Skip Navigation
Aging and Health

Ravi Varadhan, Ph.D.
Associate Professor

Academic Degrees
PhD (Biostatistics), Johns Hopkins University
PhD (Environmental Engineering), University of Toledo

Departmental Affiliation
Division of Geriatric Medicine and Gerontology
Department of Medicine
Johns Hopkins University School of Medicine

Center and Institute Affiliations
Center on Aging and Health

Departmental Address
2024 E. Monument Street, suite 2-700
Baltimore, MD 21205-2223
Telephone: (410) 502-2619
Fax: (410) 614-9625


Research Interests
My research interests lie in three widely different areas: (i) aging and frailty, (ii) patient-centered outcomes research, and (iii) computational statistics. My work in aging and frailty is aimed at development of theoretical and statistical models for studying frailty as dysregulated homeostasis that increases the vulnerability to stressors. In particular, the focus is on the markers of HPA axis, sympathetic nervous systems and inflammation dynamical systems biology in frailty. Another major focus of mine is on developing statistical methods that inform evidence-based individualized medicine. I am interested in identifying factors that are responsible for the heterogeneity in how individuals respond to interventions, which is known as the heterogeneity of treatment effect (HTE). I am also developing a new statistical approach for cross-deign synthesis that applies sophisticated statistical methods to combine randomized, controlled trial and registry data to obtain valid evidence for groups such as elderly women, who are under-represented in the trials. Last, but not least, I am interested in developing algorithms and software for solving high-dimensional optimization problems arising in statistical modeling. I have developed a new class of algorithms called SQUAREM (squared extrapolation methods) for accelerating the convergence of expectation-maximization (EM) and minorization-maximization (MM) algorithms. The EM/MM algorithms are popular approaches in statistical modeling for maximum likelihood estimation. SQUAREM algorithms increase the speed of EM/MM algorithms substantially without compromising the robustness. I have written several R packages for solving various types of problems in statistical modeling and in optimization.

Honors and Awards

2012 John M. Chambers award for best statistical software for our R package, turboEM, which provides a suite of convergence acceleration schemes for EM and MM algorithms (jointly with Jennifer Bobb)

2004 Margaret Merrell award for outstanding doctoral research in Biostatistics. Department of Biostatistics, Johns Hopkins University.

2003 Louis I. and Tomas D. Dublin award for effective use of statistical reasoning and methods in epidemiology. Department of Biostatistics, Johns Hopkins University.

Selected Publications

Varadhan R, Weiss CO, Segal JB, Wu A, Scharfstein DO, Boyd C. Evaluating health outcomes in the presence of competing risks: A review of statistical methods and clinical applications. Medical Care. 2010; 48:S96-S105. (AHRQ Contract No. 290-05-0034)

Nash JC, Varadhan R. Unifying optimization algorithms to aid software system users: optimx for R. J Statistical Software. 2011; 43:1-14.

Kalyani RR, Varadhan R, Weiss CO, Fried LP, Cappola AR. Frailty Status and Altered Glucose-Insulin Dynamics. J Gerontol A Biol Sci Med Sci. 2012; 67:1300-1306.

Varadhan R, Weiss CO, Boyd CM. The potential for increasing quality and reducing costs. Ann Intern Med. 2011;155(8):564; author reply 564-5. PMID: 22007054

Weiss CO, Segal JB, Varadhan R. Assessing the applicability of trial evidence to a target sample in the presence of heterogeneity of treatment effect. Pharmacoepidemiol Drug Saf. 2012; 21 Suppl 2:121-9. PMID: 22552987

Kalyani RR, Varadhan R, Weiss CO, Fried LP, Cappola A. Frailty Status and Altered Dynamics of Circulating Energy Metabolism Hormones after Oral Glucose in Older Women. Journal of Nutrition, Health and Aging. 2012; 16:679-686.

Weiss CO, Cappola AR, Varadhan R, Fried LP. Resting Metabolic Rate Among Old-Old Women With and Without Frailty: Variability and Estimation of Energy Requirements. J Am Geriatr. Soc. 2012; 60: 1695-1700.

Puhan MA, Singh S, Weiss CO, Varadhan R, Boyd CM. A Framework for Organizing and Selecting Quantitative Approaches for Benefit Harm Assessment. BMC Medical Research Methodology. 2012; 12:173.

Boyd CM, Singh S, Varadhan R, Weiss CO, Sharma R, Bass EB, Puhan MA. Methods for Benefit and Harm Assessment in Systematic Reviews. Agency for Healthcare Research and Quality. Rockville, MD. Nov 2012. PMID: 23326898.

Varadhan R, Yao W, Matteini A, Beamer B, Xue QL, Yang H, Manwani B, Reiner A, Jenny N, Parekh N, Fallin DM, Newman A, Bandeen-Roche K, Tracy R, Ferrucci L, Walston J. Simple biologically informed inflammatory index of two serum cytokines predicts 10 year all-cause mortality in older adults. J Gerontology Series A. 2014; 69(2):165-73. PMCID: PMC3905766

Kovalchik SA, Varadhan R. Fitting Additive Binomial Regression Models with the R package blm. J Statistical Software. Vol. 54. August 2013

Holmes HM, Min LC, Yee M, Varadhan R, Basran J, Dale W, Boyd CM. Rationalizing prescribing for older patients with multimorbidity: considering time to benefit. Drugs & Aging. 2013. PMCID: PMC3755031

Kuk D, Varadhan R. Model selection in competing risks regression. Statistics in Medicine. 2013; 32(18): 3077-88. PMID: 23436643

Varadhan R, Segal JB, Boyd CM, Wu AW, Weiss CO. A framework for the analysis of heterogeneity of treatment effect in patient-centered outcomes research. J Clin Epi. 2013. PMID: 23651763

Kovalchik SA, Varadhan R, Weiss CO. Assessing heterogeneity of treatment effect in a clinical trial with the proportional interactions model. Statistics in Medicine. 2013;66(8):818-25. PMID: 23788362

Yu T, Fain K, Boyd CM, Singh S, Weiss CO, Li T, Varadhan R, Puhan MA. Benefits and harms of roflumilast in moderate to severe COPD. Thorax. 2013. Dec 17. PMID: 24347460

Varadhan R, Wang SJ. Standardization for subgroup analysis in randomized controlled trials. J Biopharm Stat. 2014. 24(1): 154-67. PMID: 24392983

Weiss CO, Varadhan R, Puhan MA, Vickers A, Bandeen-Roche K, Boyd CM, Kent DM. Multimorbidity and evidence generation. J Gen Intern Med. 2014. PMID: 24442333.

Xue QL, Varadhan R. What is missing in the validation of frailty instruments? J Am Med Dir Assoc. 2014. 15(2): 141-2. PMID: 24405640

Click Here to View Selected Recent Publications


Here are the R packages that I have developed:

SQUAREM: Squared extrapolation methods for accelerating fixed-point iterations.
Algorithms for accelerating the convergence of slow, monotone sequences from smooth, contraction mapping such as the EM algorithm. It can be used to accelerate any smooth, linearly convergent acceleration scheme.

alabama: Constrained nonlinear optimization.
Augmented Lagrangian Adaptive Barrier Minimization Algorithm for optimizing smooth nonlinear objective functions with constraints. Linear or nonlinear equality and inequality constraints are allowed.

optimx: A replacement and extension of the optim() function.
Provides a replacement and extension of the optim() function to unify and streamline optimization capabilities in R for smooth, possibly box constrained functions of several or many parameters.

dfoptim: Derivative-free optimization.
Derivative-Free optimization algorithms. These algorithms do not require gradient information. More importantly, they can be used to solve non-smooth optimization problems.

BB: Solving and optimizing large-scale nonlinear systems.
Barzilai-Borwein spectral methods for solving nonlinear system of equations, and for optimizing nonlinear objective functions subject to simple constraints.

features: Feature extraction for discretely-sampled functional data.

eigeninv: Generates (dense) matrices that have a given set of eigenvalues. Solves the ''inverse eigenvalue problem'' which is to generate a real-valued matrix that has the specified real eigenvalue spectrum. Algorithm can also generate stochastic and doubly stochastic matrices.

turboEM: A suite of convergence acceleration schemes for EM and MM algorithms.
Algorithms for accelerating the convergence of slow, monotone sequences from smooth contraction mapping such as the EM and MM algorithms. It can also be used to accelerate any smooth, linearly convergent acceleration scheme.

anoint: Analysis of interactions
Tools for assessing multiple treatment-covariate interactions with data from a parallel-group randomized controlled clinical trial.

DSBayes: Bayesian subgroup analysis in clinical trials.
Calculates posterior modes and credible intervals of parameters of the Dixon-Simon model for subgroup analysis with binary covariates in clinical trials.

crrstep: Stepwise covariate selection for the Fine & Gray competing risks regression model.
Performs forward and backwards stepwise regression for the proportional subdistribution hazards model in competing risks (Fine & Gray 1999). Procedure uses AIC, BIC and BICcr as selection criteria.

Johns Hopkins Universitydesign elementJohns Hopkins Bloomberg School of Public Healthdesign elementJohns Hopkins Medicinedesign elementJohns Hopkins School of Nursing

©2015, Johns Hopkins University. All rights reserved.
Web policies, 615 N. Wolfe Street, Baltimore, MD 21205