[PRN_141101001]Reliability of sleep deprivation-associated spontaneous brain activity and behavior

Submitted by Lei.Gao on

Reliability of sleep deprivation-associated spontaneous brain activity and behavior

Lei Gao1,2†, Lijun Bai3†, Yuchen Zhang4, Xi-jian Dai2, Rana Netra1, Youjiang Min5, Fuqing Zhou2, Honghan Gong2*, Ming Zhang1* and Yijun Liu6,7
1. Department of Medical Imaging, the First Affiliated Hospital of Xi’an Jiaotong University, Xi’an 710061, China
2. Department of Radiology, the First Affiliated Hospital of Nanchang University, Nanchang 330006, Jiangxi Province, China
3.The Key Laboratory of Biomedical Information Engineering, Ministry of Education, Department of Biomedical Engineering, School of Life Science and Technology, Xi’an Jiaotong University, Xi’an 710049, China
4. Zonglian Experimental Class, Xi’an Jiaotong University, Xi’an 710049, China 
5. Acupuncture & Rehabilitation Department, Affiliated Hospital of Jiangxi University of Traditional Chinese Medicine, Nanchang 330006, Jiangxi Province, China
6. Southwest University, School of Psychology, Key Laboratory of Cognition and Personality, Chongqing, Chongqing 400715, China
7. Departments of Psychiatry and Neuroscience, McKnight Brain Institute, University of Florida, Gainesville, FL 32610, USA.
* Correspondence to Honghan Gong, Department of Radiology at the First Affiliated Hospital of Nanchang University, 17 Yongwaizheng Ave., Nanchang 330006, Jiangxi Province, China
Email:honghan_gong@sina.com (Honghan Gong);
and Ming Zhang, Department of Medical Imaging at the First Affiliated Hospital of Xi’an Jiaotong University, 277 West Yanta Road, Xi'an 710061, Shaanxi Province, China. Email: zmmri@163.com (Ming Zhang)
† Contribute equally to this paper.
Recent studies have indicated that sleep deprivation (SD) alters intrinsic low-frequency connectivity in the resting brain, mainly focusing on the default mode network (DMN) and its anticorrelated network (ACN). These networks hold key functions in segregating internally and externally directed awareness. However, far less attention has been paid to investigation of the altered amplitude of these low-frequency fluctuations (ALFF) at the whole-brain level and more importantly by what extent the sleep-deprived resting brain pattern can be reproducible and predict individual behavioral performance. The aim of this study was to characterize more clearly the influence of sleep on the whole brain level of ALFF changes and its relation with the performance of a lexical decision task in the sleep deprivation. Sixteen healthy participants underwent fMRI three times: once after a normal night of sleep in the rested wakefulness (RW) state and two following approximately 24 h of total SD separated by an interval of two weeks (SD1 and SD2). Our behavioral results showed that sleep stabilizes performance whereas two sleep deprivation even at an interval of two weeks consistently deteriorates it. Sleep deprivation attenuated the ALFF mainly in the bilateral orbitofrontal cortex (OFC), bilateral dorsolateral prefrontal cortex (DLPFC) and right inferior parietal lobule (IPL). By contrast, the enhanced ALFF emerged in the left sensorimotor cortex (SMA), visual cortex and left fusiform gyrus. Conjunction analysis of SD1 and SD2 versus the control maps and voxel-wise ICC analysis revealed that these SD induced ALFF changes showed a significantly high reliability (ICC>0.5). Particularly, the attenuation of the right IPL presents a significant negative relation with the behavior performance and can be reproducible for two SD at an interval of two weeks. Our results suggest that ALFF is a stable measure in study of SD, and the right IPL may represent a stable biomarker that responds to sleep loss.
Keywords: sleep deprivation, resting-state fMRI, ALFF, test-retest, reliability

YAN Chao-Gan

Mon, 12/01/2014 - 17:43

Gao et al. performed a resting-state fMRI study to examine the change of intrinsic activity of sleep deprivation (SD), using amplitude of low-frequency fluctuations (ALFF) as the main indicator. They found that SD impaired performance of a lexical decision task. They found ALFF of bilateral orbitofrontal cortex (OFC), bilateral dorsolateral prefrontal cortex (DLPFC) and right inferior parietal lobule (IPL) were decreased after SD, whereas ALFF in left sensorimotor cortex (SMA), visual cortex and left fusiform gyrus were increased. Furthermore, they found these changes are highly reliable across two SD sessions aparted by two weeks. In addition, the decrease of ALFF in right IPL negatively correlated with behavior performance. I think this study is well performed. There is a major concern of multiple comparison correction in analyses (see below). The paper is well-written, although with a check by a English native speaker may be helpful. Besides, I have the following concerns.

1.     When doing AlphSim correction, how did you estimate the smoothness? Used the smoothing kernel in preprocessing? This is an underestimate of the effective smoothness. An estimated smoothness should be used, (e.g., 3dFWHMx in AFNI, or DPABI Statistical Analysis).

2.     Line 50: significant negative relation -> significant negative relationship

3.     Line 52: is a stable measure -> is a reliable measure.

4.     Introduction: before introducing imaging findings about SD, the authors could talk more about the SD impact on psychological and psychiatric response.

5.     Line 88: anticorrelated networks (CCN) -> ACN?

6.     Line 132: They had a BMI (in kg/m2) of 17.5–22. Seems the range of BMI is very slim?

7.     Line 157: “a short period of practice with a different set of sentences was provided.” Sentences or words?

8.     Line 165. Please describe how long (or how many volumes) of the REST scan.

9.     For ALFF calculation, several nuisance regressors can be removed. See Yan et al., 2013, Neuroimage. (Yan, C.G., Cheung, B., Kelly, C., Colcombe, S., Craddock, R.C., Di Martino, A., Li, Q., Zuo, X.N., Castellanos, F.X., Milham, M.P., 2013. A comprehensive assessment of regional variation in the impact of head micromovements on functional connectomics. Neuroimage 76, 183-201.)

10. Line 208. When cite for intraclass correlation coefficients (ICC), Shrout and Fleiss, 1979 should be cited. (Shrout, P.E., Fleiss, J.L., 1979. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 86, 420-428.)

11. Line 225: “mean ALFF values of peak voxels”. Peak voxels mean all the voxels within the cluster? Furthermore, what value did you use to do correlation? The Changes in ALFF (SD1 – control)?

12. Line 244: “ICC”. Based on what values you calculated ICC? SD1- control and SD2 – control? This procedure can be expanded in the discussion, as ICC on changes can be an important point.

13. Discussion. The authors can focus more on the overlapping finds, as those are more “reproducible” results.

Hope my comments be helpful for you to improve the manuscript.



Chao-Gan YAN, Ph.D.
Research Scientist
The Nathan Kline Institute for Psychiatric Research, Orangeburg, NY 10962
Research Assistant Professor
Department of Child and Adolescent Psychiatry / NYU Langone Medical Center Child Study Center, New York University, New York, NY 10016
Have reviewed for Journal of Neuroscience; Cerebral Cortex; NeuroImage; NeuroImage: Clinical; Human Brain Mapping; PLoS ONE; Brain Connectivity; Journal of Neurophysiology; Neuroinformatics; Journal of Neuroscience Methods; Neuroscience Letters; Behavioral and Brain Functions; Frontiers in Neuroscience; Frontiers in Human Neuroscience; and Cognitive, Affective, and Behavioral Neuroscience.