat med post, beskrivet dels som en bild liknande ett trafikljus där rött visade förekomst av åderförkalkning och grönt visade att det inte fanns åderförkalkning. Med i

(12)

beskedet fanns också bild på en mätare som gick från grön via gul och orange till röd för att visa hur tjockleken på deltagarens kärl var i förhållande till deltagarens ålder. Grönt indikerade att deltagarens kärl motsvarade en individ som var minst 10 år yngre, och rött indikerade att kärlen motsvarade en individ som var mer än 10 år äldre. Med i beskedet fanns också information till deltagarna om att åderförkalkningsprocessen är påverkningsbar och att man genom en sund livsstil samt följa ordinationer av läkemedel kan minska sin framtida risk att drabbas av hjärtkärlsjukdom. Samma besked som skickades till deltagarna skickades till deras distriktsläkare. Ungefär 2 veckor efter att beskedets skickats hem till deltagarna blev de uppringda av en forskningssköterska för att diskutera resultat. Deltagarna fick sedan under de kommande 3 åren total 4 uppföljande brev med information om studien och generella råd om hälsosam livsstil. Vi följde sedan upp alla deltagare, både de som fått ta del och de som inte fått ta del av ultraljudsresultatet, 1 år och 3 år efter ultraljudsundersökningen. Vi kunde då jämföra de som fått beskedet och de som inte fått beskedet i beräknad risk för hjärtkärlsjukdom, nivåer av riskfaktorer och läkemedelsanvändning förändrats från undersökningstillfället till efter 1 och 3 år. Alla deltagare som visade allvarlig åderförkalkning fick reda på beskedet och remitterades till specialistvård (22 personer). En ny ultraljudsundersökning genomfördes igen efter 3 år och då fick alla deltagare reda på resultatet.

Till de deltagare som fått beskedet skickades också precis samma information till deras distriktsläkare. För att undersöka läkarnas upplevelse av att ta del av beskedet intervjuades 15 distriktsläkare.

Resultat

Det är vanligt med åderförkalkning i en medelålders befolkning med låg till måttlig risk för hjärtkärlsjukdom, 51 % av männen och 39 % av kvinnorna hade åderförkalkning i sina kärl på halsen. Av alla 60 åringar hade ca 50 % av kvinnorna och 60 % av männen plack (åderförkalkning). Vi kunde också se att i en modell där alla olika ultraljudsmått relateras till de olika riskfaktorer vi har mätt så förklarade riskfaktorer en större del av variationen i ett kombinerat ultraljudsmått jämfört med enbart en mätning på ett ställe av kärlet. Även om ett kombinerat mått gav en högre förklaringsgrad, förklarade riskfaktorer bara ca en fjärdedel av variationen.

Vidare fann vi att deltagarna som fått ta del av beskedet från undersökningen hade lägre SCORE-värde (mindre uppskattad risk att dö i hjärtkärlsjukdom inom 10 år) efter både 1 och 3 års uppföljning. De enskilda riskfaktorer där det var en statistiskt signifikant skillnad mellan grupperna var midjeomfång och nivåer av total kolesterol och det skadliga LDL-kolesterolet. Värdena var lägre bland de som fått beskedet än de som inte fått beskedet efter både 1 och 3 år.

(13)

De intervjuade läkarnas upplevelse av VIPIZA var att deltagarna hade större kunskap kring sin egen åderförkalkningssjukdom och även om sjukdomen generellt jämfört med patienter i övrigt. Med ökad kunskap blev patienterna mer delaktiga i sin behandling och läkarna upplevde att patienternas följsamhet till förebyggande åtgärder ökade. Vissa läkare ägnade beskedet ingen uppmärksamhet medan andra agerade aktivt genom att kontakta och behandla patienterna. Det generella intrycket var att läkarna var säkrare på hur de skulle handlägga ultraljudsresultaten senare i studien. Undersökningsresultatet från VIPVIZA upplevdes som tydligt, ”det var svart på vitt” inte bara en siffra.

Slutsatser

Åderförkalkning är vanligt hos både män och kvinnor med låg till måttlig risk för hjärtkärlsjukdom och förekomsten ökar med ålder, manligt kön och vid riskbedömning enligt traditionell riskbedömning. Riskfaktorer för hjärtkärlsjukdom är till högre grad associerad med ett kombinerat ultraljudsmått än ett enskilt mått som till exempel kärlets väggtjocklek på vänster sida. Dessa resultat indikerar att undersökning av kärlen på halsen på bägge sidor och med många projektioner med högre precision kan fånga den tidiga, fläckvisa åderförkalkningen. Distriktsläkare beskrev deras riskbedömning och patientens riskuppfattning som mer korrekt med VIPVIZA. Attityden hos läkarna till resultatet och handläggning av riskfaktorer varierade. Att informera patienter om undersökningsresultat innan kontakt med sin läkare kan förbättra och underlätta samtalet samt öka följsamheten till förebyggande åtgärder.

VIPVIZA interventionen (med bild på individens åderförkalkning till deltagare och läkare, uppföljande telefonsamtal samt informationsbrev) gav, jämfört med att inte få informationen, en minskad risk för hjärtkärlsjukdom upp till minst 3 år. Dessa resultat indikerar att detta arbetssätt har potentialen att minska insjuknandet av hjärtkärlsjukdom över tid.

(14)

(15)

Background

Cardiovascular disease and atherosclerosis

Cardiovascular disease (CVD) which includes ischemic heart disease and stroke, is the leading cause of disability and death worldwide(1). In 2017, 17.7 million out of a total of 55 million deaths derived from CVD and by 2030 the number of deaths due to CVD is estimated to grow to more than 23.6 million (2,3). On an individual level, the risk of developing future CVD when an individual has reached the age of 45 is over 60% for men and 55% for women (4). The costs of treatment and care are immense and estimated at $47 trillion worldwide in the next quarter century if there is no increased cardiovascular prevention (5). In fact, a 10% increase in hypertension treatment or cholesterol improvement could prevent 14,000 and 8000 deaths, respectively, in the USA alone (6). Nevertheless, since the 1970s there has been a reduction in age-adjusted CVD mortality (7). This decline is explained by implementations of tobacco restriction, public health initiatives, and more effective measures taken by medical and care providers. However, in recent years this trend has flattened, and the CVD mortality has actually increased in some countries and certain populations (7,8). These epidemiologic trends are associated with an increase in the prevalence of major risk factors, such as obesity, hypertension and diabetes. Even though in Sweden the age standardized death rates for CVD have decreased over the last 22 years, to a greater extent in men than in women, nevertheless in 2019 CVD was still the leading cause of death and constituted approximately 31% of all deaths in both sexes (9). (Figure 1)

Figure 1. Death rates from CVD, age standardized according to the average population in 2019, for males and females aged ≥35 years in Sweden from 1997 to 2019.

(16)

One serious concern, however, is that the declining trend of CVD death rates in Sweden will not continue, as the proportion of Swedes who are overweight or obesity is slowly increasing. In 2006 46% of the Swedes aged 16-84 years was estimated to be overweight or obese as compared to 2020 when the equivalent proportion was 51%. There are clear socioeconomic associations: in 2020 the estimated proportion of Swedes who were overweight or obese with the lowest level of education (up to 9 years of schooling) was 63% as compared to 47% in the group with the highest level of schooling (more than 12 years) (10). The previous observed trend in smoking cessation has also haltered: in 2006 14% of Swedes aged 16-84 years were estimated smokers and the proportion of smokers declined to 7% in 2018 but during 2018 up to 2020 the proportion smokers is unchanged. Based on these data it is possible that the decline in CVD deaths rates observed over the last 20 years in Sweden will level off or possible even reverse.

Atherosclerosis is by far the most important cause of CVD (8). Atherosclerosis is a slow, progressive inflammatory disease of the arteries (11). In early atherosclerosis, the changes in the arteries are subtle and heterogeneous, and may present as a minor focal thickening of the arterial wall, plaque deposition. These small early changes can be difficult to capture if only a small segment of an arterial wall is examined. The atherosclerotic process progresses over time, from subclinical atherosclerosis to clinical disease. Determinants for the process exists already in the foetal life (12) and the underlying pathophysiologic processes of CVD continues with an additive effect throughout life. Preventive actions early in life may therefore have great influence over time. This has specifically been shown for the cumulative exposure of low-density lipoprotein-cholesterol (LDL-C) (13). Reduction of risk factors can curb the progression of atherosclerosis and decrease the risk of CVD events (3,14). The atherosclerotic process is in most cases silent and continues undetected until the occurrence of a potentially lethal event such as myocardial infarction or stroke: this is the case in 40-60 % of cases (15). Detection of early subclinical atherosclerosis to identify individuals at increased risk of CVD may enables early risk-reducing actions to prevent or delay future CVD development.

The prevalence of subclinical atherosclerosis as assessed as plaque in the carotid arteries measured with ultrasound examination differs largely among populations. In the ARIC study (16) 39.6 % of the participants within the age group of 55-59 years showed the presence of plaque, as compared to the REFINE-Reykjavik study with a plaque prevalence of 70.3% among men and 54.5% among women within the same age group (17). A recent Swedish population based cross sectional study, SCAPIS, showed a plaque prevalence of 56% (18) in a population with an average age of 57.4 years. However, comparing studies regarding prevalence of plaque is challenging because of different plaque definitions and inconsistent protocols for measuring the arterial wall thickness (19,20). Population statistics often lack detailed study population characteristics, and therefore specific interaction and determinants in the development of atherosclerosis are difficult to evaluate when comparing populations (21). Moreover, populations with different degrees of atherosclerosis may not be comparable: most studies are performed on advanced atherosclerosis, suggesting that more studies on early atherosclerosis are needed.

(17)

Risk factors for CVD

Globally nine modifiable risk factors account for more than 90% of the population-

attributed risk for acute myocardial infarction (MI) (3) and stroke (14,22). At both an individual and populational level, the risk for MI is mostly contributed by hyperlipidaemia and smoking, while the risk for stroke is mostly contributed by hypertension, as shown in Figure 2 (3,22).

Figure 2. Odds ratio and Population attributed risk for modifiable risk factors attributing to Myocardial Infarction (MI) and Stroke, globally.

s

Data from Yusuf S et al. Lancet 2004;364:937-52 and O'Donnell MJ et al. Lancet 2016;388:761-75.

To determine which risk factor contributes most to CVD development is difficult and depends of whether individual or population risk is considered. Furthermore, the risk factor distributions differ between populations. In Western Europe, abdominal obesity, as

(18)

measured by waist to hip ratio, is the primary risk factor for MI and the second most prominent risk factor for Stroke (Figure 3)

Figure 3. Odds ratio and Population attributed risk for modifiable risk factors attributing to Myocardial Infarction (MI) in Western Europe and Stroke in Western Europe, North America and Australia.

Data from Yusuf S et al. Lancet 2004;364:937-52 and O'Donnell MJ et al. Lancet 2016;388:761-75.

The impact of risk factors is also dependent on how many risk factors an individual is exposed to, the amplitude and time of exposure (4,23). In average, a 45-year old with all risk factors optimized lives 14 years longer without a CVD event than someone with two or more risk factors (4). A major non-modifiable risk factor is age, with the disease risk doubling with every decade after 40 years (24). Sex is also a determinant which interacts with risk factors and affects CVD risk. For example smoking and diabetes contributes to a higher CVD risk in women than in men (25), while for myocardial infarction, age- standardized incidence and mortality was twice as high in men as in women in Sweden in 2019: as many women died of CVD as men but at an older age (9).

(19)

Primary prevention of CVD

Smoking cessation, weight reduction, lipid lowering and antihypertensive medication to prevent CVD are among the most cost-effective and evidence-based treatments in health care. However, in the real world, primary prevention often fails due to low and non-sustained adherence to prevention guidelines by practitioners and individuals (26,27). Primary prevention of CVD is concerned with haltering, or even reversing, the processes of atherosclerosis to prevent or postpone clinical events. It is important to find individuals at risk at an early stage of the disease to have the potential to reduce risk factors and subsequently reduce or postpone future CVD events, thereby improving population health, reducing cardiovascular disease care costs and reducing individual suffering and premature death. The value of primary preventive care was recognized by the World Health Organization (WHO) who established a set of 10 “Best Buys” focusing largely on low-cost preventive measures to address the growing global burden of CVD (28).

The risk factor which causes the highest individual risk for stroke in the Western world is smoking (Figure 3A). Smoking cessation is the most cost-effective strategy for CVD prevention. A lifetime smoker has a 50% probability of dying from smoking and will, on average, shorten their life by 10 years (29). However, on a population level the risk factor causing the highest proportion of risk for stroke in the western world is hypertension (Figure 3C). Hypertension can be treated with two well-established strategies: lifestyle changes and medical treatment. Lifestyle changes can lower the blood pressure (BP) but many patients with hypertension also need medical treatment. Reducing the systolic blood pressure (SBP) by 10 mmHg or diastolic blood pressure (DBP) by 5 mmHg is associated with reduction in incidence of stroke of 35% and coronary events of 20% (30,31). The 2018 guidelines issued by the European society of Cardiology (ESC)/European society of hypertension (ESH/ESH Guidelines for the Management of Arterial Hypertension) (32) recommends that when BP-lowering medication is used, the primary aim is to lower BP to <140/90 mmHg in all patients. Given that the treatment is well accepted, treated BP values should be targeted to 130/80 mmHg or lower in most patients. SBP should not be targeted to less than 120 mmHg.

The risk factor contributing to the highest individual and population risk for MI in Western Europe is abdominal obesity (Fig 3 B and D). Dietary change is a key factor in reducing obesity, diabetes mellitus type 2, hypertension and dyslipidaemia. The Dietary Approaches to Stop Hypertension (DASH) diet (33), Mediterranean diet (34,35), and vegetarian diet (36) have a strong evidence base for CVD prevention (37). Furthermore, approximately 23% of the world’s population is estimated to be insufficiently active (38). Physical activity is recommended for all women and men all through life, with a minimum of 150 to 300 min/week of moderate activity or at least 75 to 150 min/week of high intensity activity.

Although any activity is better than none, and more activity is better than some (39).

Moreover, a sedentary lifestyle is one of the major risk factors for CVD, independent of participation in physical activity (40).

(20)

Globally, hyperlipidaemia is the risk factor which contributes the highest individual and population risk for MI (Figure 2B & 2D). Numerous randomised controlled trials (RCT), and epidemiological studies have shown a relationship between changes in plasma LDL-C and the risk of atherosclerotic cardiovascular disease (41-43). Furthermore, the achieved reduction in lowering LDL-C proportionally reduces the risk of disease (13,44,45). An exposure to lower LDL-C levels for a long time is associated with a lower risk of CVD in comparison to shorter-term exposure to lower LDL-C (46,47). These data support the concept that LDL-C have both a cumulative and causal effect on the risk of CVD. There is no evidence that there is a lowest level of LDL-C in which the benefit disappears or harm starts to occur. The individual variability in the response to medical and dietary treatments on LDL-C levels show a great variability (48). Poor response to treatment with statins seems to some extent be caused by poor compliance, but might also be explained by genetic factors (49,50). The benefits associated with the LDL-C reduction are not specific for treatment with statins (51). To patients who do not tolerate the recommended dose of a statin, the addition of a non-statin lipid-modifying substance to the maximum tolerated dose of the statin is recommended (52,53). The dietary component with the largest impact on LDL-C levels is saturated fat (54). Unsaturated fat-rich oils are shown to reduce LDL-C levels when used in substitution of foods rich in saturated fats, such as butter (55). Reduction of body weight also influences total-cholesterol (TC) and LDL-C levels, but the effect is small (56,57). The reduction of levels of LDL-C caused by regular physical activity is even smaller (58,59). Nevertheless, the gains of reduction in weight and increase in physical activity on the CVD risk profile is likely to occur because of their impact on other risk factors, especially hypertension and diabetes. According to the 2019 ESC guidelines (60) patients at very high CVD risk are recommended an LDL-C reduction of >50% from baseline and an LDL-C goal of <1.4 mmol/L while for low-risk individuals a goal of <3.0 mmol/L may be considered. Presence of carotid plaque is considered as very high CVD risk. Globally, alcohol consumption is a risk factor contributing to an increased risk at both the individual and populational level for both stroke and MI (Figure 2 A-D), although this pattern is not seen in western countries. The relationship between CVD and alcohol consumption is complicated, depending on both the level and pattern of consumption. Although potentially beneficial in small doses, alcohol is harmful and may be lethal in

larger doses; heavy drinking, (>4 drinks/day) is associated with an increased risk for death

and CVD. Individuals should limit their daily intake to one drink/day for women and two

drinks/day for men. (61,62).

Today we are aware of the major risk factors for CVD, and we also know how to reduce them. The great challenge is adherence to recommendations. How can reach individuals and health care providers to increase the adherence to current preventive recommendations and thereby reduce the risk for CVD?

(21)

Opportunities in prevention of cardiovascular disease

For optimal prevention of CVD each step of the prevention stair (Figure 4) needs to be optimized.

Figure 4. The prevention stair of cardiovascular disease. Each step represents one opportunity for improved primary prevention of cardiovascular disease.

Actions are needed at both the population level and the individual level to continue to reduce CVD, the major cause of mortality in Sweden and the World. If this constant preventive work does not continue, there is a risk of increased mortality, partly as a consequence to the growing trend of obesity observed in Sweden and globally. There are several opportunities on every step of the preventive stair.

Risk factor reduction

The first step in prevention of CVD is to prevent the development of risk factors on a population level, known as primordial prevention (63). To be able to improve the health of the population, behavioural interventions and actions to reduce adverse social and environmental components are needed (64). Essential to primordial prevention is the promotion of sustainable, healthy lifestyles and environments in society to avoid the development of risk factors (65,66).

When applying the population level approach, great wins can be achieved by small behavioural changes on a large population, according to the Geoffrey Rose preventive paradox (67). A small reduction of a risk factor on a large population with low to moderate risk for disease has a greater impact than a great reduction on those individuals at very high risk (Figure 5). Population-based strategies also have the potential to reduce inequalities of health.

(22)

Figure 5. A small reduction in risk factors reduces the progression of atherosclerosis, the numbers of individuals with most severe progression of atherosclerosis is significantly reduced as the curve moves to the left.

To reach a population and change the context in which individuals interact, to make the healthier choice the “default choice” national and local authorities needs to be involved. When targeting a population, performing RCTs are difficult and generally not feasible. However, there are several studies which provide evidence for population level approaches for preventing CVD, even without RCTs. According to 2016 European Guidelines on cardiovascular disease prevention in clinical practice (68) there are class I recommendations (evidence and/or general agreement that a given treatment is beneficial, useful, effective) for population-based approaches to improve diet, increase physical activity and reduce smoking and harmful drinking.

Risk assessment

The second step in the Preventive stair model, but the first individual step, is “Risk assessment”. It is essential to assess the risk to adequately act on the risk. If the assessed risk is high, more preventive actions need to be taken as compared to a low risk. However, many individuals are unaware of their risk. This is highlighted in a multinational study with a population of 14042 participants in which 46.5% of the participants with hypertension were aware of the diagnosis (69). Screening could be one measure to improve detection of risk factors and an opportunity to reach more individuals for assessment of their CVD risk

Since atherosclerosis is usually the result of risk factors interacting, rather than the effect of one single risk factor, assessment of risk should therefore be based on the individual’s total CVD risk. There are several risk assessment tools available for use for primary prevention of CVD, including Systematic Coronary Risk Estimation (SCORE) (70) and the Framingham risk score (FRS) (71). Most risk estimation tools are similar, the difference

(23)

between them being primarily nature of the population from which the estimates were derived. The intensity of remedial actions, promotion of lifestyle modification and medical treatment should be based upon an individual´s estimated risk of atherosclerotic cardiovascular disease. The greatest benefit measured in absolute risk reduction, is accomplished in individuals with the highest risk.

The SCORE system

The SCORE system is the recommended risk assessment tool in Europe (68). It estimates the 10-year risk of the first fatal atherosclerotic event based on 5 risk factors, age, sex, present smoking, systolic blood pressure and total-cholesterol. The risk categories for SCORE are low-risk (<1%), moderate risk (≥1% and <5%), high risk (≥5% and<10%) and very high risk (≥ 10%). Individuals can be categorised into high and very high-risk categories based on factors other than SCORE. Markedly elevated single factors e.g. cholesterol >8 mmol/L or BP 180/110 mmHg, most people with diabetes mellitus (DM) and moderate chronic kidney disease are directly categorised in to high risk. Individuals with documented atherosclerotic disease, DM with organ damage or a major risk factor, or severe chronic kidney disease are categorised as very high risk. The intensity of preventive actions should increase with an increase in estimated risk. In addition to risk factors included in the risk factor models, and factors that categorise individuals into high or very high risk, there are additional factors to consider; family history of premature CVD, metabolic syndrome, chronic autoimmune disease, history of premature menopause, history of adverse pregnancy-associated outcome, high-risk race/ethnicity and high-risk biomarkers (72). Clinical judgement in performing a risk assessment is essential (73) and an individual approach where all potential risk factors are considered is necessary. Some advantages to using the SCORE risk chart include its ease of use, a common language used among health professionals and an objective multifactorial assessment of risk.

When using the risk estimate models, one has to remember that the data used for each calculated model is based on historic cohorts to estimate future risk. Even if the charts are continuously updated, the data used might not mirror the situation in the population at the moment, with a risk of either under- or over-estimation of the future risk. SCORE Sweden 2015 was evaluated in 2017 by Karjalainen et al and was shown to predict more accurately the number of deaths in the different age and risk categories than the older version of SCORE from 2003 (74). However, some age groups are not well covered in the risk models, younger and older patients. For younger patients a time frame of 10 years is short and may lead to underestimation of lifetime risk, in contrast to older patients where the time frame of 10 years is long and may lead to over estimation.

To improve the accuracy of risk prediction in older adults (>65 years), a new SCORE was

recently presented, SCORE2-OP (older people) (75). The established SCORE system estimates risk for fatal CVD, this can be misleading and the risk for non-fatal but yet disabling disease is underestimated. In June 2021 a new up-dated SCORE-model was presented, SCORE2. This new model provides estimates for the 10-year risk of both non-fatal and non-fatal CVD based on the most current and representative CVD rates available in Europe(76). This new model has the potential to overcome some of the shortcomings of the SCORE-logarithm. In addition, the SCORE system is adapted to estimate risk in different European populations not different ethnic groups within these populations.

(24)

Furthermore, the evidence for use of risk scores actually translates into a reduction of morbidity and mortality is limited (77).

When dividing individuals in risk categories based on risk calculation charts, most preventive attention is given to individuals at high or very high risk for CVD, since these individuals have most to gain from preventive actions. However, a large proportion of CVD events, 60-70%, occur in the low-intermediate risk group, a group in which adequate prevention is often overlooked (78-80). SCORE has a low specificity, with both under- and over-estimates of risk of future CVD. Furthermore, no randomized trials have shown improved results of CVD prevention in the intermediate risk group when statistical risk evaluation based on conventional factors is applied (81).

Among physicians (family physicians, general internists and cardiologist) considering primary prevention of CVD, only 41% use the recommended CVD risk estimates, even if the knowledge of these tools is high (92%). The main reasons for not using CVD risk estimate were “it is too time consuming” and” I do not find it useful in practice” (82). The variability in decisions and approaches to CVD prevention among general practitioners (GPs) is diverse (83). Furthermore, studies indicate that GPs prefer to make their own judgement of individual risk factors based on their experience, rather than using absolute risk assessment tools (83).

Visualize subclinical atherosclerosis to improve prediction of CVD risk

An alternative strategy to estimate CVD risk is to visualize the degree of actual atherosclerotic disease in combination with traditional risk factor-based assessment. This approach can personalize the risk assessment by integrating the cumulative effect of risk determinants with lifetime exposure to risk factors in each individual patient. According to the ESC Guidelines for management of dyslipidemia (60), ultrasonography to assess carotid plaque burden should be considered (84,85) and coronary artery calcium (CAC) score assessment with computed tomography (CT) may be considered as risk modifiers in the CVD risk assessment of individuals at low or moderate risk (86). In individuals at high risk based on traditional risk scores, the additional value of imaging is low since these

patients should receive an intensive risk-reducing treatment anyway.Ultrasonography and

CT scanning are non-invasive cardiovascular imaging techniques which can show the presence and extent of subclinical atherosclerosis. However, the ultrasound technique does not expose individuals to radiation and the equipment used for ultrasound examinations is smaller and may be portable, compared to CT-equipment which is normally found in hospitals or specific research settings.

When examining medium sized arteries, ultrasound can detect increased carotid intima media thickness (cIMT) and presence of plaque, both early signs of atherosclerosis. Factors associated with an increase in cIMT and presence of plaque are also associated with progression of atherosclerosis (87). Pharmacological treatment and life style modification have the potential to reduce cIMT and plaque (88). A recent meta-analysis involving 100,667 patients concluded that the extent of intervention effects on cIMT progression predicted the degree of CVD risk reduction (89). The presence of plaque has shown a stronger association to CVD than an increase of cIMT (90-94). As a complement to

(25)

traditional risk factor-based assessment, the measurement of cIMT and carotid plaque (95) has the potential to improve the precision in risk stratification (91,96).

In early subclinical atherosclerosis, there is a subtle and heterogeneous thickening of the carotid wall. To be able to detect this focality, a multi-view screening which includes differences between left and right side and also within each side, is recommended (97-99). The Three-City study, France, showed as the number of sites with focal plaques in the carotids increased, so did the incidence of CVD events (100). Newly developed semi-automatic ultrasound equipment with standardized protocols may facilitate the reproducibility of examinations (19) which has previously been an obstacle when using ultrasonography. One can hypothesise, when using modern ultrasound techniques in combination with multi-view screening, that a more accurate evaluation of the burden of subclinical atherosclerosis can be made, this could subsequently lead to improved precision in risk assessment.

Visualising subclinical atherosclerosis, as a complement to traditional risk factor-based evaluation, has the potential to increase the accuracy in risk assessment and thereby motivate practitioners to improved risk management (91,95,96).

Risk management

The third step on the preventive stair is risk management, one step up from risk assessment. To correctly manage the CVD risk, the risk assessment needs to be adequate. When the assessed risk is high, the risk management needs to be more intense; the higher the risk, the more intensive treatment is needed (Figure 4). Compliance to guidelines for CVD prevention among practitioners is not optimal. This, in combination with poor adherence to CVD preventive actions among patients leads to poor risk factor control (101). One consequence of non-compliance to guidelines by practitioners is that patients do not get sufficient treatment.

There are many possible explanations for why patient management is not always in line with evidence-based recommendations. According to a study by Fürthauer et al, the most important cause appeared to be that the physicians do not provide indicated treatment (102). One explanation to this could be a lack of familiarity with preventive guidelines(103).

A recent study from Australia highlights this problem, among individuals categorised as high-risk at the time of admission presenting with acute coronary syndrome due to

coronary artery disease, 48% were on no preventative pharmacotherapy, 32% on

insufficient single pharmacotherapy and only 20% of patients on complete guideline-recommended pharmacotherapy (104). While a majority of GPs would recommend drug prescription where appropriate, this does not automatically translate into actual prescription. For example, a study from the UK found that only 42% of patients eligible for lipid-lowering medication were prescribed it (105). The GPs’ decisions to prescribe medication can also be influenced by their perception of how likely the patient is to start and continue with the medication. In addition, some GPs expressed an unwillingness to ‘medicalise’ unhealthy lifestyles and raise a false sense of security in patients through medication (83).

(26)

GPs also find it challenging to promote lifestyle changes, especially in patients with established lifestyle habits and particularly in older or obese patients, where there is a need of a more immediate solutions such as medication (83). More than 50% of GPs rated their ability to motivate behavioural change for CVD prevention as being ‘not good’, especially for male patients older than 65 and obese (106).

By showing a pictorial representation of the atherosclerosis itself, instead of the likelihood of CVD, has been shown to influence physicians´ prescription of evidence-based interventions (107). This measure can be one potential method to improve practitioners´ risk management of the patients.

Adherence to treatment

In healthcare, adherence is often described as the “active, voluntary and collaborative involvement of the patient in a mutually acceptable course of behaviour to produce a therapeutic result” (108). It is the choice of the patient to sustain the recommended course in order to obtain the desired outcomes whether it is exercise, diet, medications or follow-up. That choice could be influenced by several factors, including health literacy, costs and social support (109). The WHO has stated that, “increasing the effectiveness of adherence interventions may have a far greater impact on the health of the population than any new improvement in specific medical treatments”. Lack of adherence to preventive guidelines is a serious health problem.

Non-adherence and perception of risk

Patient information, patient involvement and the physician’s willingness to inform and involve the patient are of great importance to improve adherence to preventive recommendations (102). A cross-sectional survey of unintentional nonadherence in more than 24,000 adults with chronic illness, including hypertension, diabetes mellitus and hyperlipidaemia, showed that, 62% did not remember to take medications and 37% ran out

of their medications within the first year (110). Medications do not work if patients do not

take them. Non-adherence among patients can be divided into three different aspects: non-initiation, sub-optimal implementation of dosing regimens and discontinuation (111). The decision to start a treatment is dependent on several factors, key to which is the perception of the risk of the disease and the risk of the treatment, further indicating that the risk-benefit balance should be in focus during the consultation (62). A factor contributing to non-adherence to primary prevention actions to reduce CVD risk might be a weak association between perceived risk and actual risk. An individual’s perception of its risk is partially based on physical signs (112); in primary prevention the lack of physical signs can make correct perception of risk more difficult. Discontinuation and sub-optimal implementation correlated to the patient’s own beliefs, with the decision to follow the prescription being evaluated by the patient after considering potential benefits and harms of the treatment. Making changes to the dosage regimen or taking medication breaks can be strategies for patients to accept the treatment (113). An Australian study of 610 adults

who received newly prescribed lipid-lowering drugs, mainly from GPs, showed that 60%

were non-adherent to their medication after 12 months; the predominant reason was that patient were not convinced about their need for treatment (114). Furthermore, a possible

(27)

explanation for the high discontinuation rates for statins might be that they are often used for primary prevention and, therefore, patients may feel that there is no immediate threat to their health (115). Factors associated with non-adherence to statins are female sex, age <50 and >70 years, low income, low level of education in men, high level of education in women, being a new statin user and non-white race (116).

The adherence to lifestyle modification interventions regarding diet and physical activity is also low and seems to cease with time. Interventions concerning diet and physical activity, especially when they are combined, can reduce the risk factor burden. Intervention effects are most prominent after one year, while intervention effects in studies with a follow-up time ≥ 24 months was, in general, modest (117,118).

Active patient involvement

Patients who prefer a more active role in healthcare, such as engagement in treatment decisions and self-management, show a significantly better adherence when matched with care-givers who promote patient self-management in comparison to when matched with care-givers who do not promote patient self-management (110). Patients actively involved in their treatment show better adherence when given control over the treatment, in contrast to patients who favour a more passive role in treatment, who show poorer adherence when given control over treatment (119).

Risk communication

Many CVD risk factors, including hypertension and high cholesterol are given in numerical values even if individuals in general find it difficult to understand probabilities and numerical information (120). Furthermore, in individuals with low health literacy, it can be challenging to comprehend written as well as verbal communication about medical conditions and to understand their personal risk (121). An inappropriate optimism has been observed among high-risk patients, in particular men (122). This might not be known by healthcare providers, who often communicate the risk message verbally, despite other formats being available, it has been shown that visual tools are more effective (123). Pictorial information has the potential to overcome some of these difficulties (124). Deficient communication concerning the CVD risk and inappropriate perception of risk are barriers for patients´ adherence to risk control measures (125).

Pictorial information of subclinical atherosclerosis has been shown to influence patients´ motivation to exercise, implement dietary change (107), and adhere to medication and prevention-seeking behaviour (126). Pictorial presentation, showing the actual atherosclerotic disease instead of the likelihood of CVD has the potential to motivate individuals to increased adherence to preventive actions (127-129).

Health dialogues: -an integration of all steps of the preventive stair Health dialogues for prevention of CVD have been shown to reduce mortality on a population level (among both non-participants and participants) but with greater success among participants, as compared to the national level (130). A health dialogue, as compared to a health check, is characterized by being integrated in the regular primary

(28)

health care, being population-based, combining a low-risk and high-risk strategy for preventing disease and focusing on the patient dialogue where the patient’s life situation, preferences and CVD risk factors are considered (130-132). All steps of the preventive stair are integrated into the health dialogues. A population approach, where also those at low risk factor load despite unhealthy lifestyle habits receive the dialogue aiming at health promotion. Also, the non-participants may have a positive effect from the intervention through dissemination of the health message in social networks and from relatives. A risk assessment is performed, and the management is partly based on the participant’s risk. If individuals would benefit from medical treatment to reduce the risk the participant gets an appointment with a physician. The dialogue is focused on the patient’s life situation, preferences and CVD risk factors are considered in order to increase adherence to recommendations.

Since the 1990s health dialogues has been offered to the population in the county of Västerbotten, Sweden. Västerbotten has performed better than Sweden in general in reducing deaths from CVD. For men, age and time-standardized mortality in Västerbotten in the 1980s was 7.2% higher than the general Swedish male population, but since the mid-1990s it has been lower than the national average (133). Partly due to the results from Västerbotten, many other counties in Sweden have implemented or are considering implementing health dialogues (Figure 6- The situation 15/3/21)

Figure 6: Implementation status of health dialogues (HD) in Swedish counties. Green are counties who offers HD to inhabitants, light green-offers a HD but have at the moment stopped, yellow-pilot studies are completed and decision regarding full implementation is expected in March 2021, orange- political decision regarding implementation is taken, blue-is a part of or having discussions with HD team group and red-have not implemented HD or have discussions with HD-team group regarding implementation.

Is it possible to further improve CVD prevention by adding pictorial information regarding the individual’s burden of subclinical atherosclerosis to the already existing preventive

(29)

programme? With the novel small, portable and semi-automatic ultrasound machines this is a feasible strategy. This is what we aim to investigate in the VIPVIZA trial. The acronym stands for Västerbotten Intervention Programme – VIsualiZation of asymptomatic Atherosclerotic disease for optimum cardiovascular prevention, and VIP first in the acronym denotes the setting.

This thesis focuses on the VIPVIZA intervention effect on clinical risk factors, prevalence of subclinical atherosclerosis and experience of GPs receiving pictorial presentation of patients’ subclinical atherosclerosis. Within the VIPVIZA trial there are several other projects running, focusing on other aspects of the trial, such as the ultrasound results and the psycho-social determinants for behavioural change.

(30)

Aims

To improve CVD primary prevention in a middle-aged population through the VIPVIZA intervention, comprising pictorial information about subclinical atherosclerosis, a follow-up phone call, and additional written information, in order to obtain more accurate risk perception and to assist in better compliance to preventive measures through improved communication about the risk of CVD.

Specific aims:

1) To assess the prevalence of subclinical atherosclerotic disease in men and women through measurement of intima-media thickness (CIMT) and carotid plaque, and to relate CIMT and plaque to clinically estimated risk factors and traditional risk stratification.

2) To explore how the VIPVIZA intervention affects physicians in their perception and communication of CVD risk to patients.

3) To investigate the impact of the VIPVIZA intervention, in addition to conventional risk factor-based risk evaluation, on traditional risk factors and pharmacological treatment after 1 year and 3 years.

(31)

Materials and design

This thesis covers different aspects of CVD primary prevention. The four studies used data collected within the VIPVIZA trial. (Table 1)

Table 1: Overview of papers I-IV of this thesis.

Paper I II III IV

Running title CVD risk association with ultrasound measurements GPs’ experience from the VIPVIZA intervention Pictorial information reduces CVD risk over 1 year Pictorial information reduces CVD risk over 3 years

Study design Cross-sectional design Individual interviews Longitudinal design of PROBE (Prospective Randomized Open Labeled Blinded Evaluator) trial Longitudinal design of PROBE (Prospective Randomized Open Labeled Blinded Evaluator) trial Participants 3532 participants recruited from VIP, either aged 40 with one first degree relative with CVD, aged 50 with one CVD risk factor or aged 60 years old. 15 GPs with experience of at least 3 VIPVIZA results. 1599 participants from the intervention group and 1576 from the control group.

1580 participants from the intervention group and 1587 from the control group.

Intervention No intervention VIPVIZA-intervention

VIPVIZA-intervention

VIPVIZA intervention Data collection Baseline data

2013-2016

Semi-structured individual interviews. 2014-2016

Baseline and 1-year follow-up, 2014-2017

Baseline,

1-year follow-up and 3-year follow-up, 2016-2019 Analysis Prevalence Partial least squares (PLS) regression Step wise linear regression modelling

Qualitative content analysis

The control and intervention groups were compared for levels of CVD risk factors and risk scores. Difference-in-change analysis Intention-to-treat analyses. Multi-level regression analysis using panel data for comparison of the levels of CVD risk factors in the three different waves. Difference-in-change analysis.

Intention-to-treat analysis

(32)

Study setting and population

VIPVIZA is nested in the Västerbotten Intervention Programme (VIP), a CVD prevention programme, provided by the regular primary health care (PHC) in the county of Västerbotten, Sweden.

Since the 1990s, the PHC in Västerbotten invites all inhabitants the year they turn 40, 50 or 60 years to participate in VIP. The programme comprises CVD risk factor screening, a health dialogue with a trained nurse (VIP nurse) to promote a healthy lifestyle and, if indicated, referral for pharmacological CVD prevention. The average participation rate in VIP during the inclusion period to VIPVIZA was approximately 68%. In an evaluation of VIP from 1990 to 2006 only small social selection bias among VIP participants was detected (134). The most noticeable differences among participants and non-participants were found in the low-income group, compared to the moderate- and high-income groups, and among those of single marital status, compared to non-single marital status, where the participation rates were 10 % lower. The differences in education, age and rural or urban residency were less than 5% (134). Studies evaluating the effectiveness of VIP have shown a reduction in smoking (135), hypertension (132), and hypercholesterolemia (136), as well as a decline of the overall trend of increasing obesity (137). Further results from VIP suggest that the programme is also able to reduce CVD mortality (130). The VIP is a well-functioning prevention program from which research data has been collected since the 1990s. All data are collected according to standardized procedures and the VIP nurses are well informed about the importance of following set up routines (131). VIP is a well-suited setting in which to conduct VIPVIZA, a pragmatic randomised controlled trial. According to VIP data from 2012, approximately 60% of all participants of VIP were estimated to be at moderate risk of CVD due to having at least one conventional CVD risk factor.

In VIP, data regarding demographics, socio-economic status, clinical risk factors for CVD and lifestyle were collected. All participants receive an individual health dialogue with a VIP nurse. The VIP nurses also controlled for eligibility for participation in VIPVIZA and participants who were eligible were given written and verbal information about the VIPVIZA study and were invited to the study.

Inclusion criteria for VIPVIZA:

i) age=40 and a history of CVD at age <60 years among first-degree

relative(s) or

ii) age=50 years and at least one of the following: a history of CVD at

age <60 years among first-degree relative(s), smoking, diabetes, hypertension, serum LDL-cholesterol ≥4.5 mmol/L, abdominal obesity defined by waist circumference >88cm for women and >102 cm for men or

iii) age=60 years

The inclusion criteria were chosen to target individuals of intermediate risk of CVD by using established clinical criteria or levels of CVD risk markers which within VIP require a follow-up.

(33)

The VIP nurses who conducted the health dialogues ensured that the study information was provided, collected and the informed consent forms were signed and posted to the research nurses to register the participants. All participants were informed about the 1- and the 3-year follow-up visits in the first information about the trial provided before collecting the informed consent forms.

Based on power calculations, the study needed approximately 3500 participants; with a dropout rate of 20% during the study, this would be sufficient to ensure a probability of 80% to detect a significant difference in cIMT between the groups at a significance level of 5% over 3 years. The limiting factor was cIMT, requesting the largest group size to show a hypothesised effect. The difference in cIMT between the groups will be evaluated at the 3-year follow-up but is not a part of this thesis.

Initiation, pilot trial and preparation for launch of VIPVIZA

The first funding for planning of VIPVIZA was received in 2011. From spring 2012, forward information about the study was sent by mail to the management of the PHC, nurses and physicians. Information was also presented repeatedly at the regular education meetings in the county for VIP nurses and physicians. Before the launch and if requested, representatives from VIPVIZA visited health care centres to inform about VIPVIZA.

Before the full-scale trial was initiated, a pilot trial involving 95 participants was performed in September to October 2012. During the pilot study, the routines for the trial were tested and the pictorial information was further developed. The original ultrasound report from the CardioHealth station was tested and refined in collaboration with VIPVIZA pilot participants. The report was also tested on patients with different levels of education and lifestyle-related metabolic diseases at the Behavioural Medicine Clinic. The research nurses’ experiences from discussing the results with participants were also considered. Together the researchers combined the information and further evolved the written information with the aim of providing easy and accurate understanding of the ultrasound result, as well as the additional information regarding atherosclerosis as a dynamic process which can be modified by lifestyle changes and pharmacological treatment. A technician adjusted the printed scripts, so that the approved information was automatically generated based on results from the Ultrasound machine system, including both graphics and text.

Design

VIPVIZA is a prospective, open labelled, randomized controlled trial with blinded evaluators (PROBE) design study (138). The result of the intervention was evaluated at follow-up visits 1 and 3 years after the intervention was initiated. Participants in the control and intervention group were given the same routine care from VIP, but the intervention group additionally received the VIPVIZA intervention. The evaluators (the biomedical scientists performing the

(34)

ultrasound examination) were blinded as to which group the participants had been randomized to, and the researchers evaluating the trial were not involved in collecting the data. The endpoints were strictly defined before the trial stated (study protocol, appendix paper 3). Primary outcomes were levels of SCORE and FRS, while secondary outcomes were levels of individual risk factors. There were no medical serious adverse events to be expected by the intervention, however mental distress and anxiety could appear after receiving the VIPVIZA result. The research nurses called all participants two weeks after sending the pictorial information, and if anxiety and distress were detected the participants were also contacted by the researchers. If a serious stenosis (>50%) was found on the ultrasound examination the participants were excluded from the trial and referred to special care (n=22).

Figure 7. Schematic figure of the intervention and follow up visits and which part of the intervention each paper in this thesis covers.

VIPVIZA trial procedure

Baseline

Participants who were eligible to participate in the study and had signed informed consent were randomized 1:1 to the intervention and control groups prior to the first ultrasound examination. The randomization list was computer generated prior to the study and a consecutive allocation to the two groups was performed by the research nurses as the informed consent was received. A unique

(35)

identification code was assigned to each participant. The research nurses then sent invitation letters by mail with a time for appointment for the baseline ultrasound examination. To non-attenders, one reminder letter with a new appointment time was sent. The randomization was masked to the participant and the evaluators (ultrasonographers) by the time the ultrasound examination was performed.

Ultrasound examination

The inclusion and the baseline ultrasound examinations were performed between 29 April 2013 to 7 June 2016 by biomedical scientists specifically trained in carotid ultrasound techniques. The examinations were undertaken in hospitals in three towns in Västerbotten county (Umeå, Skellefteå and Lycksele), participant living in those towns or neighbouring towns were examined at the hospitals. Participants living in more remote inland areas were examined and at the local primary healthcare centres (Storuman, Vilhelmina, Sorsele, Norsjö and Malå). A portable automatic carotid ultrasound equipment was used (Panasonic Healthcare Corporation of North America, Newark, NK, USA). A standardized protocol was applied for the examination (139) of the common carotid artery (CCA), bifurcation, and internal and external carotid arteries. The time reserved for each ultrasound scan was 15 minutes. A carotid plaque was defined according to the Mannheim plaque consensus and the decision about plaque was made at the time of the examination (140). The automated method used measuring cIMT was compared to conventional manual measurements and reproducibility was comparable. The intraclass correlation coefficient of the intersonographer variability of cIMT measurement was 0,95 (141) and the Kappa value of plaque detection was 0,70 for the intersonographer variability (142). Examination of cIMT was performed in the distal 1-cm of the far wall, in a plaque free segment of CCA at two predefined angles at each side. The angles of insonation were automatically provided by the system. Measurements of cIMT (max and mean values) were automatically obtained. Based on the cIMT measurements from each angle from both sides, bilateral mean and mean of maximal value were calculated. Additionally, the maximal mean value of all four angles was selected and titled maximal mean value (Figure 8). Mean and maximal values measured at the left and right sides separately, were also used. The same ultrasound machine was used during the entire study period.

(36)

Figure 8. Illustration of the ultrasound carotid intima media thickness (cIMT) measurements. The left and right carotid arteries were insonated at two standardized angles (120 and 150 vs 210 and 240). In each projection (angle), the cIMT was measured automatically by the mean of mean cIMT and max of mean cIMT. Additionally, the maximal mean value of all four angles were selected and titled ‘maximal mean value’.

Intervention from baseline to the 1-year follow-up

The control group was followed up at the 1- and 3-year follow-up and did not receive any of the following information, which was supplied to the intervention group:

i) Approximately two weeks after the ultrasound examination, pictorial

information about the carotid ultrasound results was sent by post to each participant in the intervention group and their GPs. No information was sent to participants in the control group nor their physicians.

The information included:

The cIMT was presented as individual vascular age compared to normal vascular age with a graphical presentation of atherosclerosis in colour as a gauge, proceeding from a green sector, through yellow and orange to a red sector to illustrate vascular age. Green indicated vascular age at least 10 years younger, yellow; 5-10 years younger, orange; 5-10 years older and red; at least 10 years older. No data were available from VIPVIZA’s source population, therefore the Atherosclerosis Risk In Communities study population (ARIC) (143) was selected as the reference population for calculation of vascular age by sex and age due to

Pictorial presentation of subclinical atherosclerosis : a measure to reduce the risk for cardiovascular disease