Adaptive Algorithms for Personalized Diabetes Treatment

Dynamic systems, especially in real-life applications, are often determined by inter-/intra-variability, uncertainties and time-varying components. Physiological systems are probably the most representative example in which population variability, vital signal measurement noise and uncertain dynamics render their explicit representation and optimization a rather difficult task. Systems characterized by such challenges often require the use of adaptive algorithmic solutions able to perform an iterative structural and/or parametrical update process towards optimized behavior. Adaptive optimization presents the advantages of (i) individualization through learning of basic system characteristics, (ii) ability to follow time-varying dynamics and (iii) low computational cost. In this chapter, the use of online adaptive algorithms is investigated in two basic research areas related to diabetes management: (i) real-time glucose regulation and (ii) real-time prediction of hypo-/hyperglycemia. The applicability of these methods is illustrated through the design and development of an adaptive glucose control algorithm based on reinforcement learning and optimal control and an adaptive, personalized early-warning system for the recognition and alarm generation against hypo- and hyperglycemic events.

Statin therapy and long-term adverse limb outcomes in patients with peripheral artery disease: insights from the REACH registry

AIMS Due to a high burden of systemic cardiovascular events, current guidelines recommend the use of statins in all patients with peripheral artery disease (PAD). We sought to study the impact of statin use on limb prognosis in patients with symptomatic PAD enrolled in the international REACH registry. METHODS Statin use was assessed at study enrolment, as well as a time-varying covariate. Rates of the primary adverse limb outcome (worsening claudication/new episode of critical limb ischaemia, new percutaneous/surgical revascularization, or amputation) at 4 years and the composite of cardiovascular death/myocardial infarction/stroke were compared among statin users vs. non-users. RESULTS A total of 5861 patients with symptomatic PAD were included. Statin use at baseline was 62.2%. Patients who were on statins had a significantly lower risk of the primary adverse limb outcome at 4 years when compared with those who were not taking statins [22.0 vs. 26.2%; hazard ratio (HR), 0.82; 95% confidence interval (CI), 0.72-0.92; P = 0.0013]. Results were similar when statin use was considered as a time-dependent variable (P = 0.018) and on propensity analysis (P < 0.0001). The composite of cardiovascular death/myocardial infarction/stroke was similarly reduced (HR, 0.83; 95% CI, 0.73-0.96; P = 0.01). CONCLUSION Among patients with PAD in the REACH registry, statin use was associated with an ∼18% lower rate of adverse limb outcomes, including worsening symptoms, peripheral revascularization, and ischaemic amputations. These findings suggest that statin therapy not only reduces the risk of adverse cardiovascular events, but also favourably affects limb prognosis in patients with PAD.

Polypharmacy is Associated with an Increased Risk of Bleeding in Elderly Patients with Venous Thromboembolism.

BACKGROUND Polypharmacy, defined as the concomitant use of multiple medications, is very common in the elderly and may trigger drug-drug interactions and increase the risk of falls in patients receiving vitamin K antagonists. OBJECTIVE To examine whether polypharmacy increases the risk of bleeding in elderly patients who receive vitamin K antagonists for acute venous thromboembolism (VTE). DESIGN We used a prospective cohort study. PARTICIPANTS In a multicenter Swiss cohort, we studied 830 patients aged ≥ 65 years with VTE. MAIN MEASURES We defined polypharmacy as the prescription of more than four different drugs. We assessed the association between polypharmacy and the time to a first major and clinically relevant non-major bleeding, accounting for the competing risk of death. We adjusted for known bleeding risk factors (age, gender, pulmonary embolism, active cancer, arterial hypertension, cardiac disease, cerebrovascular disease, chronic liver and renal disease, diabetes mellitus, history of major bleeding, recent surgery, anemia, thrombocytopenia) and periods of vitamin K antagonist treatment as a time-varying covariate. KEY RESULTS Overall, 413 (49.8 %) patients had polypharmacy. The mean follow-up duration was 17.8 months. Patients with polypharmacy had a significantly higher incidence of major (9.0 vs. 4.1 events/100 patient-years; incidence rate ratio [IRR] 2.18, 95 % confidence interval [CI] 1.32-3.68) and clinically relevant non-major bleeding (14.8 vs. 8.0 events/100 patient-years; IRR 1.85, 95 % CI 1.27-2.71) than patients without polypharmacy. After adjustment, polypharmacy was significantly associated with major (sub-hazard ratio [SHR] 1.83, 95 % CI 1.03-3.25) and clinically relevant non-major bleeding (SHR 1.60, 95 % CI 1.06-2.42). CONCLUSIONS Polypharmacy is associated with an increased risk of both major and clinically relevant non-major bleeding in elderly patients receiving vitamin K antagonists for VTE.

Risk of falls and bleeding in elderly patients with acute venous thromboembolism.

OBJECTIVE Whether or not a high risk of falls increases the risk of bleeding in patients receiving anticoagulants remains a matter of debate. METHODS We conducted a prospective cohort study involving 991 patients ≥ 65 years of age who received anticoagulants for acute venous thromboembolism (VTE) at nine Swiss hospitals between September 2009 and September 2012. The study outcomes were as follows: the time to a first major episode of bleeding; and clinically relevant nonmajor bleeding. We determined the associations between the risk of falls and the time to a first episode of bleeding using competing risk regression, accounting for death as a competing event. We adjusted for known bleeding risk factors and anticoagulation as a time-varying covariate. RESULTS Four hundred fifty-eight of 991 patients (46%) were at high risk of falls. The mean duration of follow-up was 16.7 months. Patients at high risk of falls had a higher incidence of major bleeding (9.6 vs. 6.6 events/100 patient-years; P = 0.05) and a significantly higher incidence of clinically relevant nonmajor bleeding (16.7 vs. 8.3 events/100 patient-years; P < 0.001) than patients at low risk of falls. After adjustment, a high risk of falls was associated with clinically relevant nonmajor bleeding [subhazard ratio (SHR) = 1.74, 95% confidence interval (CI) = 1.23-2.46], but not with major bleeding (SHR = 1.24, 95% CI = 0.83-1.86). CONCLUSION In elderly patients who receive anticoagulants because of VTE, a high risk of falls is significantly associated with clinically relevant nonmajor bleeding, but not with major bleeding. Whether or not a high risk of falls is a reason against providing anticoagulation beyond 3 months should be based on patient preferences and the risk of VTE recurrence.

Physical Activity and Risk of Bleeding in Elderly Patients Taking Anticoagulants.

BACKGROUND Although the possibility of bleeding during anticoagulant treatment may limit patients from taking part in physical activity, the association between physical activity and anticoagulation-related bleeding is uncertain. OBJECTIVES To determine whether physical activity is associated with bleeding in elderly patients taking anticoagulants. PATIENTS/METHODS In a prospective multicenter cohort study of 988 patients aged ≥65 years receiving anticoagulants for venous thromboembolism, we assessed patients' self-reported physical activity level. The primary outcome was the time to a first major bleeding, defined as fatal bleeding, symptomatic bleeding in a critical site, or bleeding causing a fall in hemoglobin or leading to transfusions. The secondary outcome was the time to a first clinically-relevant non-major bleeding. We examined the association between physical activity level and time to a first bleeding using competing risk regression, accounting for death as a competing event. We adjusted for known bleeding risk factors and anticoagulation as a time-varying covariate. RESULTS During a mean follow-up of 22 months, patients with a low, moderate, and high physical activity level had an incidence of major bleeding of 11.6, 6.3, and 3.1 events per 100 patient-years, and an incidence of clinically relevant non-major bleeding of 14.0, 10.3, and 7.7 events per 100 patient-years, respectively. A high physical activity level was significantly associated with a lower risk of major bleeding (adjusted sub-hazard ratio 0.40, 95%-CI 0.22-0.72). There was no association between physical activity and non-major bleeding. CONCLUSIONS A high level of physical activity is associated with a decreased risk of major bleeding in elderly patients receiving anticoagulant therapy. This article is protected by copyright. All rights reserved.

Atherogenic dyslipidemia and residual cardiovascular risk in statin-treated patients

BACKGROUND AND PURPOSE Treatment with statins reduces the rate of cardiovascular events in high-risk patients, but residual risk persists. At least part of that risk may be attributable to atherogenic dyslipidemia characterized by low high-density lipoprotein cholesterol (≤40 mg/dL) and high triglycerides (triglycerides≥150 mg/dL). METHODS We studied subjects with stroke or transient ischemic attack in the Prevention of Cerebrovascular and Cardiovascular Events of Ischemic Origin With Terutroban in Patients With a History of Ischemic Stroke or Transient Ischemic Attack (PERFORM; n=19,100) and Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL; n=4731) trials who were treated with a statin and who had high-density lipoprotein cholesterol and triglycerides measurements 3 months after randomization (n=10,498 and 2900, respectively). The primary outcome measure for this exploratory analysis was the occurrence of major cardiovascular events (nonfatal myocardial infarction, nonfatal stroke, or cardiovascular death). We also performed a time-varying analysis to account for all available high-density lipoprotein cholesterol and triglyceride measurements. RESULTS A total of 10% of subjects in PERFORM and 9% in SPARCL had atherogenic dyslipidemia after ≥3 months on start statin therapy. After a follow-up of 2.3 years (PERFORM) and 4.9 years (SPARCL), a major cardiovascular event occurred in 1123 and 485 patients in the 2 trials, respectively. The risk of major cardiovascular events was higher in subjects with versus those without atherogenic dyslipidemia in both PERFORM (hazard ratio, 1.36; 95% confidence interval, 1.14-1.63) and SPARCL (hazard ratio, 1.40; 95% confidence interval, 1.06-1.85). The association was attenuated after multivariable adjustment (hazard ratio, 1.23; 95% confidence interval, 1.03-1.48 in PERFORM and hazard ratio, 1.24; 95% confidence interval, 0.93-1.65 in SPARCL). Time-varying analysis confirmed these findings. CONCLUSIONS The presence of atherogenic dyslipidemia was associated with higher residual cardiovascular risk in PERFORM and SPARCL subjects with stroke or transient ischemic attack receiving statin therapy. Specific therapeutic interventions should now be trialed to address this residual risk.

Autologous hematopoietic stem cell transplantation vs intravenous pulse cyclophosphamide in diffuse cutaneous systemic sclerosis: a randomized clinical trial

IMPORTANCE High-dose immunosuppressive therapy and autologous hematopoietic stem cell transplantation (HSCT) have shown efficacy in systemic sclerosis in phase 1 and small phase 2 trials. OBJECTIVE To compare efficacy and safety of HSCT vs 12 successive monthly intravenous pulses of cyclophosphamide. DESIGN, SETTING, AND PARTICIPANTS The Autologous Stem Cell Transplantation International Scleroderma (ASTIS) trial, a phase 3, multicenter, randomized (1:1), open-label, parallel-group, clinical trial conducted in 10 countries at 29 centers with access to a European Group for Blood and Marrow Transplantation-registered transplant facility. From March 2001 to October 2009, 156 patients with early diffuse cutaneous systemic sclerosis were recruited and followed up until October 31, 2013. INTERVENTIONS HSCT vs intravenous pulse cyclophosphamide. MAIN OUTCOMES AND MEASURES The primary end point was event-free survival, defined as time from randomization until the occurrence of death or persistent major organ failure. RESULTS A total of 156 patients were randomly assigned to receive HSCT (n = 79) or cyclophosphamide (n = 77). During a median follow-up of 5.8 years, 53 events occurred: 22 in the HSCT group (19 deaths and 3 irreversible organ failures) and 31 in the control group (23 deaths and 8 irreversible organ failures). During the first year, there were more events in the HSCT group (13 events [16.5%], including 8 treatment-related deaths) than in the control group (8 events [10.4%], with no treatment-related deaths). At 2 years, 14 events (17.7%) had occurred cumulatively in the HSCT group vs 14 events (18.2%) in the control group; at 4 years, 15 events (19%) had occurred cumulatively in the HSCT group vs 20 events (26%) in the control group. Time-varying hazard ratios (modeled with treatment × time interaction) for event-free survival were 0.35 (95% CI, 0.16-0.74) at 2 years and 0.34 (95% CI, 0.16-0.74) at 4 years. CONCLUSIONS AND RELEVANCE Among patients with early diffuse cutaneous systemic sclerosis, HSCT was associated with increased treatment-related mortality in the first year after treatment. However, HCST conferred a significant long-term event-free survival benefit. TRIAL REGISTRATION isrctn.org Identifier: ISRCTN54371254.

Dynamic directed interictal connectivity in left and right temporal lobe epilepsy.

OBJECTIVE There is increasing evidence that epileptic activity involves widespread brain networks rather than single sources and that these networks contribute to interictal brain dysfunction. We investigated the fast-varying behavior of epileptic networks during interictal spikes in right and left temporal lobe epilepsy (RTLE and LTLE) at a whole-brain scale using directed connectivity. METHODS In 16 patients, 8 with LTLE and 8 with RTLE, we estimated the electrical source activity in 82 cortical regions of interest (ROIs) using high-density electroencephalography (EEG), individual head models, and a distributed linear inverse solution. A multivariate, time-varying, and frequency-resolved Granger-causal modeling (weighted Partial Directed Coherence) was applied to the source signal of all ROIs. A nonparametric statistical test assessed differences between spike and baseline epochs. Connectivity results between RTLE and LTLE were compared between RTLE and LTLE and with neuropsychological impairments. RESULTS Ipsilateral anterior temporal structures were identified as key drivers for both groups, concordant with the epileptogenic zone estimated invasively. We observed an increase in outflow from the key driver already before the spike. There were also important temporal and extratemporal ipsilateral drivers in both conditions, and contralateral only in RTLE. A different network pattern between LTLE and RTLE was found: in RTLE there was a much more prominent ipsilateral to contralateral pattern than in LTLE. Half of the RTLE patients but none of the LTLE patients had neuropsychological deficits consistent with contralateral temporal lobe dysfunction, suggesting a relationship between connectivity changes and cognitive deficits. SIGNIFICANCE The different patterns of time-varying connectivity in LTLE and RTLE suggest that they are not symmetrical entities, in line with our neuropsychological results. The highest outflow region was concordant with invasive validation of the epileptogenic zone. This enhanced characterization of dynamic connectivity patterns could better explain cognitive deficits and help the management of epilepsy surgery candidates.

Life-span development of self-esteem and its effects on important life outcomes

We examined the life-span development of self-esteem and tested whether self-esteem influences the development of important life outcomes, including relationship satisfaction, job satisfaction, occupational status, salary, positive and negative affect, depression, and physical health. Data came from the Longitudinal Study of Generations. Analyses were based on 5 assessments across a 12-year period of a sample of 1,824 individuals ages 16 to 97 years. First, growth curve analyses indicated that self-esteem increases from adolescence to middle adulthood, reaches a peak at about age 50 years, and then decreases in old age. Second, cross-lagged regression analyses indicated that self-esteem is best modeled as a cause rather than a consequence of life outcomes. Third, growth curve analyses, with self-esteem as a time-varying covariate, suggested that self-esteem has medium-sized effects on life-span trajectories of affect and depression, small to medium-sized effects on trajectories of relationship and job satisfaction, a very small effect on the trajectory of health, and no effect on the trajectory of occupational status. These findings replicated across 4 generations of participants— children, parents, grandparents, and their great-grandparents. Together, the results suggest that self-esteem has a significant prospective impact on real-world life experiences and that high and low self-esteem are not mere epiphenomena of success and failure in important life domains.

Effect of surface restoring on subsurface variability in a climate model during 1949–2005

Initializing the ocean for decadal predictability studies is a challenge, as it requires reconstructing the little observed subsurface trajectory of ocean variability. In this study we explore to what extent surface nudging using well-observed sea surface temperature (SST) can reconstruct the deeper ocean variations for the 1949–2005 period. An ensemble made with a nudged version of the IPSLCM5A model and compared to ocean reanalyses and reconstructed datasets. The SST is restored to observations using a physically-based relaxation coefficient, in contrast to earlier studies, which use a much larger value. The assessment is restricted to the regions where the ocean reanalyses agree, i.e. in the upper 500 m of the ocean, although this can be latitude and basin dependent. Significant reconstruction of the subsurface is achieved in specific regions, namely region of subduction in the subtropical Atlantic, below the thermocline in the equatorial Pacific and, in some cases, in the North Atlantic deep convection regions. Beyond the mean correlations, ocean integrals are used to explore the time evolution of the correlation over 20-year windows. Classical fixed depth heat content diagnostics do not exhibit any significant reconstruction between the different existing bservation-based references and can therefore not be used to assess global average time-varying correlations in the nudged simulations. Using the physically based average temperature above an isotherm (14°C) alleviates this issue in the tropics and subtropics and shows significant reconstruction of these quantities in the nudged simulations for several decades. This skill is attributed to the wind stress reconstruction in the tropics, as already demonstrated in a perfect model study using the same model. Thus, we also show here the robustness of this result in an historical and observational context.

Comparative effectiveness of immediate antiretroviral therapy versus CD4-based initiation in HIV-positive individuals in high-income countries: observational cohort study.

BACKGROUND Recommendations have differed nationally and internationally with respect to the best time to start antiretroviral therapy (ART). We compared effectiveness of three strategies for initiation of ART in high-income countries for HIV-positive individuals who do not have AIDS: immediate initiation, initiation at a CD4 count less than 500 cells per μL, and initiation at a CD4 count less than 350 cells per μL. METHODS We used data from the HIV-CAUSAL Collaboration of cohort studies in Europe and the USA. We included 55 826 individuals aged 18 years or older who were diagnosed with HIV-1 infection between January, 2000, and September, 2013, had not started ART, did not have AIDS, and had CD4 count and HIV-RNA viral load measurements within 6 months of HIV diagnosis. We estimated relative risks of death and of death or AIDS-defining illness, mean survival time, the proportion of individuals in need of ART, and the proportion of individuals with HIV-RNA viral load less than 50 copies per mL, as would have been recorded under each ART initiation strategy after 7 years of HIV diagnosis. We used the parametric g-formula to adjust for baseline and time-varying confounders. FINDINGS Median CD4 count at diagnosis of HIV infection was 376 cells per μL (IQR 222-551). Compared with immediate initiation, the estimated relative risk of death was 1·02 (95% CI 1·01-1·02) when ART was started at a CD4 count less than 500 cells per μL, and 1·06 (1·04-1·08) with initiation at a CD4 count less than 350 cells per μL. Corresponding estimates for death or AIDS-defining illness were 1·06 (1·06-1·07) and 1·20 (1·17-1·23), respectively. Compared with immediate initiation, the mean survival time at 7 years with a strategy of initiation at a CD4 count less than 500 cells per μL was 2 days shorter (95% CI 1-2) and at a CD4 count less than 350 cells per μL was 5 days shorter (4-6). 7 years after diagnosis of HIV, 100%, 98·7% (95% CI 98·6-98·7), and 92·6% (92·2-92·9) of individuals would have been in need of ART with immediate initiation, initiation at a CD4 count less than 500 cells per μL, and initiation at a CD4 count less than 350 cells per μL, respectively. Corresponding proportions of individuals with HIV-RNA viral load less than 50 copies per mL at 7 years were 87·3% (87·3-88·6), 87·4% (87·4-88·6), and 83·8% (83·6-84·9). INTERPRETATION The benefits of immediate initiation of ART, such as prolonged survival and AIDS-free survival and increased virological suppression, were small in this high-income setting with relatively low CD4 count at HIV diagnosis. The estimated beneficial effect on AIDS is less than in recently reported randomised trials. Increasing rates of HIV testing might be as important as a policy of early initiation of ART. FUNDING National Institutes of Health.

Multifractional Poisson process, multistable subordinator and related limit theorems

We introduce a multistable subordinator, which generalizes the stable subordinator to the case of time-varying stability index. This enables us to define a multifractional Poisson process. We study properties of these processes and establish the convergence of a continuous-time random walk to the multifractional Poisson process.

The Great Moderation and the Relationship between Output Growth and Its Volatility

This study examines the effect of the Great Moderation on the relationship between U.S. output growth and its volatility over the period 1947 to 2006. First, we consider the possible effects of structural change in the volatility process. In so doing, we employ GARCH-M and ARCH-M specifications of the process describing output growth rate and its volatility with and without a one-time structural break in volatility. Second, our data analyses and empirical results suggest no significant relationship between the output growth rate and its volatility, favoring the traditional wisdom of dichotomy in macroeconomics. Moreover, the evidence shows that the time-varying variance falls sharply or even disappears once we incorporate a one-time structural break in the unconditional variance of output starting 1982 or 1984. That is, the integrated GARCH effect proves spurious. Finally, a joint test of a trend change and a one-time shift in the volatility process finds that the one-time shift dominates.

Empirical Analysis of Credit Risk Regime Switching and Temporal Conditional Default Correlation in Credit Default Swap Valuation: The Market liquidity effect

In this paper, we extend the debate concerning Credit Default Swap valuation to include time varying correlation and co-variances. Traditional multi-variate techniques treat the correlations between covariates as constant over time; however, this view is not supported by the data. Secondly, since financial data does not follow a normal distribution because of its heavy tails, modeling the data using a Generalized Linear model (GLM) incorporating copulas emerge as a more robust technique over traditional approaches. This paper also includes an empirical analysis of the regime switching dynamics of credit risk in the presence of liquidity by following the general practice of assuming that credit and market risk follow a Markov process. The study was based on Credit Default Swap data obtained from Bloomberg that spanned the period January 1st 2004 to August 08th 2006. The empirical examination of the regime switching tendencies provided quantitative support to the anecdotal view that liquidity decreases as credit quality deteriorates. The analysis also examined the joint probability distribution of the credit risk determinants across credit quality through the use of a copula function which disaggregates the behavior embedded in the marginal gamma distributions, so as to isolate the level of dependence which is captured in the copula function. The results suggest that the time varying joint correlation matrix performed far superior as compared to the constant correlation matrix; the centerpiece of linear regression models.

Modeling the Volatility of Real GDP Growth: The Case of Japan Revisited

Previous studies (e.g., Hamori, 2000; Ho and Tsui, 2003; Fountas et al., 2004) find high volatility persistence of economic growth rates using generalized autoregressive conditional heteroskedasticity (GARCH) specifications. This paper reexamines the Japanese case, using the same approach and showing that this finding of high volatility persistence reflects the Great Moderation, which features a sharp decline in the variance as well as two falls in the mean of the growth rates identified by Bai and Perronâs (1998, 2003) multiple structural change test. Our empirical results provide new evidence. First, excess kurtosis drops substantially or disappears in the GARCH or exponential GARCH model that corrects for an additive outlier. Second, using the outlier-corrected data, the integrated GARCH effect or high volatility persistence remains in the specification once we introduce intercept-shift dummies into the mean equation. Third, the time-varying variance falls sharply, only when we incorporate the break in the variance equation. Fourth, the ARCH in mean model finds no effects of our more correct measure of output volatility on output growth or of output growth on its volatility.