Serum calcium levels are associated with novel cardiometabolic risk factors in the population-based CoLaus study.

BACKGROUND: Associations of serum calcium levels with the metabolic syndrome and other novel cardio-metabolic risk factors not classically included in the metabolic syndrome, such as those involved in oxidative stress, are largely unexplored. We analyzed the association of albumin-corrected serum calcium levels with conventional and non-conventional cardio-metabolic risk factors in a general adult population. METHODOLOGY/PRINCIPAL FINDINGS: The CoLaus study is a population-based study including Caucasians from Lausanne, Switzerland. The metabolic syndrome was defined using the Adult Treatment Panel III criteria. Non-conventional cardio-metabolic risk factors considered included: fat mass, leptin, LDL particle size, apolipoprotein B, fasting insulin, adiponectin, ultrasensitive CRP, serum uric acid, homocysteine, and gamma-glutamyltransferase. We used adjusted standardized multivariable regression to compare the association of each cardio-metabolic risk factor with albumin-corrected serum calcium. We assessed associations of albumin-corrected serum calcium with the cumulative number of non-conventional cardio-metabolic risk factors. We analyzed 4,231 subjects aged 35 to 75 years. Corrected serum calcium increased with both the number of the metabolic syndrome components and the number of non-conventional cardio-metabolic risk factors, independently of the metabolic syndrome and BMI. Among conventional and non-conventional cardio-metabolic risk factors, the strongest positive associations were found for factors related to oxidative stress (uric acid, homocysteine and gamma-glutamyltransferase). Adiponectin had the strongest negative association with corrected serum calcium. CONCLUSIONS/SIGNIFICANCE: Serum calcium was associated with the metabolic syndrome and with non-conventional cardio-metabolic risk factors independently of the metabolic syndrome. Associations with uric acid, homocysteine and gamma-glutamyltransferase were the strongest. These novel findings suggest that serum calcium levels may be associated with cardiovascular risk via oxidative stress.

Small dense lipoproteins, apolipoprotein B, and risk of coronary events in HIV-infected patients on antiretroviral therapy: the Swiss HIV Cohort Study.

OBJECTIVES: HIV infection and exposure to certain antiretroviral drugs is associated with dyslipidemia and increased risk for coronary events. Whether this risk is mediated by highly atherogenic lipoproteins is unclear. We investigated the association of highly atherogenic small dense low-density lipoproteins (LDLs) and apolipoprotein B and coronary events in HIV-infected individuals receiving antiretroviral therapy. METHODS: We conducted a case-control study nested into the Swiss HIV Cohort Study to investigate the association of small dense LDL and apolipoprotein B and coronary events in 98 antiretroviral drug-treated patients with a first coronary event (19 fatal and 79 nonfatal coronary events with 53 definite and 15 possible myocardial infarctions, 11 angioplasties or bypasses) and 393 treated controls matched for age, gender, and smoking status. Lipids were measured by ultracentrifugation. RESULTS: In models including cholesterol, triglycerides, high-density lipoprotein cholesterol, blood pressure, central obesity, diabetes, and family history, there was an independent association between small dense LDL and coronary events [odds ratio (OR) for 1 mg/dL increase: 1.06, 95% confidence interval (CI): 1.00 to 1.11] and apolipoprotein B (OR for 10 mg/dL increase: 1.16, 95% CI: 1.02 to 1.32). When adding HIV and antiretroviral therapy-related variables, ORs were 1.04 (95% CI: 0.99 to 1.10) for small dense LDL and 1.13 (95% CI: 0.99 to 1.30) for apolipoprotein B. In both models, blood pressure and HIV viral load was independently associated with the odds for coronary events. CONCLUSIONS: HIV-infected patients receiving antiretroviral therapy with elevate small dense LDL and apolipoprotein B are at increased risk for coronary events as are patients without sustained HIV suppression.

Cholesterol transport from late endosomes to the Golgi regulates t-SNARE trafficking, assembly, and function

Cholesterol regulates plasma membrane (PM) association and functioning of syntaxin-4 and soluble N-ethylmaleimide-sensitive fusion protein 23 (SNAP23) in the secretory pathway. However, the molecular mechanism and cellular cholesterol pools that determine the localization and assembly of these target membrane SNAP receptors (t-SNAREs) are largely unknown. We recently demonstrated that high levels of annexin A6 (AnxA6) induce accumulation of cholesterol in late endosomes, thereby reducing cholesterol in the Golgi and PM. This leads to an impaired supply of cholesterol needed for cytosolic phospholipase A2 (cPLA2) to drive Golgi vesiculation and caveolin transport to the cell surface. Using AnxA6-overexpressing cells as a model for cellular cholesterol imbalance, we identify impaired cholesterol egress from late endosomes and diminution of Golgi cholesterol as correlating with the sequestration of SNAP23/syntaxin-4 in Golgi membranes. Pharmacological accumulation of late endosomal cholesterol and cPLA2 inhibition induces a similar phenotype in control cells with low AnxA6 levels. Ectopic expression of Niemann-Pick C1 (NPC1) or exogenous cholesterol restores the location of SNAP23 and syntaxin-4 within the PM. Importantly, AnxA6-mediated mislocalization of these t-SNAREs correlates with reduced secretion of cargo via the SNAP23/syntaxin-4¿dependent constitutive exocytic pathway. We thus conclude that inhibition of late endosomal export and Golgi cholesterol depletion modulate t-SNARE localization and functioning along the exocytic pathway.

Cholesterol and Fatty Acids Regulate Dynamic Caveolin Trafficking through the Golgi Complex and between the Cell Surface and Lipid Bodies

Caveolins are a crucial component of plasma membrane (PM) caveolae but have also been localized to intracellular compartments, including the Golgi complex and lipid bodies. Mutant caveolins associated with human disease show aberrant trafficking to the PM and Golgi accumulation. We now show that the Golgi pool of mainly newly synthesized protein is detergent-soluble and predominantly in a monomeric state, in contrast to the surface pool. Caveolin at the PM is not recognized by specific caveolin antibodies unless PM cholesterol is depleted. Exit from the Golgi complex of wild-type caveolin-1 or -3, but not vesicular stomatitis virus-G protein, is modulated by changing cellular cholesterol levels. In contrast, a muscular dystrophy-associated mutant of caveolin-3, Cav3P104L, showed increased accumulation in the Golgi complex upon cholesterol treatment. In addition, we demonstrate that in response to fatty acid treatment caveolin can follow a previously undescribed pathway from the PM to lipid bodies and can move from lipid bodies to the PM in response to removal of fatty acids. The results suggest that cholesterol is a rate-limiting component for caveolin trafficking. Changes in caveolin flux through the exocytic pathway can therefore be an indicator of cellular cholesterol and fatty acid levels.

Inhibition of Mitogen-Activated Protein Kinase Erk1/2 Promotes Protein Degradation of ATP Binding Cassette Transporters A1 and G1 in CHO and in HuH7 cells.

Signal transduction modulates expression and activity of cholesterol transporters. We recently demonstrated that the Ras/mitogen-activated protein kinase (MAPK) signaling cascade regulates protein stability of Scavenger Receptor BI (SR-BI) through Proliferator Activator Receptor (PPARα) -dependent degradation pathways. In addition, MAPK (Mek/Erk 1/2) inhibition has been shown to influence liver X receptor (LXR) -inducible ATP Binding Cassette (ABC) transporter ABCA1 expression in macrophages. Here we investigated if Ras/MAPK signaling could alter expression and activity of ABCA1 and ABCG1 in steroidogenic and hepatic cell lines. We demonstrate that in Chinese Hamster Ovary (CHO) cells and human hepatic HuH7 cells, extracellular signal-regulated kinase 1/2 (Erk1/2) inhibition reduces PPARα-inducible ABCA1 protein levels, while ectopic expression of constitutively active H-Ras, K-Ras and MAPK/Erk kinase 1 (Mek1) increases ABCA1 protein expression, respectively. Furthermore, Mek1/2 inhibitors reduce ABCG1 protein levels in ABCG1 overexpressing CHO cells (CHO-ABCG1) and human embryonic kidney 293 (HEK293) cells treated with LXR agonist. This correlates with Mek1/2 inhibition reducing ABCG1 cell surface expression and decreasing cholesterol efflux onto High Density Lipoproteins (HDL). Real Time reverse transcriptase polymerase chain reaction (RT-PCR) and protein turnover studies reveal that Mek1/2 inhibitors do not target transcriptional regulation of ABCA1 and ABCG1, but promote ABCA1 and ABCG1 protein degradation in HuH7 and CHO cells, respectively. In line with published data from mouse macrophages, blocking Mek1/2 activity upregulates ABCA1 and ABCG1 protein levels in human THP1 macrophages, indicating opposite roles for the Ras/MAPK pathway in the regulation of ABC transporter activity in macrophages compared to steroidogenic and hepatic cell types. In summary, this study suggests that Ras/MAPK signaling modulates PPARα- and LXR-dependent protein degradation pathways in a cell-specific manner to regulate the expression levels of ABCA1 and ABCG1 transporters.

Améliorer la prescription d'un traitement hypolipémiant lors d'un syndrome coronarien aigu: absence de benefice d'une alerte automatique.

Introduction: Seulement 25-30% des patients avec syndrome coronarien aigu (SCA) atteignent les valeurs cibles de LDL-cholestérol (LDL-C) dans leur suivi. L'objectif de cette étude pré/post est de tester une alerte automatique centralisée pour améliorer les pratiques. L'alerte apparaît sur les feuilles de laboratoire pour tous les patients ayant une troponine >= 0,1 microg/l; elle précise notamment les recommandations en matière de profil lipidique (LDL-C cible) à atteindre. Méthode: Tout patient admis au CHUV pour un SCA avec troponine >= 0,1 microg/l était éligible. Durant les 2 phases de l'étude (du 23.11.2008 au 15.08.201), un bilan lipidique complet a été dosé à l'admission et à 3 mois. La phase 1 (pré) était observationnelle et le message d'alerte a été introduit pour la phase 2 (post). Résultats: Phase 1: 157 patients dont 56 (35%) étaient déjà traités par une statine: 114 hommes (âge moyen 62 ans) et 43 femmes (73 ans). LDL-C moyen: 3,4 ± 1,0 mmol/l à l'admission et 2,4 ± 0,8 mmol/l à 3 mois (p <0,001). Phase 2: 140 patients dont 42 (30%) étaient déjà traités par une statine: 116 hommes (62 ans) et 24 femmes (67 ans). LDL-C moyen: 3,4 ± 1,1 mmol/l à l'admission et 2,2 ± 1,0 mmol/l à 3 mois (p <0,001). 66 % (104 patients) atteignent un LDL-C cible < = 2,6 mmol/l à 3 mois lors de la phase 1, versus 78% (110 patients) lors de la phase 2 (p = 0,2). Les patients déjà sous statine à l'admission ont une faible diminution du LDL-C à 3 mois (de 2,8 à 2,5 mmol/l phase 1, p <0,05; de 2,5 à 2,6 mmol/l phase 2, p = 0.2), alors que les patients chez qui une statine est introduite à l'admission ont une baisse significative et plus importante du LDL-C à 3 mois (de 3,8 à 2,3 mmol/l phase 1, p <0,001; de 3,7 à 2,1 mmol/l phase 2, p <0,001) que les patients déjà sous statine au préalable. Conclusion: La phase observationnelle montre un taux élevé de patients atteignant un LDL-C cible à 3 mois. L'introduction d'une alarme automatique centralisée n'a pas permis d'améliorer ces résultats. Par contre, les patients arrivant à l'hôpital avec un SCA et étant déjà sous statine devraient avoir une intensification de leur traitement.

Prescription d'un traitement hypolipémiant lors d'un syndrome coronarien aigu en fonction du profil lipidique

Introduction: Les données épidémiologiques montrent que seuls25-30% des patients avec syndrome coronarien aigu (SCA) atteignentles valeurs cibles de LDL-cholestérol (LDL-C). Les recommandationsexistantes précisent le choix et le dosage des statines à utiliser enfonction du LDL-C cible souhaité. Le but de cette étude observationnelleétait de connaître les pratiques actuelles au CHUV avant d'introduireune étude d'intervention.Méthode: Pour être inclus, les patients devaient être admis au CHUVpour un SCA avec troponine positive (>= 0.1 microg/l) entre le23.11.2008 et le 29.05.2010. Un bilan lipidique complet (CT, HDL-C,LDL-C, TG) a été dosé à leur admission et un nouveau contrôle desparamètres lipidiques a été effectué à 3 mois. Les hypolipémiantsutilisés durant cette période ont été analysés pour chaque patient.Résultats: 141 patients, 101 hommes (âge moyen 63 ± 13 ans) et 40femmes (âge moyen 73 ± 13 ans) admis aux urgences pour un SCAavec troponine positive ont été inclus. La valeur moyenne du LDL-C àl'admission était de 3,4 ± 1,1 mmol/l (hommes 3,5 ± 1,1; femmes 3,3 ±1,1) et de 2,4 ± 0,8 mmol/l (hommes 2,4 ± 0,8; femmes 2,2 ± 0,7) aucontrôle de 3 mois. Parmi ces 141 patients, 52 (37%) étaient déjàtraités par une statine (36 hommes et 16 femmes). Leur valeur deLDL-C à l'admission était de 2,8 ± 0,9 mmol/l et de 2,5 ± 0,6 mmol/l aucontrôle de 3 mois. 7 patients (13%) ont eu une augmentation dudosage de leur statine, 14 patients (27%) ont eu un changement destatine et 31 patients (60%) n'ont eu aucune modification de leurtraitement. 89 patients n'avaient pas de statine (65 hommes et 24femmes) à leur admission mais ont quitté l'hôpital sous une statine.Leur valeur de LDL-C à l'entrée s'élevait à 3,8 ± 1 mmol/l et à 2,3 ± 0,8mmol/l au contrôle de 3 mois.Conclusion: Chez les patients hospitalisés pour un SCA mais sanstraitement par statine préalable, les résultats montrent une bonneadéquation (peut-être liée au hasard au vu d'une prescriptionstandardisée) entre le traitement prescrit et l'obtention d'un LDL-C ciblesouhaitable à 3 mois. Chez les patients déjà sous traitement de statine,les résultats montrent une faible baisse du LDL-C à 3 mois malgré leurrisque cardio-vasculaire plus élevé. Une prise en charge individualiséesemble d'autant plus nécessaire que le risque est élevé.

Risk factors of stroke mortality in the African region: a cohort study

Background: Population-based cohort studies of risk factors of stroke are scarce in developing countries and none has been done in the African region. We conducted a longitudinal study in the Seychelles (Indian Ocean, east of Kenya), a middle-income island state where the majority of the population is of African descent. Such data in Africa are important for international comparison and for advocacy in the region. Methods: Three examination surveys of cardiovascular risk factors were performed in independent samples representative of the general population aged 25-64 in 1989, 1994 and 2004 (n=1081, 1067, and 1255, respectively). Baseline risk factors data were linked with cause-specific mortality from vital statistics up to May 2007 (all deaths are medically certified in the Seychelles and kept in an electronic database). We considered stroke (any type) as a cause of death if the diagnosis was reported in any of the 4 fields in the death certificates for underlying and concomitant causes of death. Results. Among the 2479 persons aged 35-64 at baseline, 280 died including 56 with stroke during follow up (maximum: 18.2 years; mean: 10.2 years). In this age range, age-adjusted mortality rates (/100'000/year) were 969 for all cause and 187 for stroke; age-adjusted prevalence of high blood pressure (≥140/90 mmHg) was 48%. In multivariate Cox survival time regression, stroke mortality was increased by 18% and 35% for a 10-mmHg increase in systolic, respectively diastolic BP (p<0.001). Stroke mortality was also associated with age, smoking ≥5 cigarettes vs. no smoking (HR: 2.4; 95% CI: 1.2-4.8) and diabetes (HR: 1.9; 1.02-3.6) but not with sex, LDL-cholesterol intake, alcohol intake and professional occupation. Conclusion. This first population-based cohort study in the African region demonstrates high mortality rates from stroke in middle-aged adults and confirms associations with high BP and other risk factors. This emphasizes the importance of reducing BP and other modifiable risk factors in high risk individuals and in the general population as a main strategy to reduce the burden of stroke.

Age and gender differences in the social patterning of cardiovascular risk factors in Switzerland: the CoLaus study.

OBJECTIVES: We examined the social distribution of a comprehensive range of cardiovascular risk factors (CVRF) in a Swiss population and assessed whether socioeconomic differences varied by age and gender. METHODS: Participants were 2960 men and 3343 women aged 35-75 years from a population-based survey conducted in Lausanne, Switzerland (CoLaus study). Educational level was the indicator of socioeconomic status used in this study. Analyses were stratified by gender and age group (35-54 years; 55-75 years). RESULTS: There were large educational differences in the prevalence of CVRF such as current smoking (Δ = absolute difference in prevalence between highest and lowest educational group:15.1%/12.6% in men/women aged 35-54 years), physical inactivity (Δ = 25.3%/22.7% in men/women aged 35-54 years), overweight and obesity (Δ = 14.6%/14.8% in men/women aged 55-75 years for obesity), hypertension (Δ = 16.7%/11.4% in men/women aged 55-75 years), dyslipidemia (Δ = 2.8%/6.2% in men/women aged 35-54 years for high LDL-cholesterol) and diabetes (Δ = 6.0%/2.6% in men/women aged 55-75 years). Educational inequalities in the distribution of CVRF were larger in women than in men for alcohol consumption, obesity, hypertension and dyslipidemia (p<0.05). Relative educational inequalities in CVRF tended to be greater among the younger (35-54 years) than among the older age group (55-75 years), particularly for behavioral CVRF and abdominal obesity among men and for physiological CVRF among women (p<0.05). CONCLUSION: Large absolute differences in the prevalence of CVRF according to education categories were observed in this Swiss population. The socioeconomic gradient in CVRF tended to be larger in women and in younger persons.

The association between inflammatory biomarkers and metabolically healthy obesity depends of the definition used.

BACKGROUND/OBJECTIVES: To assess the distribution of interleukin (IL)-1β, IL-6, tumour necrosis factor (TNF)-α and C-reactive protein (CRP) according to the different definitions of metabolically healthy obesity (MHO). SUBJECTS/METHODS: A total of 881 obese (body mass index (BMI) > or =30 kg/m2) subjects derived from the population-based CoLaus Study participated in this study. MHO was defined using six sets of criteria including different combinations of waist, blood pressure, total high-density lipoprotein cholesterol or low-density lipoprotein -cholesterol, triglycerides, fasting glucose, homeostasis model, high-sensitivity CRP, and personal history of cardiovascular, respiratory or metabolic diseases. IL-1β, IL-6 and TNF-α were assessed by multiplexed flow cytometric assay. CRP was assessed by immunoassay. RESULTS: On bivariate analysis some, but not all, definitions of MHO led to significantly lower levels of IL-6, TNF-α and CRP compared with non-MH obese subjects. Most of these differences became nonsignificant after multivariate analysis. An posteriori analysis showed a statistical power between 9 and 79%, depending on the inflammatory biomarker and MHO definition considered. Further increasing sample size to overweight+obese individuals (BMI > or =25 kg/m2, n=2917) showed metabolically healthy status to be significantly associated with lower levels of CRP, while no association was found for IL-1β. Significantly lower IL-6 and TNF-α levels were also found with some but not all MHO definitions, the differences in IL-6 becoming nonsignificant after adjusting for abdominal obesity or percent body fat. CONCLUSIONS: MHO individuals present with decreased levels of CRP and, depending on MHO definition, also with decreased levels in IL-6 and TNF-α. Conversely, no association with IL-1β levels was found.

Obesity and the incidence of chronic diseases: a seemingly unrelated probit approach

Western societies can reduce avoidable mortality and morbidity by better understanding the relationship between obesity and chronic disease. This paper examines the joint determinants of obesity and of heart disease, diabetes, hypertension, and elevated cholesterol. It analyzes a broadly representative Spanish dataset, the 1999 Survey on Disabilities, Impairments and Health Status, using a health production theoretical framework together with a seemingly unrelated probit model approach that controls for unobserved heterogeneity and endogeneity. Its findings provide suggestive evidence of a positive and significant, although specification-dependent, association between obesity and the prevalence of chronic illness

Socioeconomics status and quality of care in a population-based sample of swiss diabetic patients

Introduction: Low socioeconomic status (SES) is associated with higher prevalence of diabetes and worse outcomes; it has also been shown to be associated with worse quality of care. We aimed to explore the relationship between SES and quality of care in the Swiss context. Methods: We used data from a population-based survey including 519 adult diabetic patients living in the canton of Vaud. Self-reported data on patients' and diabetes characteristics, indicators of process and outcomes of care and quality of life were collected. Dependent variables included 6 processes of care (PoC) received during the last 12 months (HbA1C, lipid, microalbuminuria, fundoscopy, feet examination and influenza vaccination) and selected clinical outcomes (blood pressure, LDL, HbA1C, diabetes-specific (ADDQoL) and generic quality of life (SF-12)). Regression analyses were performed to assess the relationship between education and income, respectively, and quality of care as measured by PoC and clinical outcomes. Adjustment was made for age, gender and comorbidities. Results: Mean age was 64.5 years, 40% were women; 19%, 56% and 25% of the patients reported primary (I), secondary (II) and tertiary (III) education. Fundoscopy was the only PoC significantly associated with education, with III education patients more likely to get the exam than those with primary education (adjOR 1.8, 95% CI 1.0-3.3). Use of composite indicators of PoC showed that compared to patients with primary education, patients with III education were more likely to receive ≥5/6 PoC (adjOR 1.9, 95% CI 1.1-3.4), and that those with II or III education were more likely to receive 4/4 PoC (adjOR 1.9, 95% CI 1.0-3.3; adjOR 2.1, 95% CI 1.1-4.1, respectively). Quality of life was the only clinical outcome significantly associated with education, with II and III education patients reporting better quality of life compared to primary education patients, as measured by the ADDQoL (β 0.6, 95% CI 0.3-1.0, β 0.6, 95% CI 0.2-1.0, respectively) and the physical component score of the SF-12 (β 2.5, 95% CI 0.2-4.8, β 3.6, 95% CI 0.9-6.4, respectively). No associations were found between income and quality of care. Conclusion: Social inequalities have been demonstrated in Switzerland for global health indicators. Our results suggest that similar associations are found when considering quality of care measures in individuals with diabetes, but only for a few indicators.

Biological, clinical and population relevance of 95 loci for blood lipids.

Plasma concentrations of total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol and triglycerides are among the most important risk factors for coronary artery disease (CAD) and are targets for therapeutic intervention. We screened the genome for common variants associated with plasma lipids in >100,000 individuals of European ancestry. Here we report 95 significantly associated loci (P < 5 x 10(-8)), with 59 showing genome-wide significant association with lipid traits for the first time. The newly reported associations include single nucleotide polymorphisms (SNPs) near known lipid regulators (for example, CYP7A1, NPC1L1 and SCARB1) as well as in scores of loci not previously implicated in lipoprotein metabolism. The 95 loci contribute not only to normal variation in lipid traits but also to extreme lipid phenotypes and have an impact on lipid traits in three non-European populations (East Asians, South Asians and African Americans). Our results identify several novel loci associated with plasma lipids that are also associated with CAD. Finally, we validated three of the novel genes-GALNT2, PPP1R3B and TTC39B-with experiments in mouse models. Taken together, our findings provide the foundation to develop a broader biological understanding of lipoprotein metabolism and to identify new therapeutic opportunities for the prevention of CAD.

Metastable liquid-liquid phase transition in a single-component system with only one crystal phase and no density anomaly

We investigate the phase behavior of a single-component system in three dimensions with spherically-symmetric, pairwise-additive, soft-core interactions with an attractive well at a long distance, a repulsive soft-core shoulder at an intermediate distance, and a hard-core repulsion at a short distance, similar to potentials used to describe liquid systems such as colloids, protein solutions, or liquid metals. We showed [Nature (London) 409, 692 (2001)] that, even with no evidence of the density anomaly, the phase diagram has two first-order fluid-fluid phase transitions, one ending in a gas¿low-density-liquid (LDL) critical point, and the other in a gas¿high-density-liquid (HDL) critical point, with a LDL-HDL phase transition at low temperatures. Here we use integral equation calculations to explore the three-parameter space of the soft-core potential and perform molecular dynamics simulations in the interesting region of parameters. For the equilibrium phase diagram, we analyze the structure of the crystal phase and find that, within the considered range of densities, the structure is independent of the density. Then, we analyze in detail the fluid metastable phases and, by explicit thermodynamic calculation in the supercooled phase, we show the absence of the density anomaly. We suggest that this absence is related to the presence of only one stable crystal structure.

Finite-size scaling investigation of the liquid-liquid critical point in ST2 water and its stability with respect to crystallization.

The liquid-liquid critical point scenario of water hypothesizes the existence of two metastable liq- uid phases low-density liquid (LDL) and high-density liquid (HDL) deep within the supercooled region. The hypothesis originates from computer simulations of the ST2 water model, but the stabil- ity of the LDL phase with respect to the crystal is still being debated. We simulate supercooled ST2 water at constant pressure, constant temperature, and constant number of molecules N for N ≤ 729 and times up to 1 μs. We observe clear differences between the two liquids, both structural and dynamical. Using several methods, including finite-size scaling, we confirm the presence of a liquid-liquid phase transition ending in a critical point. We find that the LDL is stable with respect to the crystal in 98% of our runs (we perform 372 runs for LDL or LDL-like states), and in 100% of our runs for the two largest system sizes (N = 512 and 729, for which we perform 136 runs for LDL or LDL-like states). In all these runs, tiny crystallites grow and then melt within 1 μs. Only for N ≤ 343 we observe six events (over 236 runs for LDL or LDL-like states) of spontaneous crystal- lization after crystallites reach an estimated critical size of about 70 ± 10 molecules.