There is growing interest in studying the environmental, biologic and genetic correlates of the components of motor activity as well as the relationships between motor activity with sleep, exercise, mood, and cognitive functioning. Aggregation of the findings across these studies is complicated by the substantial differences in both the study goals, procedures and statistical methods. There is a need for greater coordination across studies in the procedures and analytic methods that can consider advances in analysis of functional data. We have devoted the past year to analyses of the findings of the actigraphy and electronic diary, and their combination. We have used multiple analytic approaches to characterize the actigraphy data including generalized estimating equations, mixed models, functional principal components analysis and function on scalar regression models. We found that those with bipolar disorder (BD) have greater cross-domain reactivity across homeostatic regulatory systems of motor activity, sleep, mood and energy (Merikangas et al, 2019). Additionally, we are also analyzing patterns of motor activity and their association with mood disorders. We found two main patterns (contrast between daytime and nighttime activity/sleep; and contrast activity during morning and afternoon) and people with bipolar 1 disorder tend to have lower activity levels during the day compared to controls (Glaus et al, in preparation). Since the formation of the mMARCH initiative, we have presented findings at scientific meetings including the International Society for BD Research in Washington, DC; the International Society for Affective Disorders in Amsterdam; the Gordon Conference on Cognition and Circadian Rhythms in Hong Kong; the Society for Biological Psychiatry in San Diego, CA; the International Conference on Ambulatory Monitoring of Physical Activity and Movement in Bethesda, MD; the American College of Neuropsychopharmacology in Hollywood, FL; the World Psychiatric Association Epidemiology and Public Health Section in New York, NY; the Johns Hopkins Annual Sleep and Circadian Research Day in Baltimore, MD; and several academic settings including Harvard Medical School, Weill-Cornell Medical School, and the University of Toronto. We have also done extensive analyses of the electronic diary data using ecological momentary assessment (EMA). A post-baccalaureate training fellow from our group (Yao Xiao, B.S.) developed a Shiny application using R language that facilitates visualization of the data including frequencies, cross tabulation of multiple variables, subject-level visuals, and preliminary analyses of the data. Our analyses of the data showed that people with BD have greater reactivity to positive events, whereas variability in mood and anxiety appears to be a trait marker of people with a history of mood disorders in general (Lamers et al, 2018). We have also employed a novel statistical technique to study the stability of emotional and attention states using fragmentation models (Johns et al, 2019). We have now developed new EMA scripts that will be used in several sites in order to increase the generalizability of the samples and the power of these analyses. We are also analyzing sleep patterns and disorders as well as physical activity and eating patterns and their associations with mood disorder-subtypes. This work has primarily been focused on data from the NIMH Family Study. However, we have also analyzed data from the Brisbane Twin Study, the Australian BMRI and the two Swiss studies (Family Study and CoLaus/PsyCoLaus). Our analytic group has conducted several webinars that instruct mMARCH site investigators on the procedures for processing, setting up data sets and conducting analyses of the GENEActiv actigraphy data. We have also convened an international team of biostatisticians and mathematicians directed by Drs. Zipunnikov and Shou that is working on algorithms for processing and analyzing the actigraphy data. The goal of this group is to simultaneously characterize multiple landmarks in circadian rhythms and will model inter- and intra-day interactions and dynamics in rest/activity and sleep patterns, thereby augmenting the available information and will empower the analytical framework uniformly applied across mMARCH sites. Recent publications resulting from the mMARCH initiative: -Merikangas KR, Swendsen J, Hickie IB, Cui L, Shou H, Merikangas AK, Zhang J, Lamers F, Crainiceanu C, Volkow ND, and Zipunnikov V. Real-time mobile monitoring of the dynamic associations among motor activity, energy, mood and sleep in adults with BD. JAMA Psychiatry. 2019. -Glaus J, et al (in preparation). Patterns of motor activity and mood disorders using two family studies. -Lamers F, Swendsen J, Cui L, Husky M, Johns J, Zipunnikov V, and Merikangas KR. Mood reactivity and affective dynamics in mood and anxiety disorders. Journal of Abnormal Psychology. 2018. -Johns JT, Di J, Merikangas K, Cui L, Swendsen J, and Zipunnikov V. Fragmentation as a novel measure of stability in normalized trajectories of mood and attention measured by EMA. Psychol Assess. 2019. Public Health Impact: The formation of the mMARCH initiative will enable groups to efficiently share and combine data to learn more about how activity affects different disorders and diseases across many populations, including mood disorders, sleep patterns, circadian rhythms, genetic studies, emotion, eating, etc. This work will also define targets for prevention and intervention studies. Future Plans: During the next year we are focusing on four major activities: (1) joint analysis of the mMARCH core group data including the CoLaus/PsyCoLaus study of comorbidity of depression and cardiovascular disease, the NESDA study in the Netherlands, the Australian studies of twin and youth with emerging mood disorders, and the Hong Kong circadian rhythms study; (2) follow up of the NIMH Family Study and the CoLaus/PsyCoLaus samples to investigate the stability of the findings from the first wave of participants; (3) initiation of new studies of youth in seven sites (miniMARCH collaboration); and (4) development of translational studies to identify the regulatory systems underlying motor activity and sleep across species. For the first goal, we will continue to develop standard methods of analyses for cross-site data, and the topical work groups on mood disorders, sleep patterns, circadian rhythms, developmental trajectories, and genetic studies will establish specific aims and analyze the multi-site data. For the second goal, we have developed new scripts for the EMA data and will examine the stability of the findings by repeated assessments twice per year. The third goal involves harmonization of clinical and mobile technology procedures in three studies of offspring of parents with BD (NIMH, Bethesda, MD; Lausanne, CH; Hong Kong, CN), three population studies (Sao Paulo, BR; Queensland, AU; Staten Island, NY) and three clinical studies of youth (Toronto, CA; Philadelphia, PA; Sydney, AU). We are starting to have regular meetings of the latter sites to discuss common aims, measures and procedures. The fourth goal will be achieved through collaborations with behavioral neuroscientists and circadian biologists including NIMH collaborators Samer Hattar, PhD, Yogita Chudasama, PhD, and Susan Harbison, PhD, along with Judy L. Cameron, PhD from the University of Pittsburgh. Additionally, we are creating a mMARCH website for both the general public and professionals, in order to share the findings and to exchange procedures, methods, analytic approaches and programs, and literature. The goal of these studies is to identify the longitudinal evolution of patterns of motor activity/sleep and their relationships with mood and comorbid conditions.
|Ramirez, Veronica; Shokri-Kojori, Ehsan; Cabrera, Elizabeth A et al. (2018) Physical activity measured with wrist and ankle accelerometers: Age, gender, and BMI effects. PLoS One 13:e0195996|