Abstract
Physical activity has long been shown to be associated with biological and physiological performance and risk of diseases. It is of great interest to assess whether the effect of an exposure or intervention on an outcome is mediated through physical activity measured by modern wearable devices such as actigraphy. However, existing methods for mediation analysis focus almost exclusively on mediation variable that is in the Euclidean space, which cannot be applied directly to physical activity measured by wearable devices. Such data is best summarized in the form of a random distribution. In this paper, we develop a structural equation model to the setting where a random distribution is treated as the mediator. We provide sufficient conditions for identifying the average causal effects of a distribution mediator and present methods for estimating the direct and mediating effects of a random distribution mediator on the outcome. We apply our method to analyze the iCOMPARE data set that compares flexible duty-hour policies and standard duty-hour policies on interns’ sleep related outcomes and to investigate the mediation effect of physical activity on the causal path between flexible duty-hour policies and sleep related outcomes.
Similar content being viewed by others
References
Gomez-Pinilla F, Hillman C (2013) The influence of exercise on cognitive abilities. Compr Physiol 3(1):403–428
Hillman C, Erickson K, Kramer A (2008) Be smart, exercise your heart: exercise effects on brain and cognition. Nat Rev Neurosci 9:58–65
Ten Have TR, Joffe MM, Lynch KG, Brown GK, Maisto SA, Beck AT (2007) Causal mediation analyses with rank preserving models. Biometrics 63(3):926–934
Albert JM (2008) Mediation analysis via potential outcomes models. Stat Med 27(8):1282–1304
Jo B (2008) Causal inference in randomized experiments with mediational processes. Psychol Methods 13(4):314
Sobel ME (2008) Identification of causal parameters in randomized studies with mediating variables. J Educ Behav Stat 33(2):230–251
Preacher KJ, Hayes AF (2008) Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behav Res Methods 40(3):879–891
VanderWeele T, Vansteelandt S (2014) Mediation analysis with multiple mediators. Epidemiol Methods 2(1):95–115
Sohn MB, Li H (2019) Compositional mediation analysis for microbiome studies. Ann Appl Stat 13(1):661–681
Lindquist MA (2012) Functional causal mediation analysis with an application to brain connectivity. J Am Stat Assoc 107(500):1297–1309
Zeng S, Rosenbaum S, Alberts SC, Archie EA, Li F (2021) Causal mediation analysis for sparse and irregular longitudinal data. Ann Appl Stat 15(2):747–767
Basner M, Asch DA, Shea JA, Bellini LM, Carlin M, Ecker AJ, Malone SK, Desai SV, Sternberg AL, Tonascia J (2019) Sleep and alertness in a duty-hour flexibility trial in internal medicine. N Engl J Med 380(10):915–923
Basner M, Mollicone D, Dinges DF (2011) Validity and sensitivity of a brief psychomotor vigilance test (PVT-B) to total and partial sleep deprivation. Acta Astronaut 69(11–12):949–959
Varma VR, Dey D, Leroux A, Di J, Urbanek J, **ao L, Zipunnikov V (2018) Total volume of physical activity: TAC, TLAC or TAC(λ). Prev Med 106:233–235
Faurholt-Jepsen M, Brage S, Vinberg M, Christensen EM, Knorr U, Jensen HM, Kessing LV (2012) Differences in psychomotor activity in patients suffering from unipolar and bipolar affective disorder in the remitted or mild/moderate depressive state. J Affect Disord 141(2–3):457–463
De Crescenzo F, Economou A, Sharpley AL, Gormez A, Quested DJ (2017) Actigraphic features of bipolar disorder: a systematic review and meta-analysis. Sleep Med Rev 33:58–69
Murray G, Gottlieb J, Hidalgo MP, Etain B, Ritter P, Skene DJ, Garbazza C, Bullock B, Merikangas K, Zipunnikov V, Shou H, Gonzalez R, Scott J, Geoffroy PA, Frey BN (2020) Measuring circadian function in bipolar disorders: empirical and conceptual review of physiological, actigraphic, and self-report approaches. Bipolar Disord 22:693–710
Keadle SK, Shiroma EJ, Freedson PS, Lee I-M (2014) Impact of accelerometer data processing decisions on the sample size, wear time and physical activity level of a large cohort study. BMC Public Health 14:1210
Schrack JA, Cooper R, Koster A, Shiroma EJ, Murabito JM, Rejeski WJ, Ferrucci L, Harris TB (2016) Assessing daily physical activity in older adults: unraveling the complexity of monitors, measures, and methods. J Gerontol Ser A 71(8):1039–1048
Delicado P (2007) Functional \(k\)-sample problem when data are density functions. Comput Stat 22(3):391–410
Petersen A, Müller H-G (2016) Functional data analysis for density functions by transformation to a Hilbert space. Ann Stat 44(1):183–218
Talská R, Menafoglio A, Machalová J, Hron K, Fišerová E (2018) Compositional regression with functional response. Comput Stat Data Anal 123:66–85
Petersen A, Müller H-G (2019) Fréchet regression for random objects with Euclidean predictors. Ann Stat 47:691–719
Lin Z, Kong D, Wang L (2023) Causal inference on distribution functions. J R Stat Soc Ser B 85(2):378–398
Wang J-L, Chiou J-M, Müller H-G (2016) Functional data analysis. Annu Rev Stat Appl 3:257–295
Han K, Müller H-G, Park BU (2020) Additive functional regression for densities as responses. J Am Stat Assoc 115(530):997–1010
Müller H-G, Stadtmüller U (2005) Generalized functional linear models. Ann Stat 33(2):774–805
Goldsmith J, Bobb J, Crainiceanu CM, Caffo B, Reich D (2011) Penalized functional regression. J Comput Gr Stat 20:830–851
Goldsmith J, Zipunnikov V, Schrack J (2015) Generalized multilevel function-on-scalar regression and principal component analysis. Biometrics 71:344–353
Shrout PE, Bolger N (2002) Mediation in experimental and nonexperimental studies: new procedures and recommendations. Psychol Methods 7(4):422
Chomistek AK, Yuan C, Matthews CE, Troiano RP, Bowles HR, Rood J, Barnett JB, Willett WC, Rimm EB, Bassett DR Jr (2017) Physical activity assessment with the ActiGraph GT3X and doubly labeled water. Med Sci Sports Exerc 49(9):1935
Basner M, Dinges DF (2011) Maximizing sensitivity of the psychomotor vigilance test (PVT) to sleep loss. Sleep 34(5):581–591
Li G, Shi P, Li G (1995) Global convergence rates of b-spline m-estimators in nonparametric regression. Stat Sin 5:303–318
Funding
The authors gratefully acknowledge supports from the NIH grants GM129781 and GM123056. The iCompare trial was supported by grants U01HL125388 and U01HL126088 from the National Heart, Lung, and Blood Institute and grants from the American Council for Graduate Medical Education.
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
1.1 Sensitivity Analysis in Sect. 3.3
Given \(\rho (t)\), we can estimate \(\beta _\rho ^{(s)}(t)\) as follows.
Step 1. Estimate \(\delta ^{(s)}(t,X_i)\), \(\alpha ^{(s)}(t,X_i)\), and \(\epsilon _i(t)\) in (5) by using the method in Sect. 3.1. Then \(\Sigma (u,v)=E(\epsilon _i(u)\epsilon _i(v))\) could be estimated.
Step 2. Given \(\beta _\rho ^{(s)}(t)\), let
Regress \({\tilde{Y}}_i\{Z_i,\varvec{Q}_i(Z_i)\}\) on \((X_i^T,Z_iX_i^T)\) and obtain the estimate of \(U_i={\tilde{Y}}_i\{Z_i,\varvec{Q}_i(Z_i)\}-X_i^T\xi ^{(s)}-Z_iX_i^T\gamma ^{(s)}\).
Step 3. Obtain \(\eta _i=U_i-\int \beta _\rho ^{(s)}(t)\epsilon _i(t)dt\) and \(\sigma _1^2=var(\eta _i)\).
Step 4. Estimate \(\beta _\rho ^{(s)}(t)\) by plugging \(cov\{U_i,\epsilon _i(t)\}\), \(\Sigma (u,v)\), \(\sigma _1\), and \(\rho (t)\) into (12).
Step 5. Repeat Steps 2–4 until convergence.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Zhang, J., Basner, M., Jones, C.W. et al. Mediation Analysis with Random Distribution as Mediator with an Application to iCOMPARE Trial. Stat Biosci 16, 107–128 (2024). https://doi.org/10.1007/s12561-023-09383-9
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12561-023-09383-9