Life expectancy associated with different ages at diagnosis of type 2 diabetes in high-income countries: 23 million person-years of observation

BACKGROUND

The prevalence of type 2 diabetes is increasing rapidly, particularly among younger age groups. Estimates suggest that people with diabetes die, on average, 6 years earlier than people without diabetes. We aimed to provide reliable estimates of the associations between age at diagnosis of diabetes and all-cause mortality, cause-specific mortality, and reductions in life expectancy.

METHODS

For this observational study, we conducted a combined analysis of individual-participant data from 19 high-income countries using two large-scale data sources: the Emerging Risk Factors Collaboration (96 cohorts, median baseline years 1961-2007, median latest follow-up years 1980-2013) and the UK Biobank (median baseline year 2006, median latest follow-up year 2020). We calculated age-adjusted and sex-adjusted hazard ratios (HRs) for all-cause mortality according to age at diagnosis of diabetes using data from 1 515 718 participants, in whom deaths were recorded during 23·1 million person-years of follow-up. We estimated cumulative survival by applying age-specific HRs to age-specific death rates from 2015 for the USA and the EU.

FINDINGS

For participants with diabetes, we observed a linear dose-response association between earlier age at diagnosis and higher risk of all-cause mortality compared with participants without diabetes. HRs were 2·69 (95% CI 2·43-2·97) when diagnosed at 30-39 years, 2·26 (2·08-2·45) at 40-49 years, 1·84 (1·72-1·97) at 50-59 years, 1·57 (1·47-1·67) at 60-69 years, and 1·39 (1·29-1·51) at 70 years and older. HRs per decade of earlier diagnosis were similar for men and women. Using death rates from the USA, a 50-year-old individual with diabetes died on average 14 years earlier when diagnosed aged 30 years, 10 years earlier when diagnosed aged 40 years, or 6 years earlier when diagnosed aged 50 years than an individual without diabetes. Using EU death rates, the corresponding estimates were 13, 9, or 5 years earlier.

INTERPRETATION

Every decade of earlier diagnosis of diabetes was associated with about 3-4 years of lower life expectancy, highlighting the need to develop and implement interventions that prevent or delay the onset of diabetes and to intensify the treatment of risk factors among young adults diagnosed with diabetes.

FUNDING

British Heart Foundation, Medical Research Council, National Institute for Health and Care Research, and Health Data Research UK.

Very long-term outcomes after acute myocardial infarction in young men and women

Aims

There is a paucity of data on very long-term outcomes in young women and men experiencing acute myocardial infarction (AMI).

Methods and results

The FAST-MI program consists of three nationwide French surveys carried out 5 years apart from 2005 to 2015, including consecutive AMI patients over a 1-month period with up to 10-year follow-up. The present analysis focused on adults ≤50 yo according to their gender. Women accounted for 17.5% (N=335) of the 1912 patients under 50 yo and were as old as men (43.9±5.5 vs. 43.9±5.1yo, p=0.92). Non-significant coronary artery disease was more frequent in women (12.8% vs. 5.8%, P<0.001). Women with significant coronary artery disease underwent less percutaneous coronary internvention (PCI) than men (85.9% vs. 91.3%, p=0.005), though primary PCI in ST-elevated myocardial infarction was as frequent in women (67.3% vs. 66.8%). Recommended secondary prevention medications were less prescribed at discharge in women (40.6% vs. 52.8%, p<0.001), a trend that persisted in 2015 (59.1% vs. 72.8% in 2015, p<0.001). Still, ten-year survival was similar in men (90.5%) and women (92.6%) (crude HR 0.87 [95% CI 0.55–1.57], adjusted HR 0.77 [95% CI 0.48–1.23], p=0.27) even among hospital survivors (adjusted HR 0.64 [95% CI 0.32–1.30], p=0.22). Ten-year survival was similar in patients with myocardial infarction with non-obstructive coronary arteries compared with those with significant coronary artery disease (91.8% vs. 91.0%, P=0.46)), both among men (91.6% vs. 90.7%) and women (92.9% vs. 93.0%). However, when taking into account early revascularization and discharge medications, in patients with significant coronary artery disease, adjusted 10-year mortality was significantly lower in women (HR 0.53, 95% CI 0.29–0.96, P=0.04).

Conclusions

Ten-year survival in young women with AMI is similar to that of men. However, in those with significant coronary artery disease, improving secondary prevention in women should result in better long-term outcome.

Funding Acknowledgement

Type of funding sources: Public Institution(s). Main funding source(s): French Society of Cardiology

Sex differences in acute coronary syndrome management and in 12-month case-fatality trends: data from the French MONICA registries

Funding Acknowledgements

Type of funding sources: Public hospital(s). Main funding source(s): Univ. Lille, Inserm, CHU Lille, Institut Pasteur de Lille, Santé publique France, U1167 - RID-AGE - Facteurs de risque et déterminants moléculaires des maladies liées au vieillissement, Lille, France.

Background

Earlier studies, have reported sex differences in clinical presentation, management and outcomes of acute coronary syndrome (ACS) which have prompted the medical community to take actions to erase these differences. To our knowledge, there has been no recent analysis of sex difference trends in ACS management, to assess whether these differences have been attenuated over time.

Aim

To assess recent sex differences trends in ACS characteristics, management and associated mortality.

Methods

We assessed all men and women (aged 35-74) hospitalized for an incident (first) ACS, in the 3 distinct geographical areas covered by the MONICA registries in the north, east and south-west of France, during a 12-month period in 2006 and a 6-month period in 2016. We analyzed the patients’ clinical, biochemical, electrocardiographic and care-related data, and their vital status 12 months after the ACS.

Results

We analyzed 2023 incident ACSs in 2006 and 1173 in 2016. The proportion of men was three times higher than that of women in both periods. In 2016, women were younger (62.0 y in 2006 and 60.4 y in 2016; p=0.06) and men were older (57.6 y in 2006 and 59.0 y in 2016; p<0.01). Women had no longer more atypical symptoms than men in 2016. In both men and women, the proportion of patients with NSTEMI increased from 26% in 2006 to 39% in 2016 (p<0.0001), whereas the proportion of patients with unstable angina decreased from 14% in 2006 to 7% in 2016 (p<0.0001). Between 2006 and 2016, the proportion of thrombolysis fell from 10% to 1% (p<0.0001); conversely, the proportion of patients receiving angioplasty increased from 67% to 75% (p<0.001). However, men were still more likely than women to receive revascularization therapy (+19% in 2006 and +18% in 2016). Between 2006 and 2016, prescriptions of angiotensin-converting enzyme inhibitors at discharge decreased from 68% to 64% (p< 0.01), and prescriptions of statins increased from 89% to 91% (p=0.02), as did prescription of functional rehabilitation from 34% to 40% (p<0.0001). Despite these trends, platelet aggregation inhibitors, statins and functional rehabilitation were still less prescribed in women than men in 2016 (p<0.01), independently of confounders. Finally, the 12-month case fatality rate was 11% in 2006 and 10% in 2016 (p=0.15), without sex differences.

Conclusions

The results of the present study evidenced an improvement over time in the management of ACS. However, although there were no longer sex differences in the patients’ age and clinical presentation, women with ACS were still less likely than men to receive revascularization and pharmacological treatments in 2016.

Long-term follow-up of incident acute coronary syndrome: results from the French MONICA registries over the 2009-2017 period

Funding Acknowledgements

Type of funding sources: Public hospital(s). Main funding source(s): Univ. Lille, Inserm, CHU Lille, Institut Pasteur de Lille, Santé Publique France, U1167 - RID-AGE - Facteurs de risque et déterminants moléculaires des maladies liées au vieillissement, Lille, France.

Background

The prognosis of an acute coronary syndrome (ACS) is strongly affected by the clinical, biological, and angiographic features of the event. However, few studies have characterized long-term recurrences of ACS after an incident (first) event.

Aim

The goals of this study were (i) to estimate the long-term (9 years) risk of ACS recurrence, including fatal ACS recurrence, among survivors of a first-ever ACS, according to its diagnosis subtype (STEMI / NSTEMI / Unstable angina (UA)) and (ii) to identify factors associated with these risks.

Methods

We assessed all men and women (aged 35-74) hospitalized between January 2009 and December 2016 for an incident (first-ever) ACS, in the 3 distinct geographical areas covered by the MONICA registries in the north, east and south-west of France, and still alive at discharge (Index event). ACSs were classified as STEMI, NSTEMI and UA. Patients were followed-up until December 2017. Recurrent events were defined as the first non-fatal or fatal ACS occurring after hospital discharge from the index event. Multivariate Cox regression models were used to assess the relationships between recurrent ACS and variables of interest.

Results

A total of 15,739 incident ACSs were included. The study comprised a total of 63,777 patients-years and a median duration of follow-up of 3.8 [1.6-6.0] years. There were 1,963 (12,4%) recurrent ACSs of which almost half (1,046; 53%) occurred during the first year. The 1-, 5- and 9- year cumulative probabilities of recurrent ACS were 6.7% [6.3-7.1%], 13.4% [12.8-14.0%] and 18.4% [17.4-19.5%], respectively, and those of fatal recurrent ACS were 1.4% [1.2-1.5%], 2.7% [2.3-3.0%] and 4.3% [3.6-4.9%], respectively. Annual 1-year recurrence rates decreased between 2009 and 2016, from 7.4% to 4.0% (p Cochran-Armitage test <0.001). After an index STEMI and NSTEMI, the most frequent form of recurrent event was a NSTEMI, whereas UA was more likely event after an index UA. The age at the time of the event, the geographical region (North to South gradient), the presence of a major event (i.e. resuscitated cardiac arrest, acute pulmonary oedema or cardiogenic shock), and an impaired left ventricular ejection fraction (LVEF) were significantly associated with the risk of recurrence and fatal recurrence. ACS subtype was not associated with recurrent risk after adjustment for confounders.

Conclusions

In conclusion, after an incident ACS the recurrence rate remained elevated, with one in five patients experiencing a recurrent ACS during a 9-years follow-up. Half of recurrent events occurred within the first year after the index event and NSTEMI was the most frequent form of recurrent event. Age at the time of the event, region, major event and impaired LVEF are factors associated with a higher risk of recurrence, the most important one being an LVEF <35%.

Breast cancer radiotherapy and risk of pacemaker implantation: an epidemiologic analysis using the French nationwide claims database

Background

Radiotherapy (RT) is a major treatment for breast cancer (BC), but it is also associated with an increased long-term (>5–10 years) risk of cardiac complications. Specific data on cardiac arrhythmias and conduction disorders are warranted.

Purpose

To evaluate whether pacemaker (PM) implantation risk is higher in BC patients treated with RT compared to the general population or population of BC patients treated without RT.

Methods

Our study was based on the Echantillon Généraliste de Bénéficiaires (EGB), the 1/97 random sample of the French nationwide health insurance claims database. The population of patients treated for BC consisted of adult women with a first BC identified by the ICD-10 diagnosis codes between 2008 and 2016, and followed through 2018. PM implantations and RT were identified with medical procedures codes in the french Common Classification of Medical Procedures (CCAM). For each year from 2008 to 2018, the reference population included all adult women of the EGB database. History of PM or BC was an exclusion criteria. Annual incidence rates of PM implantations were evaluated between 2008 and 2018 in the reference populations. The number of PM cases observed in BC patients at least one year after BC diagnosis was compared with the expected number of PM cases calculated by using reference incidence rates with age standardization. We then computed the standardized incidence ratio (SIR) as the ratio of the observed number of PM to the expected number of PM, with 95% confidence intervals (CIs). Thereafter, a survival analysis was performed in BC patients using a Cox regression model (Hazard Ratio – HR) in order to evaluate the risk of PM associated with treatment including RT (RT) compared with treatment not including RT (no RT).

Results

A total of 3,853 patients were included in the population of patients treated for BC between 2008 and 2016 with a mean age of 60 years old and a mean follow-up of 5.6 years. In this population, 35 cases of PM were observed, compared with 19.9 expected cases of PM, corresponding to a significant SIR=1.76 (95% CI: 1.22 - 2.39). This excess risk was carried by the RT group of 2973 patients (77% of BC population): 28 observed PM vs 12.9 expected, SIR=2.18 (95% CI: 1.45–3.06). In Cox regression analysis, after adjustment on age, RT was associated with an increased risk of PM compared with no RT, but this result did not reach statistical significance (HR=1.73; p-value=0.21).

Conclusion

Based on our study, BC patients receiving RT appeared to be at higher risk of PM implantation than general female population and BC patients without RT, illustrating a potential association between cardiac exposure during BC RT and conduction disorders.

Funding Acknowledgement

Type of funding sources: None.

Declines in in- and out-of-hospital coronary mortality from 2000 to 2016: results from the French MONICA registries

Background

In the past decades, improvements in the management of acute coronary events (ACE) have dramatically reduced in-hospital Coronary Heart Disease (CHD) mortality. Little is known on recent evolutions in out-of-hospital mortality and its contribution to overall ACE mortality.

Purpose

To estimate trends of in- and out-of-hospital ACE mortality rates from 2000 to 2016 and their respective contribution to total CHD mortality.

Methods

Using data from the three CHD MONICA registries, all fatal myocardial infarctions, coronary deaths and sudden deaths occurring between January 1st, 2000 and December 31st, 2016 in patients age 35 to 74 were recorded. Age-standardized mortality rates (SMRs) were estimated by gender and calendar year. Crude mortality rates (CMRs) were estimated by 20-years age group. Trends were expressed as annual percentage changes (APCs).

Results

During the study period, 20,822 fatal events were recorded, of which 69.4% took place out-of hospital (Figure 1). Almost 90% of out-of-hospital deaths occurred at home. Standardized ACE mortality rates decreased more steeply in women as compared to men, with an APC of −4.4%, versus −3.5% in men. Decreases in ACE mortality were more pronounced inside than outside the hospital (APC: −4.6% versus −3.1% in men; −5.9% versus −3.7% in women), with an increase in the contribution of out-of-hospital mortality to overall ACE mortality from 64.7% to 71.1% in men, and from 64.4% to 71.4% in women. Decreases in CMRs were more pronounced for people age 55–74 compared to those age 35–54. In the 55–74 age group, in-hospital SMRs decreased more steeply inside than outside the hospital (APC −5.2% versus −3.6% in men; −6.3% versus −3.9% in women, Figure 2). This was not observed among the youngest subjects, in whom the decreases for in- and out-of-hospital CMRs were similar (APC −1.5% versus −1.6% in men; −3.6% versus −3.1% in women).

Conclusions

Over the study period, out-of-hospital mortality accounted for more than two-thirds of overall ACE mortality. Declines were more pronounced for in-hospital mortality than for out-of hospital, except in the youngest subjects, in whom they were similar. These results highlight the need to strengthen primary prevention.

Funding Acknowledgement

Type of funding sources: Public Institution(s). Main funding source(s): - Santé Publique France (SPF) - Institut national de la santé et de la recherche médicale (Inserm)

How many CCS- and ACS-patients might reach the newly recommended LDL-C-target <55mg/dl in clinical practice if guidelines were applied – an estimate from the DYSIS II study population

Background

The 2019 ESC guidelines for the management of dyslipidemia even further lowered the LDL-C-target values for the very high-risk population from <70mg/dl to <55mg/dl. Population based studies already had shown that the previous target was difficult to reach. It is yet unclear how many patients in clinical practice might be treated to the new target.

Methods

The Dyslipidemia International Study (DYSIS II) prospectively collected data of patients with chronic coronary syndromes (CCS) and acute coronary syndromes (ACS) (all on statins) in 18 countries in Europe, the Middle East, South- and East Asia to document patient characteristics, medication and a current lipid profile from 2012 to 2014 under real life conditions in physicians' offices and hospitals. We took these real-life lipid profiles and data on the kind/dose of used statins to estimate how treatment escalation such as changing statin treatment to a high dose (atorvastatin ≥40mg / rosuvastatin≥20mg), adding ezetimibe and adding a PCSK9-inhibitor might help to bring LDL-C-levels to the recommended <55mg/dl target.

Results

A total of 7,865 patients were enrolled into DYSIS II, 6,794 had CCS and 1,071 ACS. Under the documented statin treatment in DYSIS only 12.7% of patients reached an LDL-C <55mg/dl. Putting all patients on high dose statins in combination with ezetimibe, 64.1% would reach the target. If PCSK9-inhibitors would be used in the remaining patients not at goal a total of 94.0% would match the goal.

Conclusion

Our analysis indicates that in real life practice the use available lipid-lowering medications would substantially increase the percentage of CCS- and ACS-patients reaching the newly recommended 2019 ESC guideline LDL-C-target of <55 mg/dl from less than 20% to more than 90% of the population.

Funding Acknowledgement

Type of funding source: Private grant(s) and/or Sponsorship. Main funding source(s): MSD

Participation in a research study related to acute myocardial infarction is not a guarantee to live more longer: results from the FAST-MI registries

Background

The main potential benefits of participating in a clinical trial is to have access to a treatment that is not available yet and to have a regular and careful attention from physicians. Several data have suggested that inclusion in a research study was associated with better clinical outcome.

Aims

The aim of this study is to describe the prevalence of inclusions in a research study (i.e., device or medication), clinical characteristics, management and clinical outcome in patients admitted for acute myocardial infarction (AMI) according to participation in a research study (versus not) using data from the French registries of Acute ST-or non-ST-elevation Myocardial infarction (FAST-MI) 2010 and 2015.

Methods

We used data from 2 one-month French registries, conducted 5 years apart, including 9,414 AMI admitted to coronary or intensive care units. We analyzed baseline characteristics, management and one-year survival according to participation in a research study.

Results

From 2010 to 2015, the prevalence of patients included in a research study decreased from 6.8% to 3.6% (P<0.001). Inclusions were performed mainly in university hospitals (8%). Clinical characteristics according to participation in a research study were strongly different. Overall, patients included in a research study were younger (61.2±12.7 vs 65.7±14.1; P<0.001) with less previous medical history and co-morbidities. Clinical presentation was preferentially a ST-elevation myocardial infarction (STEMI: 70% vs 52%; P<0.001) in these patients who had a lower GRACE score (133±32 vs. 141±35; P<0.001). The use of invasive strategies was more used in patients included in a research study (coronary angiogram: 99% vs 95%, P<0.001) as prescriptions of recommended medications (i.e., antiplatelet agents, beta-blockers, angiotensin-converting-enzyme inhibitor (ACE-I) or angiotensin receptor blocker (ARB) and statins) at discharge (72% vs 63%; P<0.001).

In a cox multivariate analysis, participation in a research study was not associated with lower mortality at one-year (HR= 0.68, 95% CI, 0.39–1.18, P=0.17). Similar results were observed in patients discharge alive (HR= 0.81, 95% CI, 0.44–1.48, P=0.49). Recommended medications were however more used in patients included in a clinical trial (OR=1.34; 95% CI, 1.09–1.65; P=0.007).

Conclusions

The number of inclusions in a research study related to AMI in France is low. Our data suggest that patients included in a research study are selected and received more recommended medications and invasive strategies. However, this management is not associated with a lower mortality at one-year.

Funding Acknowledgement

Type of funding source: None

Association between lipid lowering regimen intensity at discharge and long-term mortality in optimally-treated patients with acute myocardial infraction. The FAST-MI programme

Background

Randomised trials evaluate the efficacy of individual medications, irrespective of overall patient management. We assessed the association between lipid-lowering therapy (LLT) intensity and long-term mortality in otherwise optimally-treated patients with acute myocardial infarction (AMI).

Methods

FAST-MI consists in one-month nationwide French surveys of patients admitted for a recent AMI, repeated every 5 years. We used the 2010 and 2015 data with 3-year follow-up. Background optimal therapy was defined as use of PCI, together with ESC guideline-recommended treatment with beta-blockers, ACEi/ARB, when indicated, and optimal antithrombotic medications including type of P2Y12-i; of 9,460 patients included, 4,042 were optimally-treated, with 478 (12%), 1120 (28%), and 2,444 (60%) respectively receiving conventional-dose statins (Gr 1), moderate-intensity statins (atorvastatin 40 mg or rosuvastatin 10 mg) (Gr2) or high-dose LLT (atorvastatin 80 mg, rosuvastatin ≥20 mg or statin-ezetimibe combination) (Gr3).

Results

Baseline characteristics markedly differed in the 3 groups (Table 1).

Three-year Kaplan-Meier survival was 88.5%, 93.5% and 96.3% respectively for gr 1, 2 and 3, with Cox-adjusted HR of 0.75 (0.51–1.10), P=0.137, and 0.59 (0.41–0.86), P=0.006 for gr 2 and 3 compared with Gr1 (Figure).

Conclusion

In otherwise optimally-treated AMI patients, lipid-lowering regimen intensity at discharge was inversely associated with 3-year mortality. These results confirm that high-intensity lipid lowering therapy at discharge is likely beneficial even in patients receiving otherwise optimal therapy.

Figure 1. 3-year survival

Funding Acknowledgement

Type of funding source: Private grant(s) and/or Sponsorship. Main funding source(s): MSD, AstraZeneca

Correlates and prognostic impact of acute heart failure at the acute stage of ST-elevation and non-ST-elevation myocardial infarction according to diabetic status: the FAST-MI registries

Background

Diabetes mellitus (DM) predisposes to cardiovascular diseases including acute myocardial infarction (AMI) and acute heart failure (AHF).

Purpose

Analysing the French Registries of Acute ST-elevation and non-ST-elevation Myocardial Infarction (FAST-MI) 2005 and 2010, we assessed correlates of AHF occurring at the acute stage of ST-elevation AMI (STEMI) and non-ST-elevation AMI (NSTEMI), as well as the prognostic impact of AHF on 5-year mortality according to diabetic status.

Methods

The FAST-MI 2005 and 2010 registries included 7,839 consecutive patients admitted for AMI (4,250 STEMI and 3,589 NSTEMI). Vital status at 5 years was available in >96% of the patients. Binary logistic regression analysis was used to determine independent correlates of AHF and Cox multivariate analysis was used to determine independent correlates of 5-year mortality. Long-term survival curves were estimated using the Kaplan Meier method and comparisons were made using log-rank tests.

Results

2,151 patients presented with DM (27,4%) and 629 patients (8,0%) were treated by insulin (DMi). DM patients were older (70.0 vs. 64.6 years; p<0.001), with more comorbidities and more severe coronary artery disease. AHF (pulmonary edema or cardiogenic shock) was the most frequent in-hospital complication (12.5%) and was twice as frequent in DM patients (20.2% vs. 9.6%; adjusted OR=1.66; 95% confidence interval: 1.43–1.94; P<0.001). AHF was more frequently observed in DM patients on insulin therapy compared with DM patients not receiving insulin (29.1% vs 16.6%; adjusted OR=1.53; 95% CI: 1.20–1.96; P=0.001). The significant difference in AHF between DM patients and patients without DM was found in both STEMI (18.8% vs 8.0%; P=0.001) and in NSTEMI (21.3% vs 11.9%; P=0.001) patients.

After multivariate analysis on confounders (risk factors, previous medical history, type of AMI, year of survey and medications used before the index AMI), compared with patients without DM nor AHF, those with AHF without DM and those with DM without AHF had a 50% increase in 5-year mortality (adjusted HR=1.50; 95% CI: 1.32–1.69; P<0.001 and adjusted HR=1.46; 95% CI: 1.23–1.74; P<0.001) while the risk of 5-year death was doubled in those with both DM and AHF (adjusted HR=1.97; 95% CI: 1.66–2.34; P<0.0001).

Conclusion

AHF is the most frequent complication of AMI and is twice as common in DM patients. It is associated with reduced 5-year survival in non-diabetic and DM patients, with the worst outcomes in patients with both conditions (AHF and DM). In AMI, new management strategies are needed to prevent AHF and improve survival in DM patients with AHF.

Figure 1

Funding Acknowledgement

Type of funding source: Public grant(s) – National budget only. Main funding source(s): The FAST-MI 2005 and 2010 registries are the propriety of the French Society of Cardiology and were funded by grants from the following companies: Amgen, AstraZeneca, Bayer, BMS, Daiichi-Sankyo, Eli-Lilly, GSK, MSD, Novartis, Pfizer, Sanofi, and Servier, and by a grant from the French National Health Insurance body (CNAM-TS).

Professional drivers and cardiovascular risk: APPTIV cohort

Background

Professional drivers have a high risk of morbidity, including myocardial infarction. Bus, taxi and truck drivers have most often constituted the studied populations. However, if the transport sector faces road risk, 70% of employees in all activity sectors are exposed to this risk in France. Most often studied by the analysis of global working time, the real driving time and the distance travelled are rarely taken into account, whereas they are potentially modifiable factors.

Purpose

To assess the associations between exposure to road risk by taking into account driving time and the number of kilometers travelled professionally and certain cardiovascular risk factors, in different sectors of professional activity.

Method

Based on the APPTIV cohort, data from 984 men and women employees from 94 companies in France was collected from 2018 to 2019, during occupational health medical visits focused on road risk. The number of Kms / month (<400 (ref); <1600; ≥1600) and the duration of driving (hours / d <1 (ref); <4; ≥4) were studied according to the respective tertiles. A variable with 9 classes was created. Logistic regressions adjusted for age, gender and activity sectors were carried out to explain the associations with certain cardiovascular factors (Hypertension, overweight (≥25 kg/m2), obesity (30 kg/m2) Hypertriglyceridemia (≥1.7 mmol/l), Hypercholesterolemia, HypoHDLemia).

Results

The population studied came from 23.30% of the transport sector, 23.30% of the construction sector, 26.75 of the public administration and 4.78% of the trade sector. After adjustments, the estimated risks of hypertension according to the driving time or the number of km travelled are not significant. Driving between 1 and 4 hours and more than 4 hours per day increases the risk of obesity, by: OR: 1.85 (1.22–2.80); 1.85 (1.19–2.86), the risk of hypertriglyceridemia of 1.39 (1.0–1.95); 1.54 (1.09–2.18). Traveling more than 1600 km / m increases the risk of hypertriglyceridemia by 1.5. People driving between 400 and 1600 km / m with a driving time between 1 and 4 h / d or those covering more than 1600 km with long driving times (>4 h / d) have an estimated risk of obesity at 1.82 (1.05–3.14) and 1.87 (1.13–3.11) respectively. Those traveling over 1600 km / m with long driving times (>4 hrs / day) have a 1.5 times risk of hypertriglyceridemia.

Conclusion

Taking into account the number of hours of driving and the kilometers travelled professionally, this study highlighted some employee profiles with high cardiovascular risk, accessible to medical and professional prevention measures.

Funding Acknowledgement

Type of funding source: None

Compared prognostic impact of incident atrial fibrillation versus history of atrial fibrillation in patients with AMI: the FAST-MI programme

Background

History of atrial fibrillation (HxAF) and new onset atrial fibrillation (NOAF) at acute stage of MI are associated with poorer survival. Whether both entities carry an increased risk of stroke is uncertain.

Using data from the FAST-MI 2010 and 2015 registries, we analysed the associations between HxAF and NOAF and risk of 3-year death, nonfatal stroke or combined death or stroke.

Methods

The FAST-MI registries are nationwide French cohorts consecutively including AMI patients admitted over a 1-month period every 5 years. Baseline characteristics, acute management and medications at discharge are collected. Among 9460 patients with STEMI or NSTEMI, 610 (6.4%) had HxAF, and 626 (6.6%) developed NOAF.

Main characteristics

Table 1 Overall, NOAF was associated with larger and more severe AMIs.

Results

In hospital survivors, 3-year death was 8.6% in patients without AF, 23.2% in those with NOAF and 29.2% in those with HxAF. 3-year Kaplan-Meier rates of non-fatal stroke were 1.1%, 0.3% and 3.6%, respectively (Figure).

Compared with no AF, NOAF was not associated with non-fatal stroke (Cox HR, 95% CI: 0.17, 0.02–1.21), while HxAF was (HR, 95% CI 2.04, 1.13–3.66, P=0.017). Risk of death or stroke was increased for both NOAF (HR, 95% CI 1.35, 1.10–1.65, P=0.004) and HxAF (HR 95% CI, 1.37, 1.14–1.65, P=0.001). Risk of all-cause death at 3 years was increased for NOAF (HR, 95% CI 1.32, 1.09–1.60) and HxAF (HR, 95% CI 1.30, 1.09–1.55). The results were concordant in patients not receiving oral anticoagulants at discharge.

Conclusion

Both NOAF and HxAF are associated with increased risk of death at 3 years after AMI. NOAF, however, is not associated with an increased risk of non-fatal stroke.

Figure 1. Non-fatal stroke

Funding Acknowledgement

Type of funding source: Other. Main funding source(s): Pharma companies

Simple cardiovascular risk stratification using anthropometric measures instead of serum cholesterol. The MORGAM Prospective Cohort Project

Background

Body composition predicts cardiovascular outcomes, but it is uncertain whether anthropometric measures can replace the more expensive serum total cholesterol for cardiovascular risk stratification in low resource settings.

Purpose

The purpose of the study was to compare the additive prognostic ability of serum total cholesterol with that of body mass index (BMI), waist/hip ratio (WHR), and estimated fat mass (EFM, calculated using a validated prediction equation), individually and combined.

Methods

We used data from the MORGAM (MONICA, Risk, Genetics, Archiving, and Monograph) Prospective Cohort Project, an international pooling of cardiovascular cohorts, to determine the relationship between anthropometric measures, serum cholesterol, and cardiovascular events, using multivariable Cox proportional-hazards regression analysis. We further investigated the ability of these measures to enhance prognostication beyond a simpler prediction model, consisting of age, sex, smoking status, systolic blood pressures, and country, using comparison of area under the receiver operating characteristics curve (AUCROC) derived from binary logistic regression models. The primary endpoint was major adverse cardiovascular events (MACE), defined as a composite of death from coronary heart disease, myocardial infarction, or stroke.

Results

The study population consisted of 52,188 apparently healthy subjects (56.3% men) aged 47±12 years ranging from 20 to 84, derived from 37 European cohorts, with baseline between 1982–2002 all followed for 10 years during which MACE occurred in 2465 (4.7%) subjects. All anthropometric measures (BMI: hazard ratio (HR) 1.04 [95% confidence interval (CI): 1.03–1.05] per kg/m2; WHR: HR 7.5 [4.0–14.0] per unit; EFM: HR 1.02 [1.01–1.02] per kg) as well as serum total cholesterol (HR 1.20 [1.16–1.24] per mmol/l) were significantly associated with MACE (P<0.001 for all), independently of age, sex, smoking status, systolic blood pressures, and country. The addition of serum cholesterol significantly improved the predictive ability of the simple model (AUCROC 0.818 vs. 0.814, P<0.001), as did the combination of WHR, BMI, and EFM (AUCROC 0.817 vs. 0.814, P=0.004). When assessed individually, BMI (AUCROC 0.816 vs. 0.814, P=0.004) and WHR (AUCROC 0.815 vs. 0.814, P=0.02) improved model performance, while EFM narrowly missed significance (AUCROC 0.815 vs. 0.814, P=0.06). There was no significant difference in the predictive ability of a model including serum cholesterol versus that including all three anthropometric measures (AUCROC 0.818 vs. 0.817, P=0.13). The figure shows the pertinent areas under the ROC curve in predicting MACE.

Conclusion

In this large population-based cohort study, the addition of a combination of anthropometric measures, i.e. BMI, WHR, and EFM, raised the predictive ability of a simple prognostic model comparable to that obtained by the addition of serum total cholesterol.

Figure 1

Funding Acknowledgement

Type of funding source: None

Cardio-vascular risk factors according to different profiles of night work and shift work

Background

The shift-night work concerns around 17% workers in our modern society with evident need of this type of organisation for productivity and economic reasons for the companies. if the impact of this global organisation pattern on cardiovascular disease or risk factors is more and more documented, it remains unclear which types of rotating or night work are really involved and which are the impact on cardiovascular risk of being exposed currently or in the past to these different types of shift-night work.

Purpose

To study possible relationships between arterial blood pressure, anthropometric markers, fasting biomarkers, and sleep disturbances, and different exposure profiles of night work and/or shift work.

Methods

A cross-sectional study was conducted among 65.126 workers aged 18–70 at baseline in the French random population-based Constances cohort. Four exposure groups were defined: (1) workers with current permanent night work, (2) current night shift workers 3x8, (3) day workers with past night and/or shift work, (4) day workers without night and/or shift work during their work history (reference group). Logistic regression models adjusted for age, sex, socio-economic position, and personal history of ischemic cardiovascular disease (CVD) were performed to assess associations between clinical risk markers for CVD at baseline and these exposure groups.

Results

Compared with the reference group, current and past night workers had a significant higher risk of obesity (Body Mass Index≥30). Permanent night workers had a significant higher risk of metabolic syndrome, mostly due to abdominal obesity, hypertriglyceridemia (>1.7 mmol/l) and low level of high density lipoprotein. Even if day workers with past night/shift work had no statistically significant over-risk of metabolic syndrome, they presented significantly higher risks for the same three criteria as observed for the permanent night workers. However, only hypertriglyceridemia was significantly more frequent among 3x8 workers. Among current night workers, 3x8 had a higher risk of sleep disturbances while permanent had not. Sleep disturbances over-risk was also observed among day workers with past night/shift work. No over-risk of hypertension, hypercholesterolemia, hyperglycemia, and elevated levels of gamma-glutamyl-transferase was observed among the three groups.

Conclusion

The results support the need for enhanced cardiovascular follow-up of workers exposed to night and/or shift work. However, it is also necessary to follow regularly and specifically current day workers with a past night and/or shift work history.

Funding Acknowledgement

Type of funding source: Public Institution(s). Main funding source(s): INRS: institut National de Recherche et de Sécurité

Compared presentation, management and long-term outcomes after acute myocardial infraction in men and women <50 years of age. The FAST-MI programme

 

The number of young women suffering acute myocardial infarction (AMI) is increasing. There are conflicting results on the prognostic impact of gender in young patients with AMI. The aim of the present study was to assess in-hospital and 5-year outcomes in men and women ≤50 years old admitted for AMI and participating in the FAST-MI programme.

Methods

FAST-MI consists in one-month nationwide French surveys of patients admitted for a recent AMI, repeated every 5 years since 2005. We used the 2005, 2010 and 2015 data with up to 5-year follow-up, to describe baseline characteristics and outcomes according to gender in young patients. Of 13,130 patients included, 1,912 were ≤50 years old (335 women, 17.5%).

Results

Men and women had a similar age (44±5 years), BMI, current smoking (72% vs 75%), diabetes, family history, prior history of CVD, GRACE score (104±22 vs 106±23), and LVEF (54±10%). LDL at admission was lower in women (127±46 vs 138±46 mg/dl). Women had a higher admission heart rate (81±19 vs 78±17 BPM) and lower systolic blood pressure (133±26 vs 136±25 mm Hg). Presenting AMI was STEMI in 63% in women vs 68% in men (P=0.08).

All in-hospital complications except reinfarction (1.8 vs 0.6%, p<0.03) were similar in women and men; in-hospital death was 1.5% vs 0.8%, P=0.20. Coronary angiography was performed slightly less often (97.6% vs 99.0%, P=0.047) and more women had non-significant coronary artery disease (CAD) (13% vs 6%, P<0.001); PCI was less often used (74% vs 85.5%, P<0.001). At discharge, ESC guidelines-recommended medical treatment was less often prescribed in women (41% vs 53%, P<0.001), even in patients with significant CAD (46% vs 55%, P=0.004).

Kaplan-Meier 5-year survival did not differ in women (94.7%) and men (95.2%), P=0.56 (Figure). The respective figures for hospital survivors were 96.1% and 96.0% (HR 1.00, 95% CI 0.52–1.91; HR adjusted on age, type of MI, previous history of CAD, presence of significant CAD, LVEF, Killip class and appropriate medications at discharge: 0.99, 95% CI 0.51–0.92). Similar results were found for the combined end-point of death, AMI or stroke.

Conclusion

There were more similarities than differences between women and men who sustained an AMI at age ≤50 years. Women had lower LDL-c, lower haemoglobin, higher heart rate and lower blood pressure on admission. Non-significant CAD was more frequently found in women, who were less often treated with PCI and recommended medications at discharge. Five-year outcomes, however, did not differ according to gender.

Five-year mortality

Funding Acknowledgement

Type of funding source: Private grant(s) and/or Sponsorship. Main funding source(s): Pfizer, MSD, AstraZeneca

Are the results of clinical trials relevant in the real world? The applicability of REDUCE-IT to the FAST-MI Registry

Background

Large clinical trials are often criticized for testing therapeutic strategies in selected populations. The REDUCE-IT data revealed robust atherosclerotic cardiovascular risk reduction with a strategy comprising of high dose omega-3 icosapent ethyl versus placebo in statin treated patients with triglyceride values (TG) between 150 and 499 mg/dl and controlled LDL-C (41–100 mg/ dl).

Purpose

In order to evaluate the applicability of REDUCE-IT in a French population, we applied the inclusion and exclusion criteria in the French Registry on Acute ST- elevation and non-ST-elevation Myocardial Infarction (FAST-MI) [2010 and 2015].

Methods

From the FAST-MI registry, we included patients over 45 years, who had detailed lipid values post-acute hospitalization [11.1 months (median) after hospitalization for myocardial infarction]. We thus compared the applicability of REDUCE-IT [in patients with TG between 150 (or 200) and 500 mg/dl and an LDL-C between 40 and 100 mg/dl while treated with statin therapy] in the FAST-MI registry focusing on the general characteristics, the risk factors, and the cardiovascular prognosis i.e. the rate of total mortality, myocardial infarction and stroke.

Results

12.5% of the patients (5.5% if TG between 200 and 500 mg/dl) met the eligibility criteria for REDUCE-IT. The differences between the REDUCE-IT like (n=472) and REDUCE-IT excluded (n=3317) samples were related to age (61 vs 65, NS), male gender (79.5% vs 73.8%, p<0.01), body mass index in kg/m2 (28 vs 26.2, p<0.001), current smoking (45% vs 31.4%, p<0.001), hypertension (56.5% vs 50.2%, p<0.01), and the percentage of diabetic patients (29.5% vs 15.6%, p<0.001). In the REDUCE-IT like sample, the mean values of total cholesterol, TG (median) and the HDL-C were respectively 159, 192, 43 and the LDL-C value was 72 mg/dl. The distribution of the statin regimens in the REDUCE-IT like sample was as follows: 65.3%, 32.4% and 2.3% had high, moderate and low intensity statin therapy. The distribution of TG categories in the REDUCE-IT like sample was as follows: 55.3% (150–199), 36.0% (200–299) and 8.7% (300–500 mg/dl). The subgroup with TGs greater than 200 mg/dl and HDL-C less than 35 mg/dl was 16.1% in the REDUCE-IT like group and 2.1% in the REDUCE-IT excluded group (p<0.001). The cardiovascular event rate (death, nonfatal MI, nonfatal stroke) was respectively 36.7 for the FAST-MI REDUCE-IT like group and 36.9 persons-years (CVD death, nonfatal MI, nonfatal stroke) for the REDUCE-IT trial.

Conclusion

The cardiovascular residual risk related to elevated TG in the applicable patient population in the FAST-MI registry was similar to the risk in REDUCE-IT. If the results of REDUCE-IT are applied to patients hospitalized for a myocardial infarction in France, on top of statins, 12.5% of these patients could benefit from a strategy of high dose omega-3 icosapent ethyl on top of contemporary medical therapy to improve their future cardiovascular health.

Funding Acknowledgement

Type of funding source: None

A cost-utility analysis of increasing percutaneous coronary intervention use in elderly patients with acute coronary syndromes in six European countries

AIMS

Percutaneous coronary intervention reduces mortality in acute coronary syndrome patients but the cost-utility of increasing its use in elderly acute coronary syndrome patients is unknown.

METHODS

We assessed the efficiency of increased percutaneous coronary intervention use compared to current practice in patients aged ≥75 years admitted for acute coronary syndrome in France, Germany, Greece, Italy, Portugal and Spain with a semi-Markov state transition model. In-hospital mortality reduction estimates by percutaneous coronary intervention use and costs were derived from the EUROpean Treatment & Reduction of Acute Coronary Syndromes cost analysis EU project (n = 28,600). Risk of recurrence and out-of-hospital all-cause mortality were obtained from the Information System for the Development of Research in Primary Care (SIDIAP) database from North-Eastern Spain (n = 55,564). In-hospital mortality was modelled using stratified propensity score analysis. The 8-year acute coronary syndrome recurrence risk and out-of-hospital mortality were estimated with a multistate survival model. The scenarios analysed were to increase percutaneous coronary intervention use among patients with the highest, moderate and lowest probability of receiving percutaneous coronary intervention based on the propensity score analysis.

RESULTS

France, Greece and Portugal showed similar total costs/1000 individuals (7.29-11.05 m €); while in Germany, Italy and Spain, costs were higher (13.53-22.57 m €). Incremental cost-utility ratios of providing percutaneous coronary intervention to all patients ranged from 2262.8 €/quality adjusted life year gained for German males to 6324.3 €/quality adjusted life year gained for Italian females. Increasing percutaneous coronary intervention use was cost-effective at a willingness-to-pay threshold of 10,000 €/quality adjusted life year gained for all scenarios in the six countries, in males and females.

CONCLUSION

Compared to current clinical practice, broadening percutaneous coronary intervention use in elderly acute coronary syndrome patients would be cost-effective across different healthcare systems in Europe, regardless of the selected strategy.

P651Prevalence of hypertriglyceridemia in statin treated high risk patients who might benefit from treatment with icosapent ethyl for secondary prevention in clinical practice - Results of DYSIS

Background

Patients with elevated triglyceride (TG) levels are at increased risk for ischemic events. In the recently published REDUCE-IT-Study among patients with elevated triglyceride levels who were receiving statin therapy, the risk of major ischemic events, including cardiovascular death, was significantly lower with 2 g of icosapent ethyl twice daily than with placebo. Little is known about how many patients in clinical practice might benefit from this additional treatment.

Methods

The cross sectional, observational Dyslipidemia International Study (DYSIS) examined lipid goal attainment among statin-treated very high cardiovascular risk patients (defined as per 2011 EAS/ESC guidelines, including patients suffering from coronary heart disease, diabetes, chronic kidney disease or peripheral atherosclerotic disease) in Canada, Europe, Middle East countries and China. Data were collected under real life conditions in physicians' offices and hospital outpatient wards between 2008–2012. We examined the prevalence of mixed dyslipidemia with TG >135 mg/dl in high statin treated high risk patients in clinical practice.

Results

Of a total of 44,593 patients on very high cardiovascular risk, all on chronic statin treatment, 21,312 (47.8%) had mixed dyslipidemia with TG values >135mg/dl. Patients with elevated TG also had higher levels of total and LDL-cholesterol. The overall use of additional fibrates on top of statins was low (4.0%), but more frequent in patients with TG >135 mg/dl.

CVD patients with TG>135 mg/dl Pts with TG >135 mg/dl Pts with TG ≤135 mg/dl p-value OR (95% CI) n=21,312 (47.8%) n=23,281 (52.2%) Age (years) 65.4±10.1, 68.1±10.2 <0.0001 Females 42.2% 38.6% <0.0001 1.16 (1.12–1.21) Sedentary lifestyle 43.0% 37.6% <0.0001 1.25 (1.20–1.30) Risk factors/CV-disease   Hypertension 79.0% 74.9% <0.0001 1.26 (1.21–1.32)   Diabetes mellitus 56.3% 45.3% <0.0001 1.55 (1.50–1.61)   Ischemic heart disease 50.6% 54.4% <0.0001 0.86 (0.83–0.89)   Cerebrovascular disease 16.1% 17.7% <0.0001 0.90 (0.85–0.94)   Peripheral artery disease 8.1% 6.9% <0.0001 1.18 (1.10–1.27) Lipid lowering therapy   Rosuvastatin 11.3% 11.8% 0.13 0.95 (0.90–1.01)   Atorvastatin 38.2% 40.9% <0.0001 0.89 (0.86–0.93)   Simvastatin 41.9% 38.9% <0.0001 1.13 (1.09–1.18)   Other statins 8.6% 8.4% 0.12 0.96 (0.88–1.05)   Fibrates 5.5% 1.5% <0.0001 3.86 (3.40–4.37) Lipid Profile   Total Cholesterol (mg/dl) 185.6 159.0 <0.0001   LDL-Cholesterol (mg/dl) 102.1 88.9 <0.0001   Triglyderides (mg/dl) 189.5 95.7 <0.0001

Conclusion

Almost half of consecutive patients with cardiovascular disease treated with statins for secondary prevention in clinical practice suffer from TG levels >135 mg/dl and might benefit from additional treatment with icosapent ethyl with further reduction in subsequent major ischemic events, including cardiovascular death.

Acknowledgement/Funding

MSD

468Patients eligible for proprotein convertase subtilisin/kexin type9 inhibitors (pcsk9i) after acute myocardial infarction. The ESC position applied to patients included in the FAST-MI 2015 registry

Background

PCSK9i on top of high intensity statins have shown clinical benefit in patients after Acute Myocardial Infarction (AMI) who are not at LDL-c target. The ESC Task Force has defined guidance for the prescription of PCSK9i. Among patients discharged after AMI, the rate of those eligible for PCSCK9i is poorly documented.

Methods

We used data from the nationwide French FAST-MI 2015 registry. PSCK9-eligible patients were defined as those discharged with high intensity statins with expected-LDL>140 mg/dL, or >100mg/dL if they had additional high risk features such as diabetes with renal dysfunction or hypertension, multivessel coronary disease, associated peripheral artery disease or recurrent MI. The expected LDL-c was estimated from admission LDL-c and changes in lipid-lowering treatment. The rate of eligible patients was estimated from actual treatment and optimized treatment (i.e. addition of ezetimibe).

Results

Among 5291 pts included, 4715 (89%) were discharged with statins, at high intensity in 3655 (71%). Expected LDL was 71mg/dL (IQ 56, 95). Among patients discharged with high intensity statins, 3146 (59%) had an expected LDL-c<100/mg (figure, in green). PCSK9-eligible patients were those with LDL-c>140mg/dL (n=178, 3.3%, in red) and, among those with LDL-c 100–140mg/dL (n=331, 6.2%, in yellow), patients who had additional risk features (n=227 (4%)). As a result, the population eligible for PSCK9i according to the ESC guidance would represent 7.6% (405 pts) of the population admitted with AMI. Expanding the indication to patients with statins, but not at high intensity would add 159 (3%). Conversely, optimizing discharge treatment with ezetimibe would reduce the rate of eligible patients to 3% (181 pts, in brown).

Conclusions

In real life, according to the ESC Task Force, 7.6% of the whole population admitted for AMI would be eligible for PCK9i. This rate could be reduced to 3% with the addition of ezetimibe.

P4569Differential prognostic impact of blood glucose levels at the acute stage of myocardial infarction according to HbA1c. The FAST-MI programme

Background

Hyperglycemia is a well-known prognostic marker in patients with acute myocardial infarction (AMI), associated with higher mortality compared with normoglycemia. Whether the prognostic impact of glycemic status at the acute stage of AMI is similar in patients with chronic dysglycemia has not been extensively explored.

Aims and methods

Using data from the nationwide French FAST-MI cohorts (2005, 2010 and 2015), we analysed the association between glycemia at entry and 30-day death, according to HbA1c level. From the 13,130 patients included, 5,452 had both glycemia and HbA1c assessed at entry. Of those, 1173 (21.5%) had an HbA1c ≥7%.

Results

In patients with HbA1c <7%, LVEF was inversely correlated with glycemic levels (55±11% for glycemia <100, 52±11% for glycemia 100–140, 50±12% for glycemia 140–160 and 49±12% for glycemia >180 mg/dl); a graded association between admission glycemia and 30-day mortality was observed, ranging from 0.7% in normoglycemic patients to 6.3% in patients with admission glycemia >180 mg/dl. In contrast, in patients with HbA1c ≥7%, LVEF was not correlated with glycemia (<100 mg/dl: 49±14%, >180 mg/dl 49±12%), and mortality was the highest in patients with normoglycemia (9.2%) and the lowest in patients with glycemia between 140 and 180 mg/dl (3.1%) (Figure). In multivariate analyses adjusting for baseline characteristics and early management, normoglycemia was associated with a decreased risk of 30-day mortality in patients with HbA1c <7% (HR 0.27, 95% confidence interval 0.10–0.73, P=0.01), while it was associated with a two-fold increase in mortality in patients with HbA1c ≥7% (HR 2.49, 95% confidence interval 1.02–6.09, P=0.046).

Figure 1. 30-day death

Conclusion

In AMI patients with high HbA1c levels on admission, normoglycemia is associated with higher early mortality than hyperglycemia. In contrast, a graded correlation is observed between admission glycemia and early mortality in patients with HbA1c <7%. Management of glycemia at the acute stage of MI might require different measures according to the initial HbA1c level.

Acknowledgement/Funding

Amgen, AstraZeneca, Bayer, Daiichi-Sankyo, Eli-Lilly, GSK, MSD, Novartis, Pfizer, Sanofi, Servier

P2480Sub-optimal anti-diabetic treatment in patients with pre-existing and newly diagnosed diabetes admitted for acute myocardial infarction. A study from the FAST-MI 2015 registry

Background

In patients (pts) with Acute Myocardial Infarction (AMI), we evaluate how diabetes is detected and treated at discharge.

Methods

Using the French FAST-MI 2015 registry, pts were classed as non-diabetic (NonDiab), pre-existing (PreEx) or newly diagnosed (NewDiab) diabetes. PreEx was defined by history or pretreatment; NewDiab as no history, no anti-diabetic treatment plus HbA1C>6.5% or admission glucose ≥200mg/dL. Characteristics and adjusted 1-year mortality were compared.

Results

In 5291 FAST MI pts, 3857 (73%) were NonDiab, 1145 (21.5%) PreEx and 289 (5.5%) NewDiab (176 had glucose ≥200mg/dL; 143 had HbA1C >6.5%). PreEx pts were older, had more comorbidities, and higher GRACE score vs NonDiab. NewDiab pts had higher HbA1C (8.9%±9.2 vs 5.65%±0.38 in NonDiab; 7.5%±2.2 in PreEx). At 1 yr, diabetic pts (PreEx and/or NewDiab) had a 2.5-times higher adjusted risk of death. At discharge, DAPT, statins, ACEi and betablockers were less often prescribed in PreEx vs NonDiab or NewDiab pts. In diabetic pts, vs admission, the number of anti-diabetic treatments decreased at discharge in 11%, no change in 69% and increased in 20%. Despite higher HbA1C in NewDiab vs PreEx pts, anti-diabetics were less often prescribed in NewDiab (23%) vs PreEx pts (75%) (table). In pts with HbA1C>8%, treatment intensification was observed in 30%.

Admission treatment Discharge treatment NonDiab PreEx NewDiab NonDiab PreEx NewDiab 3759 (71%) 1283 (24%) 249 (5%) 3759 (71%) 1283 (24%) 249 (5%) Any Anti Diabetic Tx 0 975 (76%) 0 51 (1%) 931 (75%) 53 (23%) Insulin 0 339 (26%) 0 2 404 (33%) 13 (6%) Biguanide 0 456 (36%) 0 2 412 (34%) 30 (13%) Sulfonylureas 0 258 (20%) 0 2 244 (20%) 4 DPP4i 0 175 (14%) 0 2 176 (14%) 6 (3%) GLP1 RAs 0 108 (8%) 0 1 117 (10%) 10 (1%) DAPT 3212 (88%) 1009 (90%) 201 (87%) Aspirin 778 (21%) 550 (45%) 66 (27%) 3533 (96%) 1170 (95%) 218 (94%) Statins 964 (25%) 721 (56%) 69 (28%) 2394 (92%) 1109 (82%) 216 (93%) ACEi/ARB 1028 (27%) 672 (53%) 69 (28%) 2744 (74%) 925 (75%) 174 (87%) Coronary Angiography 3684 (98%) 1202 (94%) 240 (96%) PCI 3010 (80%) 907 (71%) 201 (81%)

Conclusions

In AMI pts, 5.5% have previously unknown diabetes and have a higher risk of death, similar to that of pts with PreEx diabetes. Treatment initiation and intensification are sub-optimal.

P5473Perceived well-being after acute myocardial infarction according to diabetic status and its impact long-term mortality. The FAST-MI programme

Background

Little information is available regarding subjective well-being according to diabetic status in patients surviving an acute myocardial infarction (AMI).

Aims and methods

A health status questionnaire was sent to all patients participating in the nationwide French FAST-MI cohorts (2005, 2010 and 2015) one year after the acute episode of AMI, with a specific question on overall health condition (how do you feel: very well, well, fair, poor) (n=6082), and additional questions on the presence of any chest pain (n=4590), presence nuisance bleedings (n=3968), and reporting of perceived medication side-effects (n=2220). Answers were analysed according to diabetic status. We also analysed in the 2005 and 2010 cohorts whether subjective health status at one year was a correlate of 5-year mortality.

Results

Compared with non-diabetic patients (n=4692), those with diabetes (n=1390) had a poorer perceived health status (poor/fair health status: 40% vs 30%, P<0.001), more presence of chest pain (31% vs 23%, P<0.001), but less nuisance bleedings (47% vs 56%, P<0.001) and similar reported medication side-effects (37% vs 41%, P=0.12). Using logistic regression analysis, diabetes was an independent correlate of poorer health status (any diabetes OR vs non-diabetic 1.29, 1.13–1.49, P<0.001; non-insulin-treated: 1.17, 1.00–1.36, P=0.05, insulin-treated: 1.58, 1.29–1.94, P<0.001). In diabetic patients, perceived health condition at one year was an independent correlate of 5-year death (Cox multivariate analysis): compared with patients reporting very good health, HR (95% CI) 2.16 (1.08–4.32) for good health, 3.06 (1.53–6.11) for fair health, and 3.63 (1.55–8.55) for poor health; in non-diabetic patients, 5-year survival was similar in those reporting good or very good health status, but lower in those reporting fair or poor health status (Figure).

Figure 1. 5-year survival by health status

Conclusion

Compared with non-diabetic patients, patients with diabetes described a poorer general health and more residual chest pain after AMI, but had less nuisance bleedings. Health condition reported one year after AMI was an independent correlate of subsequent long-term mortality, particularly for diabetic patients.

Acknowledgement/Funding

Amgen, AstraZeneca, Bayer, Daiichi-Sankyo, Eli-Lilly, GSK, MSD, Novartis, Pfizer, Sanofi, Servier

5197Correlates and prognostic significance of nuisance bleeding after acute myocardial infarction. The FAST-MI programme

Background

Following acute myocardial infarction (AMI), most patients receive potent antithrombotic medications, which may promote nuisance bleedings (ecchymoses, minor nose or dental bleeds etc.). Little information is available on the factors related to nuisance bleedings, nor on their prognostic significance in post-AMI patients.

Aims and methods

A health status questionnaire was sent to all patients participating in the nationwide French FAST-MI cohorts (2010 and 2015) one year after the acute episode, with a specific question on the presence of nuisance bleedings. Overall, 3968 patients answered the question on the presence nuisance bleedings. In the 2010 cohorts, we also analysed whether the presence of nuisance bleedings at one year was an independent correlate of 5-year mortality.

Results

54% of the patients reported the presence of nuisance bleedings (59% in 2010 and 51% in 2015). In univariate analyses, nuisance bleedings were more frequently found in younger patients, women, patients with STEMI, current smokers, patients treated with PCI, those receiving newer P2Y12 inhibitors or ACE-inhibitors but less frequent in patients with diabetes, hypertension, or those receiving ARBs or direct oral anticoagulants. Using logistic regression analysis, however, the only independent correlates of nuisance bleedings were: female gender (OR 1.45, 1.25–1.68), age ≤60 years (OR 1.22, 1.06–1.41), VKAs (OR 1.72, 1.28–2.31), clopidogrel (OR 1.62, 1.29–2.03), prasugrel (OR 3.16, 2.43–4.09), ticagrelor (OR 2.61, 2.04–3.35) at discharge, diabetes (OR 0.74, 0.63–0.88) and year 2015 vs 2010 (OR 0.62, 0.53–0.73).

In the 2010 cohort, the presence of nuisance bleeding at one year was not a predictor of mortality at 5 years (90% survival in both patients with or without nuisance bleedings; adjusted HR 0.96, 95% CI 0.69–1.33) (Figure).

Figure 1. 5-yr survival by nuisance bleed

Conclusion

Nuisance bleedings one year after AMI are extremely frequent. They are more common in women, younger patients, in patients receiving P2Y12 inhibitors, especially newer P2Y12-i, and in those receiving VKAs; in contrast, diabetic patients report less nuisance bleedings. The presence of nuisance bleeding at one year does not appear to impact 5-year mortality.

Acknowledgement/Funding

Amgen, AstraZeneca, Bayer, Daiichi-Sankyo, Eli-Lilly, GSK, MSD, Novartis, Pfizer, Sanofi, Servier

P648Screening and treatment of familial hypercholesterolemia in a French sample of ambulatory care: a retrospective longitudinal cohort study

Background

Familial hypercholesterolemia (FH) is largely underdiagnosed as there are typically no clinical symptoms prior to the first cardiovascular (CV) event. We conducted a study which utilised ambulatory care electronic medical record (EMRs) to alert physicians to possible cases of FH. Specifically, physicians were alerted to LDL-C levels >190 mg/dL (suggesting a risk of FH) and invited to complete the Dutch Lipid Clinic Network score (DLCN).

Purpose

Describe characteristics, comorbidities and clinical management of patients diagnosed with definite or probable FH in an ambulatory care setting.

Methods

All patients with a DLCN score of definite/probable FH (score higher or equal than 6; index event) between January 2016 and September 2018 were identified in the THIN® database (The Health Improvement Network; an anonymized EMR powered by GERSDATA, a Cegedim Health Data Division). These fully anonymized data were collected by 2000 General Practitioners (GP), 130 cardiologists and 40 endocrinologists, receiving 5.5 million patients regularly in their office. Sociodemographic, laboratory measurements, comorbidities, lipid-lowering therapies (LLT), visits to specialists, LDL-C and hospitalizations were collected and analysed at baseline, and 1, 2, 3, 6 months, and 1 year thereafter.

Results

From 999 anonymous patients with a DLCN score, 98 (10%) FH patients were identified (38 [39%] definite FH, 60 [61%] probable FH) while remaining fully anonymous, 9 (9%) of whom already had genetic testing. Mean (SD) age was 57.4 (14.3) years; 56 (57%) patients were female, half (51/98 [52%]) were diagnosed with pure hypercholesterolemia (ICD-10 code: E78.0) and 9 (9%) had a personal history of CV event. 93 patients (95%) had a LDL-C measurement prior to DLCN assessment (definite FH, 36/38 [95%]; probable FH, 57/60 [95%]). Among screened FH patients, 61.2% had LDL-C between 190 to 250 mg/dL and 16.3% had LDL-C higher than 250 mg/dL. At the time of DLCN assessment, one third (30/98 [31%]) of patients were not receiving any LLT, one third ([35%] 34/98) were receiving statins alone, 19% (19/98) receiving LLT combination with statin, and 15% (15) other LLTs. Moderate statin intensity was prescribed in 20% (20/98) of patients; high intensity statin, 17%, (17/98); low intensity, 10% (10/98). No improvement on LLT use (including use of high statin intensity) was observed over the 12-month follow-up.

Conclusion

This is the first study in France that use EMR to screen possible FH patients and support GPs in identifying patients that need to be treated. Our data highlight the need to screen, diagnosis and treat potential FH patients in ambulatory care settings. Longer follow-up is needed to evaluate the impact of FH assessment on referral to specialists, LLT and clinical outcomes.

Acknowledgement/Funding

Amgen

A Role for Behavior in the Relationships Between Depression and Hostility and Cardiovascular Disease Incidence, Mortality, and All-Cause Mortality: the Prime Study

Background

Behavioral factors are important in disease incidence and mortality and may explain associations between mortality and various psychological traits.

Purpose

These analyses investigated the impact of behavioral factors on the associations between depression, hostility and cardiovascular disease(CVD) incidence, CVD mortality, and all-cause mortality.

Methods

Data from the PRIME Study (N = 6953 men) were analyzed using Cox proportional hazards models, following adjustment for demographic and biological CVD risk factors, and other psychological traits, including social support.

Results

Following initial adjustment, both depression and hostility were significantly associated with both mortality outcomes (smallest SHR = 1.24, p < 0.001). Following adjustment for behavioral factors, all relationships were attenuated both when accounting for and not accounting for other psychological variables. Associations with all-cause mortality remained significant (smallest SHR = 1.14, p = 0.04). Of the behaviors included, the most significant contribution to outcomes was found for smoking, but a role was also found for fruit and vegetable intakes and high alcohol consumption.

Conclusions

These findings demonstrate well-known associations between depression, hostility, and mortality and suggest the potential importance of behaviors in explaining these relationships.