Correcting for selective participation in cohort studies using auxiliary register data without identification of non-participants.

Aims: Selective participation may hamper the validity of population-based cohort studies. The resulting bias can be alleviated by linking auxiliary register data to both the participants and the non-participants of the study, estimating propensity scores for participation and correcting for participation based on these. However, registry holders may not be allowed to disclose sensitive data on (invited) non-participants. Our aim is to provide guidance on how adequate bias correction can be achieved by using auxiliary register data but without disclosing information that could be linked to the subset of non-participants. Methods: We show how existing methods can be used to estimate generalisation weights under various data disclosure scenarios where invited non-participants are indistinguishable from uninvited ones. We also demonstrate how the methods can be implemented using Nordic register data. Results: Inverse-probability-of-sampling weights estimated within a random sample of the target population in which the non-respondents are disclosed are equivalent in expectation to analogous weights in a scenario where the non-participants and uninvited individuals from the population are indistinguishable. To minimise the risk of disclosure when the entire population is invited to participate, investigators should instead consider inverse-odds-of-sampling weights, a method that has previously been suggested for transporting study results to external populations. Conclusions: Generalisation weights can be estimated from auxiliary register data without disclosing information on invited non-participants.

Predicting participation in the population-based Swedish cardiopulmonary bio-image study (SCAPIS) using register data.

AIMS: To illustrate the importance of access to register data on determinants and predictors of study participation to assess validity of population-based studies. In the present investigation, we use data on sociodemographic conditions and disease history among individuals invited to the Swedish cardiopulmonary bio-image study (SCAPIS) in order to establish a model that predicts study participation. METHODS: The pilot study of SCAPIS was conducted within the city of Gothenburg, Sweden, in 2012, with 2243 invited individuals (50% participation rate). An anonymous data set for the total target population ( n = 24,502) was made available by register authorities (Statistics Sweden and the National Board of Health and Welfare) and included indicators of invitation to and participation in SCAPIS along with register data on residential area, sociodemographic variables, and disease history. Propensity scores for participation were estimated using logistic regression. RESULTS: Residential area, country of birth, civil status, education, occupational status, and disposable income were all associated with participation in multivariable models. Adding data on disease history only increased overall classification ability marginally. The associations with disease history were diverse with some disease groups negatively associated with participation whereas some others tended to increase participation. CONCLUSIONS: The present investigation stresses the importance of a careful consideration of selection effects in population-based studies. Access to detailed register data also for non-participants can in the statistical analysis be used to control for selection bias and enhance generalizability, thereby making the results more relevant for policy decisions.

Very urgent carotid endarterectomy confers increased procedural risk.

BACKGROUND AND PURPOSE: Current Swedish guidelines recommend that carotid endarterectomy should be performed within 14 days of a qualifying neurological event, but it is not clear if very urgent surgery after an event is associated with increased perioperative risk. The aim of this study was to determine how the time between the event and carotid endarterectomy affects the procedural risk of mortality and stroke. METHODS: We prospectively analyzed data on all patients who underwent carotid endarterectomies for symptomatic carotid stenosis between May 12, 2008, and May 31, 2011, with records in the Swedish Vascular Registry (Swedvasc). Patients were divided according to time between the qualifying event and surgery (0-2 days, 3-7 days, 8-14 days, 15-180 days). Stroke rate and mortality at 30 days postsurgery were determined. RESULTS: We analyzed data for 2596 patients and found that the combined mortality and stroke rate for patients treated 0 to 2 days after qualifying event was 11.5% (17 of 148) versus 3.6% (29 of 804), 4.0% (27 of 677), and 5.4% (52 of 967) for the groups treated at 3 to 7 days, 8 to 14 days, and 15 to 180 days, respectively. In a multivariate analysis, time was an independent risk factor for perioperative complications: patients treated at 0 to 2 days had a relative OR of 4.24 (CI, 2.07-8.70; P<0.001) compared with the reference 3- to 7-day group. CONCLUSIONS: In this study of patients treated for symptomatic carotid disease, it was safe to perform surgery as early as Day 3 after a qualifying neurological event in contrast to patients treated within 0 to 2 days, which has a significantly increased perioperative risk.

Increased vascularization of shoulder regions of carotid atherosclerotic plaques from patients with diabetes.

OBJECTIVE: Increased vascularization is considered an important contributing factor for plaque vulnerability. Microvascular proliferative disease in patients with diabetes results in renal damage and visual loss. We assessed the hypothesis that vascularization in carotid atherosclerotic tissue is increased in diabetic patients, especially in the critical shoulder regions of the plaque. METHODS: Carotid endarterectomy specimens, clinical data, and blood samples were collected from patients with symptomatic carotid artery stenosis (median 85 days after clinical event) and pharmacologic treatment for diabetes (n = 26) or no diabetes (n = 85). Plaques were fixed in formalin and transverse tissue sections prepared. Histopathology and immunohistochemistry were performed for detection of endothelial cells (anti-CD34), macrophages (anti-CD68), vascular endothelial growth factor (VEGF), and its receptor (VEGFR-2). Neovascularization was assessed as CD34(+) neovessel density in the entire section area and by the presence or absence of CD34(+) vessels in the shoulder and cap regions of the plaques. RESULTS: The patient groups did not differ significantly in neovascularization in the entire transverse sections (2.0 vs 2.1 vessels/mm(2); P = .61) or in the fibrous cap (52% of the patients in both groups; P = .95). Neovascularization of the plaque shoulder regions was observed in 52% of the diabetic patients and in 26% of the nondiabetic patients (P = .028). VEGF-stained areas were similar in the two patient groups (0.4% and 0.2% of shoulder area; P = .61). Patients with diabetes had more VEGFR-2 (1.0% vs 0.2% of shoulder area; P < .016) and less CD68 staining (0.4% vs 3.6% of shoulder area; P < .008). Time from clinical event to surgery was positively associated with neovascularization of the plaque shoulder regions (≤90 days, 18% of patients; >90 days, 50% of patients; P = .002), independently of diabetes status. CONCLUSIONS: Diabetes was associated with increased vascularization of the shoulder regions in patients with symptomatic carotid atherosclerotic plaques. This was accompanied by increased expression of VEGFR-2. The increased vascularization of the plaque shoulder regions may help explain why patients with diabetes are at increased risk of atherosclerotic complications.

Erythrocyte sodium-lithium countertransport activity is inversely correlated to adiponectin, retinol binding protein 4 and body height.

BACKGROUND: We have previously described that the sodium/lithium countertransport (SLC) in the erythrocyte cell membrane is closely linked to obesity and insulin resistance. Adiponectin and retinol-binding protein 4 (RBP-4) are believed to affect obesity and insulin resistance. In the present study, we aimed to further characterize the relationship between SLC, inflammatory markers, adiponectin and RBP-4. METHODS: We included 93 clinically healthy 58-year-old men selected to display variations in insulin sensitivity. High sensitivity C-reactive protein (hs-CRP), TNF-alpha, soluble TNF-alpha-receptors (sTNFR) 1 and 2, IL-6 and RBP-4 were measured using antibody-based techniques. Adiponectin was determined by a radioimmunoassay kit. The lithium concentration in the special flux medium was measured by atomic absorption spectrophotometry. RESULTS: In univariate analyses, SLC correlated negatively with RBP-4 (r(s) = -0.256, p = -0.017) and to adiponectin (r(s) = -0.316, p = 0.003) and positively with TNF-alpha (r(s) = 0.346, p = 0.001) and hs-CRP (r(s) = 0.288, p = 0.005). There were no statistically significant correlations with sTNFR 1 or 2 or IL-6. SLC was negatively associated to body height (r(s) = -0.256, p = 0.013). CONCLUSIONS: We are the first to report that SLC correlates negatively with adiponectin and RBP-4. This finding is intriguing, as adiponectin is anti-inflammatory and anti-diabetic whereas RBP-4 supposedly decreases insulin sensitivity. We also observed a negative association between SLC activity and body height indicating that SLC activity is not primarily influenced by fat mass. The positive association of SLC with markers of inflammatory activity such as TNF-alpha and hs-CRP is in line with the proposed link between inflammation and insulin resistance.

Participation weighting based on sociodemographic register data improved external validity in a population-based cohort study.

OBJECTIVE: To investigate whether inverse probability of participation weighting (IPPW) using register data on sociodemographic and disease history variables can improve external validity in a cohort study with selective participation. STUDY DESIGN AND SETTING: We fitted various IPPW models by logistic regression using register data for the participants (n = 1,111) and nonparticipants (n = 1,132) of a Swedish cohort study. For each of six diagnostic groups, we then estimated (1) weighted disease prevalence proportions and (2) weighted cross-sectional associations (odds ratios) between sociodemographic variables and disease prevalence. Using register data on the remaining individuals of the entire study population of men and women aged 50-64 years (n = 22,259), we addressed how the choice of variables used for IPPW influenced estimation errors. RESULTS: Disease prevalence proportions were generally underestimated in the absence of IPPW but became markedly closer to population values after IPPW using sociodemographic variables. We found limited evidence of selective participation bias in association estimates, but IPPW improved external validity when bias was present. CONCLUSIONS: IPPW using sociodemographic register data can improve the external validity of disease prevalence estimates in cohort studies with selective participation. The performance of IPPW for association estimates merits further investigations in longitudinal settings and larger cohorts.

Inverse association between size of the lipid-rich necrotic core and vascularization in human carotid plaques.

PURPOSE: To study the relationship between the size of the lipid-rich necrotic core measured by MRI (magnetic resonance imaging) and the level of plaque vascularization measured by contrast-enhanced ultrasound, in human carotid plaques. Further, to compare the size of lipid-rich necrotic core from MRI to plaque echogenicity. METHODS: Thirty-one subjects with carotid plaques underwent standard B-mode ultrasound, contrast-enhanced ultrasound and MRI. The lipid-rich necrotic core was quantified using MRI. Contrast-enhanced ultrasound was used to measure carotid plaque vascularization. Standard B-mode ultrasound was used to measure plaque echogenicity as greyscale median. RESULTS: The amount of lipid-rich necrotic core correlated inversely with the degree of plaque vascularization (r = -0·40, P = 0·03). There were no correlations between the degree of plaque vascularization and the amount of fibrous tissue or calcifications. There were no correlations between greyscale median and the lipid-rich necrotic core, fibrous tissue or calcifications. CONCLUSIONS: We show that more dense plaque vascularization is associated with a lower plaque content of lipid-rich necrotic core. A large lipid-rich necrotic core and high plaque vascularization are both proposed as predictors of vulnerability, and our finding is therefore odds with some earlier observations. Our finding can be explained by the fact that the necrotic core of the plaque contains no viable tissue and therefore less of the plaque can be vascularized if the lipid-rich necrotic core is large. Our study suggests that the true relation between plaque vascularization and other indices of vulnerability is more complex than initially thought.

Comparison between magnetic resonance imaging and B-mode ultrasound in detecting and estimating the extent of human carotid atherosclerosis.

BACKGROUND: In MRI studies of carotid plaques, ultrasound is used to find plaques, which are later imaged using MRI. The performance in plaque detection has not been compared between the modalities. The aim of the current study was to compare the performance of MRI and ultrasound in detecting carotid artery plaques and measuring extent of atherosclerosis. METHODS: Subjects with at least one plaque (height≥2·5 mm) on ultrasound were imaged using MRI. The number of plaques and their height was measured in both modalities; plaque area and volume were analysed on ultrasound and MRI, respectively. RESULTS: Thirty-eight subjects were included. MRI detected plaques in 95% of carotid arteries with a plaque height of ≥2·5 mm on ultrasound and in all carotid arteries with a plaque exceeding 2·5 mm. MRI detected 53% of the plaques with a height below 2·5 mm. The plaque height measured with both techniques correlated significantly, 0·59, P<0·0001. Ultrasound-derived plaque height and plaque area correlated similarly to MRI-derived plaque volume, r = 0·52; P<0·0001 and r = 0·47; P = 0·001, respectively. CONCLUSIONS: We conclude that MRI has a similar sensitivity to ultrasound in finding carotid artery plaques that are 2·5 mm or higher. In smaller plaques, MRI detects fewer plaques. Multiple carotid plaques seen on ultrasound most often are a misinterpretation of the anatomy and correspond to a single plaque. Plaque height on ultrasound is comparable to plaque height on MRI and correlates fairly well with plaque volume on MRI making it an interesting proxy for plaque burden.

Increased Vascularization in the Vulnerable Upstream Regions of Both Early and Advanced Human Carotid Atherosclerosis.

BACKGROUND: Vascularization of atherosclerotic plaques has been linked to plaque vulnerability. The aim of this study was to test if the vascularization was increased in upstream regions of early atherosclerotic carotid plaques and also to test if the same pattern of vascularization was seen in complicated, symptomatic plaques. METHODS: We enrolled 45 subjects with early atherosclerotic lesions for contrast enhanced ultrasound and evaluated the percentage of plaque area in a longitudinal ultrasound section which contained contrast agent. Contrast-agent uptake was evaluated in both the upstream and downstream regions of the plaque. We also collected carotid endarterectomy specimens from 56 subjects and upstream and downstream regions were localized using magnetic resonance angiography and analyzed using histopathology and immunohistochemistry. RESULTS: Vascularization was increased in the upstream regions of early carotid plaques compared with downstream regions (30% vs. 23%, p = 0.033). Vascularization was also increased in the upstream regions of advanced atherosclerotic lesions compared with downstream regions (4.6 vs. 1.4 vessels/mm2, p = 0.001) and was associated with intra-plaque hemorrhage and inflammation. CONCLUSIONS: Vascularization is increased in the upstream regions of both early and advanced plaques and is in advanced lesions mainly driven by inflammation.

Metoprolol reduces proinflammatory cytokines and atherosclerosis in ApoE-/- mice.

A few studies in animals and humans suggest that metoprolol (ß1-selective adrenoceptor antagonist) may have a direct antiatherosclerotic effect. However, the mechanism behind this protective effect has not been established. The aim of the present study was to evaluate the effect of metoprolol on development of atherosclerosis in ApoE(-/-) mice and investigate its effect on the release of proinflammatory cytokines. Male ApoE(-/-) mice were treated with metoprolol (2.5 mg/kg/h) or saline for 11 weeks via osmotic minipumps. Atherosclerosis was assessed in thoracic aorta and aortic root. Total cholesterol levels and Th1/Th2 cytokines were analyzed in serum and macrophage content in lesions by immunohistochemistry. Metoprolol significantly reduced atherosclerotic plaque area in thoracic aorta (P < 0.05 versus Control). Further, metoprolol reduced serum TNFα and the chemokine CXCL1 (P < 0.01 versus Control for both) as well as decreasing the macrophage content in the plaques (P < 0.01 versus Control). Total cholesterol levels were not affected. In this study we found that a moderate dose of metoprolol significantly reduced atherosclerotic plaque area in thoracic aorta of ApoE(-/-) mice. Metoprolol also decreased serum levels of proinflammatory cytokines TNFα and CXCL1 and macrophage content in the plaques, showing that metoprolol has an anti-inflammatory effect.

A study of plaque vascularization and inflammation using quantitative contrast-enhanced US and PET/CT.

BACKGROUND: Contrast-enhanced ultrasound (CEUS) is an in vivo methodology to quantify carotid plaque vascularization. Increased metabolism in plaques, measured as FDG uptake in PET/CT examination, has been associated with markers of inflammation in histological samples. In this study, we tested the association between FDG uptake and vascularization measured by CEUS to assess whether CEUS can be used as an in vivo marker of plaque vulnerability. METHODS: After informed consent, subjects aged >60 years with carotid plaque height exceeding 2.5mm were recruited. CEUS was performed and analyzed using earlier described protocol and software, Contrast Quantification Program, which calculates the fraction of the plaque being contrast positive (CQP value). PET/CT examination was performed within 3 months of CEUS (median time 7 days). PET/CT images were acquired 90 min after FDG injection (2.7 MBq/kg). FDG uptake was measured as tissue background index (TBI), calculated using Spearman's rho as mean standard uptake value (SUV) of the plaque divided by mean SUV in the jugular vein (mean of 7 measuring points). Local ethics committee approved the study. RESULTS: We recruited 13 subjects (5 women) with a mean age of 71 years, 6 had a history of stroke or TIA, 1 had a history of ipsilateral stroke. CQP values showed a significant, positive correlation with TBI of carotid plaques, r=0.67, p<0.02. CONCLUSIONS: Plaque vascularization measured by CEUS correlates positively with FDG uptake measured by PET/CT in humans. This indicates an association between vascularization and inflammation and/or hypoxia, supporting the use of CEUS as a non-invasive method to detect plaque vulnerability.

The metabolic syndrome predicts cardiovascular events: results of a 13-year follow-up in initially healthy 58-year-old men.

BACKGROUND: The metabolic syndrome is a global health problem and is associated with subsequent development of cardiovascular disease (CVD). However, data are scarce concerning prospective association of the syndrome and CVD in populations free of diabetes and previous CVD that also is free of all cardiovascular drugs. The aim of this study was to assess the risk of cardiovascular events due to the metabolic syndrome in a population-based cohort of initially healthy, low-risk, and medication-free 58-year-old Swedish men during 13-years of follow-up. METHODS: From a total population sample of 1728 subjects, a stratified and randomly selected group of 391 subjects was included. The metabolic syndrome was defined according to the National Cholesterol Education Program. Cardiovascular events and cause of death were investigated by contact with the Centre of Epidemiology at the National Board of Health and Welfare. RESULTS: The metabolic syndrome increased the risk of cardiovascular events with a hazard ratio of 2.1 [95% confidence interval (CI) 1.3-3.4], P=0.003. When adjusted for the factors of leisure-time physical activity, smoking habits, alcohol intake, and low-density lipoprotein cholesterol (LDL-C), the hazard ratio was 2.0 (95% CI 1.1-3.4), P=0.016. CONCLUSION: In a 13-year follow-up in initially healthy men, the metabolic syndrome increases the risk of cardiovascular events by two-fold. This risk was maintained also after adjustment for lifestyle factors.

Social disruption stress increases IL-6 levels and accelerates atherosclerosis in ApoE-/- mice.

INTRODUCTION: We have previously shown that different forms of stress have distinctive effects on atherogenesis in mice. We showed that social stress increase atherosclerosis in ApoE(-/-) mice, while more physical forms of stress do not. Here we evaluated the effect of social disruption (SDR) stress on atherogenesis and evaluated cytokine release after SDR-stress and five more physical stressors. METHODS: Male ApoE(-/-) mice were exposed to SDR-stress during 12 weeks, and atherosclerotic plaque area was assessed in aorta, aortic root and innominate artery. Further, male C57BL/6 mice were exposed to SDR-stress or five physical stressors, and cytokine and corticosterone levels were analyzed in plasma/serum samples immediately after stress. RESULTS: We found a correlation between the level of SDR-stress and atherosclerotic plaque area in aorta and a numerical increased plaque area in aortic root. SDR stress did not affect histological features of plaque composition. However, SDR-stress increased levels of corticosterone, IL-6 and CXCL1. Plasma corticosterone increased for all five physical stressors, but IL-6 and CXCL1 only increased in the group exposed to restraint combined with rat odor. CONCLUSIONS: These findings suggest that SDR-stress is indeed atherogenic, in contrast to our previous results using the physical stressors. A possible explanation to this difference is that SDR-stress, but not physical stressors, leads to release of the pro-inflammatory cytokines IL-6 and CXCL1.

Slightly elevated high-sensitivity C-reactive protein (hsCRP) concentrations are associated with carotid atherosclerosis in women with varying degrees of glucose tolerance.

We examined whether high-sensitivity C-reactive protein (hsCRP) ≥2.0 mg/L was associated with increased intima-media thickness (IMT), plaque burden, and plaque echolucency in carotid arteries. Women (n = 635) from a population sample of 64-year-old females with varying degrees of glucose tolerance underwent risk factor assessment, measurement of hsCRP, and ultrasound examinations of the carotid arteries. Participants with hsCRP levels ≥2.0 mg/L had elevated carotid bulb IMT independently of other cardiovascular risk factors compared with those with hsCRP <2.0 mg/L. The participants with plaques in the highhsCRP group had larger total plaque area compared to those with plaque in the lower hsCRP group. Plaque echolucency did not differ between groups. High-sensitivity CRP levels ≥2.0 mg/L were accompanied by elevated IMT in the carotid bulbs independently of other cardiovascular risk factors. Total plaque area was larger among women with plaques in the high hsCRP group versus the lower hsCRP group.