Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Mar 1.
Published in final edited form as: Clin Biomech (Bristol). 2022 Feb 4;93:105593. doi: 10.1016/j.clinbiomech.2022.105593

Sample entropy discriminates balance performance of older cannabis users from non-users

Craig D Workman a,*, Jacob J Sosnoff b, Thorsten Rudroff a,c
PMCID: PMC8960352  NIHMSID: NIHMS1779290  PMID: 35151108

Abstract

Background:

Maintaining an upright stance involves a complex interaction of sensory processing and motor outputs to adequately perform this fundamental motor skill. Aging and cannabis use independently disrupt balance performance, but our recent data did not find differences in static balance performance between older cannabis Users and older Non-Users using traditional linear measures (i.e., characteristics of the center of pressure sway). The purpose of this analysis was to determine whether an unbiased entropy measure (sample entropy) can differentiate postural control (standing posture) strategies between older cannabis users and non-users when typical linear measures could not.

Methods:

Eight medical cannabis Users and eight age- and sex-matched controls completed static posturography testing in an eyes-open condition for 60 s. Linear measures included pathlength of the anterior-posterior and medio-lateral directions and an ellipse that encapsulates 95% of the 2D area explored. The nonlinear measure was the sample entropy of the center of pressure time-series in anterior-posterior and medio-lateral directions. Group comparisons were accomplished via pairwise testing and effect size calculations.

Findings:

The statistical testing revealed that sample entropy in the anterior-posterior direction was significantly larger in the Users (mean ± SD = 0.29 ± 0.08) compared to the Non-Users (0.19 ± 0.05; P = 0.01, d = 1.55).

Interpretation:

This finding indicates that the Users had a decreased regularity of their center of pressure signal in the anterior-posterior direction, which might reflect reduced balance adaptability and accompanies the increased fall risk observed in our recent report on these same subjects.

Keywords: cannabis, nonlinear, SampEn, balance, older adults

1. Introduction

The ability to maintain an upright stance is contingent upon a complex interaction of sensory processing and motor outputs (Rasman et al., 2018), and standing balance represents a fundamental motor skill required for many activities of daily living (Mlinac and Feng, 2016). Alterations in sensory processing and/or motor output often result in a decline in postural control. Indeed, balance is frequently impaired in older adults and age-related declines in postural control are associated with adverse events (e.g., falls) (Osoba et al., 2019). Postural control is traditionally assessed with linear posturography measures of the body’s center of pressure (CoP), such as path length and sway area, which quantify the amount or magnitude of postural sway (Stevens and Tomlinson, 1971). During static postural assessments, increases in sway variables are usually interpreted as declines in balance capabilities.

Although these parameters are sensitive to changes in postural control with aging (Visser et al., 2008), researchers have suggested that they do not fully characterize age-related impairments (Ghofrani et al., 2017), nor do they account for the various control processes operating across distinct timescales. In contrast, nonlinear entropy estimations, like approximate entropy (ApEn) or sample entropy (SampEn), are capable of quantifying the temporal aspects of the interacting control processes characteristic of postural stability (Borg and Laxaback, 2010). ApEn (m, r, N) provides an estimation of the regularity of a time-series by assessing the probability that a series of length N and dimension m is similar to a time-series of length N and dimension m + 1 within a tolerance window (estimated by r times the SD of the time-series) (Pincus et al., 1991). Similarly, SampEn (m, r, N) was developed as a non-biased representation of ApEn, and is less dependent on a sufficiently large N and results in more consistent outcomes (Richman and Moorman, 2000). Both statistics produce a unitless number that ranges from 0 to 2, with 0 representing perfect regularity (e.g., a pure sine wave) and 2 equivalent to random/Gaussian noise (Pincus et al., 1991; Richman and Moorman, 2000). In general, these entropy measures are increased in older adults (Borg and Laxaback, 2010; Montesinos et al., 2018) and age-related increases have been interpreted as a lack of adaptability and loss of complexity in the motor control of older adults (Goldberger et al., 2002; Yamagata et al., 2017). Thus, balance assessments that include nonlinear measures might bolster dysfunctional postural control determinations in this population.

Aging is also associated with increased chances of acquiring diseases or disabilities, many of which might be amenable to medical cannabis (e.g., pain, cachexia, nausea, and conditions associated with cancer) (Kaskie et al., 2017). Thus, given the increasing availability and use of cannabis by older adults in the US (Blazer and Wu, 2009; Han et al., 2017; Kaskie et al., 2017), researchers have started to examine the effect of chronic cannabis use on motor function in this group. For example, we recently reported that older cannabis users had an increased fall risk, worse unipedal balance performance, and slower gait speed compared to their non-cannabis using peers (Workman et al., 2021). However, there were no group differences in postural control, as indexed by CoP path length or sway area measures. This null result was surprising given that linear sway variables were reported to be impaired in chronic cannabis users (Bolbecker et al., 2018). Thus, we postulated that nonlinear measures might be able to distinguish between the static posturography balance performances of our two groups of older adults.

Therefore, the purpose of this analysis was to determine whether an unbiased entropy measure (SampEn) can differentiate postural control (standing posture) strategies between older cannabis users and non-users when typical linear CoP measures could not (Workman et al., 2021). It was hypothesized that older cannabis users would have higher SampEn values – indicative of reduced balance adaptability (Goldberger et al., 2002; Yamagata et al., 2017) – compared to their age- and sex-matched non-using peers.

2. Methods

2.1. Subjects

Eight medical cannabis users (Users) and eight age- and sex-matched cannabis abstinent controls (Non-Users) were recruited from the community according to the following criteria (see Table 1 for subject demographics): between 50 – 80 years old, part of the Iowa Medical Cannabidiol program and using cannabis for ≥ 6 months (Users) or cannabis abstinent for ≥ 5 years (Non-Users), able to comprehend the study protocol (i.e., responded to questions about the study after reading the consent form), able to complete a questionnaire in English, not pregnant, without a history of traumatic brain injury, and without a history of other drug use or alcoholism. This study was performed per the Declaration of Helsinki, approved by the University of Iowa Institutional Review Board, and all subjects read and signed the consent form before participating.

Table 1.

Subject demographic information stratified by user status. Data are mean ± SD.

Demographic Non-Users Users P-value
Sex (M/F) 3/5 3/5 n/a
Age (years) 60.5 ± 4.7 59.6 ± 4.8 0.72
Height (cm) 168.9 ± 10.6 169.5 ± 11.4 0.91
Weight (kg) 82.5 ± 20.7 92.8 ± 24.9 0.38
Duration of Cannabis Use (years) n/a 10.40 ± 12.6 n/a
Uses per week (days) n/a 4.9 ± 2.5 n/a
Uses per day (times) n/a 1.4 ± 0.7 n/a
THC Dominant (n) n/a 4 n/a
THC = CBD (n) n/a 2 n/a
CBD Dominant (n) n/a 1 n/a
Multiple Types (n) n/a 1 n/a
Medical reasons for use (n) Pain (5)* Pain (7), PD (1) n/a
Open in a new tab
*

Represents the number of subjects that have at least one of the approved conditions for medical cannabis use in Iowa but are not using cannabis as a treatment.

Group comparisons were performed with an unpaired t-test. THC = Δ-9-Tetrahydrocannabinol, CBD = cannabidiol, PD = Parkinson’s disease.

2.2. Protocol

The static posturography data for this preliminary analysis were part of another study that assessed fall risk differences between cannabis users and non-users (Workman et al., 2021); only the information relevant for the present analysis will be discussed here. Testing was completed in a single data collection session. A urine test to detect the presence of cannabis (iScreen IS1THC dipstick; Alere Toxicology, MA) was performed to verify group assignment. A single trial of static posturography was undertaken on a BTrackS balance board (Balance Tracking Systems, San Diego, CA, USA). The subjects were instructed to stand as still as possible for 60 s with their arms folded and their eyes open and looking at a marker placed at eye-level ~0.9 m in front of them. Subjects were also asked to refrain from speaking and heavy breathing/sighing throughout the trial (Dault et al., 2003). Posturography data (i.e., position of the center of pressure (CoP) in the anterior-posterior and medio-lateral directions) were collected at 25 Hz.

2.3. Measurements

The primary linear outcomes were CoP path length in the anterior-posterior (AP-Path) and medio-lateral (ML-Path) directions, and the area of an ellipse that encapsulated 95% of the 2D area explored (CoParea). These measures were calculated and recorded by the BTrackS software. For the non-linear analysis (SampEn), the CoP data were exported from the BTrackS software, imported into MATLAB (The MathWorks, Natick, MA), and analyzed with custom scripts. SampEn (m, r, N) was performed separately on the unfiltered (Giovanini et al., 2017; Rhea et al., 2015) AP and ML CoP time-series. The choice of the SampEn input parameters m, r, and N critically influence the regularity outcome and likely depend on the physiological source of the time-series (e.g., a cardiac/respiratory series (Pincus et al., 1991; Richman and Moorman, 2000) vs. spatiotemporal gait measures (Yentes et al., 2013)). Importantly, Montesinos et al. (2018) systematically evaluated how changing m, r, and N inputs in a SampEn calculation influenced the regularity estimations of CoP time-series data of healthy young adults, older fallers, and older non-fallers. Their analysis revealed that the choice of input parameters significantly influenced the entropy estimation such that larger m and r resulted in smaller regularity estimations (Montesinos et al., 2018). The authors also confirmed that the behavior of SampEn was more consistent than ApEn and that SampEn within ranges of m = 4 or 5 and r = 0.25 – 0.5 (by 0.05 increments) was the only measure (linear or non-linear) that correctly distinguished static balance performances between all three study groups. Therefore, the SampEn input parameters for the present analysis were N = 1500 (25 Hz X 60 s = 1500), m = 4, and r = 0.25; the smallest m and r within the ranges recommended by Montesinos et al. (2018) were chosen to avoid the possibility of floor effects (i.e., larger m and r consistently yielded smaller SampEn values (Montesinos et al., 2018)).

2.4. Statistical Analysis

The data were analyzed for normality by inspecting Q-Q Plots and Kolmogorov-Smirnov and Shapiro-Wilk tests. As reported previously (Workman et al., 2021), CoP-area did not pass the normality tests and non-parametric (Mann-Whitney U) testing was performed. For the remaining outcomes, unpaired t-tests between the Users and the Non-Users (e.g., AP-Users vs. AP-Non-Users), accompanied by Cohen’s d as an effect size, were performed. Group matching was confirmed by subjecting the demographic variables to unpaired t-tests. Significance was accepted at P < 0.05 and the statistical analysis was performed in GraphPad Prism 9.0 (GraphPad Software, San Diego, CA, USA).

3. Results

The subject groups were equivalent in age (P = 0.72), height (P = 0.91), and weight (P = 0.38). The results of the linear analyses have been reported elsewhere (Workman et al., 2021) and will not be detailed here. The file for one of the Users corrupted at export and was not able to be included in the non-linear analysis. To maintain an equivalent n, this subject’s Non-User match was also removed from the statistical analyses of the non-linear outcomes. The statistical testing revealed that SampEn in the AP direction was significantly larger in the Users (mean ± SD = 0.29 ± 0.08) compared to the Non-Users (0.19 ± 0.05; P = 0.01, d = 1.55; Fig. 1), indicating that the Users had a decreased regularity of their CoP signal in the AP direction. There were no differences between the groups for SampEn ML (Users = 0.25 ± 0.09, Non-Users = 0.27 ± 0.13, P = 0.77, d = 0.17) or for any of the linear CoP measures (P = 0.28 – 0.47 (Workman et al., 2021)).

Fig. 1.

Fig. 1.

Open in a new tab

Sample Entropy (SampEn) for the cannabis Users and Non-Users in anterior-posterior (AP) and medio-lateral (ML) directions. A higher SampEn indicates less regularity of the center of pressure (CoP) time-series. * = significantly different (P = 0.01).

4. Discussion

The purpose of this analysis was to determine whether non-linear metrics such as SampEn could differentiate postural control (standing posture) performances between older cannabis users and non-users when typical linear CoP measures did not (Workman et al., 2021). The novel result was that SampEn of the AP signal was significantly larger in the Users compared with the Non-Users, indicating decreased regularity. These data provide support for our hypothesis that older cannabis users would have reduced balance adaptability (i.e., higher SampEn values (Goldberger et al., 2002; Yamagata et al., 2017)), compared to their age- and sex-matched non-using peers. Interestingly, the current finding accompanies the increased fall risk observed in these same subjects (Workman et al., 2021).

Postural control in static, eyes-open stance includes reactive movements of the ankle and hip joints, as was convincingly evidenced by Winter et al. (1996) who found that AP and ML CoP control were independently managed by the ankle and hip joints, respectively. However, the subjects in their study were young adults (Winter et al., 1996) and balance control in older adults might have an increased contribution from the hip joints (Kasahara and Saito, 2021), especially in perturbation conditions (Nashner and Mccollum, 1985). This heightened role of the hips in balance maintenance with age might stem from insufficient postural control of the ankle joints (Kasahara and Saito, 2021; Sturnieks et al., 2008) and suggests that excessive use of the hip joints in older adults (i.e., for both AP and ML balance maintenance) might worsen overall balance management. Indeed, increases in ML postural control variables have been associated with falls in older adults (Piirtola and Era, 2006) and similar maladaptations have been documented in samples with impaired postural control (e.g., multiple sclerosis; (Morrison et al., 2016; Sosnoff et al., 2010)).

Based on the above evidence, it might have been expected that alterations in ML postural control would have been evident. However, the linear balance data from the current dataset (Workman et al., 2021) did not reveal any differences in either AP or ML directions. In fact, SampEn of the AP time-series was the only CoP measure, linear or non-linear, that was able to discriminate between the groups (see Fig. 1 and (Workman et al., 2021)). This finding agrees with several other reports. Specifically, Borg and Laxaback (2010) found that AP SampEn was larger in older compared with younger adults. However, they also reported that SampEn in AP was generally larger than in ML (Borg and Laxaback, 2010), which opposes the present findings (Fig. 1). Additionally, another study (Montesinos et al., 2018) found that linear CoP measures were unable to differentiate older fallers from older non-fallers, but that the fallers had significantly larger AP SampEn values than non-fallers. Another study (Pantall et al., 2018) used triaxial data from an accelerometer placed on the lower back and found significantly increased SampEn in AP, ML, and vertical directions in Parkinson’s disease subjects compared to neurologically healthy older adults. On the contrary, Decker et al. (2015) reported smaller AP SampEn in older lower physical functioning women compared with higher physical functioning women – however, both groups had relatively large SampEn values (mean ≥ 0.96) that were calculated on the CoP velocity series rather than the CoP position series (Decker et al., 2015). Additionally, AP and ML regularity were also found to be differently impacted by the various task demands of the sensory organization test and history of mild traumatic brain injury (i.e., concussion) in young adults (Sosnoff et al., 2011). Altogether, the former studies indicate that decreased regularity (larger SampEn) might be reflective of poorer balance performance via reduced balance adaptability (Goldberger et al., 2002; Yamagata et al., 2017). However, the findings of the latter studies suggest caution in using definitive statements about higher or lower signal regularity equating to respectively worse and better balance performance.

Linear measures have previously been reported to distinguish balance performance in cannabis users (Bolbecker et al., 2018) but were inadequate for the present analysis. However, there some important study differences to consider: 1) Normal aging is associated with deteriorations in balance performance which independently leads to an increased likelihood of falling (Osoba et al., 2019). This means that postural control in the Non-Users might already be suboptimal and more challenging balance conditions, like unipedal stance (Workman et al., 2021), may be necessary to adequately discriminate between these groups using linear measures. 2) The subjects in the present analysis were older adults while those in the Bolbecker et al. (2018) study were young adults (early 20’s). Building on the supposition above, postural differences between younger cannabis users and non-users, whose balance capabilities are unaffected by age, might be easier to describe with linear CoP measures. For example, younger subjects maintain static balance with a smaller CoParea (e.g., 8.3 cm2 vs. 12.7 cm2 (older) and 14.3 cm2 (older fallers) (Montesinos et al., 2018)) and would, in theory, have greater distances to explore before reaching stability margins. By extension, young subjects would have a greater potential for large differences in linear balance measures than older subjects and nonlinear measures might be required for disambiguation when group differences are theoretically smaller (i.e., in older adults). 3) Another potential confounding difference is the user status, type of cannabis used (including route of administration), and reasons for use by the subjects. In addition to recruiting young subjects, Bolbecker et al. (2018) also recruited subjects that used combustible cannabis (i.e., “joints,” which tend to be THC-dominant and are primarily used for intoxication) and were only required to be consistently using for ≥ 1 month. This contrasts sharply with the current sample that were using ingestible preparations (i.e., capsules or tinctures) with various THC:CBD ratios for medical purposes (e.g., pain) and were only included after ≥ 6 months of use (Table 1). These are relevant differences because ingested cannabis, even THC-dominant products, have a lower THC bioavailability for intoxication (Huestis, 2007), many medical cannabis users – especially older adults –titrate their use or use non-intoxicating formulations to avoid acute effects (Kaufmann et al., 2020), and chronicity of use and age have both been associated with increased side-effect tolerance (Colizzi and Bhattacharyya, 2018; Mueller et al., 2021).

There are some relevant limitations to this analysis. The primary limitation is the small sample size, which tempers the generalizability of the results, and future studies should aim to include more subjects to confirm or refute these findings. Additionally, the results are based on a single posturography trial and average responses from two or three trials might have yielded different results. Additionally, the Users all had underlying conditions (e.g., pain) that motivated their cannabis use, and these conditions might have independently influenced the results; still, most of the Non-Users had similar conditions (Table 1), which weakens the impact of this limitation. Similarly, the various THC:CBD ratios of the products consumed by the Users might have created heterogeneity within the data and an investigation of if/how these different subgroups contributed to the larger AP SampEn values was impractical in this sample. Lastly, biological sex might influence both the type of cannabis used (Kalaba and Ware, 2021) and the subject’s response (Blanton et al., 2021; Fattore et al., 2020). Future studies should institute longitudinal investigations, seek to discriminate the influence of THC:CBD ratio on functional outcomes, and recruit carefully matched (age, sex, condition/disease) control subjects. Investigations that prospectively follow medical users from initiation through chronic use, assessing both symptomatic relief, functional/cognitive side-effects, and changes in brain activity (Colizzi and Bhattacharyya, 2020), would be especially interesting and more directly inform the risk-to-benefit profile of older cannabis user. In addition, such investigations would benefit from including non-linear measures of multiple balance trials to bolster their findings and potentially reveal relevant functional differences.

5. Conclusion

In summary, typically used linear measures of balance were unable to distinguish between Users and Non-Users during static balance with the eyes open and only SampEn of the AP time-series revealed significantly decreased predictability (larger SampEn) balance control in the Users. This decreased regularity might reflect reduced balance adaptability and accompanies the increased fall risk previously observed in these subjects (Workman et al., 2021). Future longitudinal and mechanistic (e.g., neuroimaging) studies investigating the risk-to-benefit profile of medical cannabis use are strongly suggested and would be strengthened by including nonlinear measures. Such investigations will help clinicians, caregivers, and current/prospective users make more informed decisions regarding the suitability of including cannabis as a medical adjuvant.

  • Cannabis use in older adults was previously associated with increased fall risk

  • Linear measures did not discriminate static balance between users and non-users

  • The nonlinear measure sample entropy distinguished group balance differences

Acknowledgments

The authors thank the study participants for the time and effort to complete the testing. In addition, we thank Justin Deters, Hannah Welper, Annie Cooper, Delaney McDowell, and Nicole Majerus for their assistance with data collection, and Patrick Ten Eyck for his assistance with the a priori statistical design.

Funding

This work was supported by the National Institutes of Health [grant number AG0643308-01].

Data statement

The data that supports this article will be made available by contacting the senior author, Thorsten Rudroff, at thorsten-rudroff@uiowa.edu.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of interest statement

The authors declare no conflicts of interest.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper

6. References

  1. Blanton HL, Barnes RC, McHann MC, Bilbrey JA, Wilkerson JL, Guindon J, 2021. Sex differences and the endocannabinoid system in pain. Pharmacol. Biochem. Behav 202, 173107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Blazer DG, Wu LT, 2009. The epidemiology of substance use and disorders among middle aged and elderly community adults: national survey on drug use and health. Am. J. Geriatr. Psychiatry 17, 237–245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bolbecker AR, Apthorp D, Martin AS, Tahayori B, Moravec L, Gomez KL, O’Donnell BF, Newman SD, Hetrick WP, 2018. Disturbances of postural sway components in cannabis users. Drug Alcohol Depend. 190, 54–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Borg FG, Laxaback G, 2010. Entropy of balance--some recent results. J. Neuroeng. Rehabil 7, 38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Colizzi M, Bhattacharyya S, 2018. Cannabis use and the development of tolerance: a systematic review of human evidence. Neurosci. Biobehav. Rev 93, 1–25. [DOI] [PubMed] [Google Scholar]
  6. Colizzi M, Bhattacharyya S, 2020. Cannabis: Neuropsychiatry and Its Effects on Brain and Behavior. Brain Sci. 10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Dault MC, Yardley L, Frank JS, 2003. Does articulation contribute to modifications of postural control during dual-task paradigms? Cogn. Brain Res 16, 434–440. [DOI] [PubMed] [Google Scholar]
  8. Decker LM, Ramdani S, Tallon G, Jaussent A, Picot MC, Bernard PL, Blain H, 2015. Physical function decline and degradation of postural sway dynamics in asymptomatic sedentary postmenopausal women. J. Nutr. Health Aging 19, 348–355. [DOI] [PubMed] [Google Scholar]
  9. Fattore L, Marti M, Mostallino R, Castelli MP, 2020. Sex and Gender Differences in the Effects of Novel Psychoactive Substances. Brain Sci. 10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ghofrani M, Olyaei G, Talebian S, Bagheri H, Malmir K, 2017. Test-retest reliability of linear and nonlinear measures of postural stability during visual deprivation in healthy subjects. J. Phys. Ther. Sci 29, 1766–1771. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Giovanini LHF, Silva SM, Manffra EF, Nievola JC, 2017. Sampling and Digital Filtering Effects When Recognizing Postural Control with Statistical Tools and the Decision Tree Classifier. International Conference on Computational Science (ICCS 2017) 108, 129–138. [Google Scholar]
  12. Goldberger AL, Peng CK, Lipsitz LA, 2002. What is physiologic complexity and how does it change with aging and disease? Neurobiol. Aging 23, 23–26. [DOI] [PubMed] [Google Scholar]
  13. Han BH, Sherman S, Mauro PM, Martins SS, Rotenberg J, Palamar JJ, 2017. Demographic trends among older cannabis users in the United States, 2006-13. Addiction 112, 516–525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Huestis MA, 2007. Human cannabinoid pharmacokinetics. Chem. Biodivers 4, 1770–1804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kalaba M, Ware MA, 2021. Cannabinoid Profiles in Medical Cannabis Users: Effects of Age, Gender, Symptoms, and Duration of Use. Cannabis Cannabinoid Res. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kasahara S, Saito H, 2021. Mechanisms of postural control in older adults based on surface electromyography data. Hum. Movement. Sci 78, 102803. [DOI] [PubMed] [Google Scholar]
  17. Kaskie B, Ayyagari P, Milavetz G, Shane D, Arora K, 2017. The Increasing Use of Cannabis Among Older Americans: A Public Health Crisis or Viable Policy Alternative? Gerontologist 57, 1166–1172. [DOI] [PubMed] [Google Scholar]
  18. Kaufmann CN, Kim A, Miyoshi M, Han BH, 2020. Patterns of Medical Cannabis Use Among Older Adults from a Cannabis Dispensary in New York State. Cannabis Cannabinoid Res. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Mlinac ME, Feng MC, 2016. Assessment of Activities of Daily Living, Self-Care, and Independence. Arch. Clin. Neuropsychol 31, 506–516. [DOI] [PubMed] [Google Scholar]
  20. Montesinos L, Castaldo R, Pecchia L, 2018. On the use of approximate entropy and sample entropy with centre of pressure time-series. J. Neuroeng. Rehabil 15, 116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Morrison S, Rynders CA, Sosnoff JJ, 2016. Deficits in medio-lateral balance control and the implications for falls in individuals with multiple sclerosis. Gait Posture 49, 148–154. [DOI] [PubMed] [Google Scholar]
  22. Mueller RL, Ellingson JM, Bidwell LC, Bryan AD, Hutchison KE, 2021. Are the Acute Effects of THC Different in Aging Adults? Brain Sci. 11, 590. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Nashner LM, Mccollum G, 1985. The Organization of Human Postural Movements - a Formal Basis and Experimental Synthesis. Behav. Brain Sci 8, 135–150. [Google Scholar]
  24. Osoba MY, Rao AK, Agrawal SK, Lalwani AK, 2019. Balance and gait in the elderly: A contemporary review. Laryngoscope Investig. Otolaryngol 4, 143–153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Pantall A, Del Din S, Rochester L, 2018. Longitudinal changes over thirty-six months in postural control dynamics and cognitive function in people with Parkinson’s disease. Gait Posture 62, 468–474. [DOI] [PubMed] [Google Scholar]
  26. Piirtola M, Era P, 2006. Force platform measurements as predictors of falls among older people - a review. Gerontology 52, 1–16. [DOI] [PubMed] [Google Scholar]
  27. Pincus SM, Gladstone IM, Ehrenkranz RA, 1991. A regularity statistic for medical data analysis. J. Clin. Monit 7, 335–345. [DOI] [PubMed] [Google Scholar]
  28. Rasman BG, Forbes PA, Tisserand R, Blouin JS, 2018. Sensorimotor Manipulations of the Balance Control Loop-Beyond Imposed External Perturbations. Front. Neurol 9, 899. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Rhea CK, Kiefer AW, Wright WG, Raisbeck LD, Haran FJ, 2015. Interpretation of postural control may change due to data processing techniques. Gait Posture 41, 731–735. [DOI] [PubMed] [Google Scholar]
  30. Richman JS, Moorman JR, 2000. Physiological time-series analysis using approximate entropy and sample entropy. Am. J. Physiol. Heart Circ. Physiol 278, H2039–2049. [DOI] [PubMed] [Google Scholar]
  31. Sosnoff JJ, Broglio SP, Shin S, Ferrara MS, 2011. Previous mild traumatic brain injury and postural-control dynamics. J. Athl. Train 46, 85–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Sosnoff JJ, Shin S, Motl RW, 2010. Multiple sclerosis and postural control: the role of spasticity. Arch. Phys. Med. Rehabil 91, 93–99. [DOI] [PubMed] [Google Scholar]
  33. Stevens DL, Tomlinson GE, 1971. Measurement of human postural sway. Proc. R. Soc. Med 64, 653–655. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Sturnieks DL, St George R, Lord SR, 2008. Balance disorders in the elderly. Clin. Neurophysiol 38, 467–478. [DOI] [PubMed] [Google Scholar]
  35. Visser JE, Carpenter MG, van der Kooij H, Bloem BR, 2008. The clinical utility of posturography. Clin. Neurophysiol 119, 2424–2436. [DOI] [PubMed] [Google Scholar]
  36. Winter DA, Prince F, Frank JS, Powell C, Zabjek KF, 1996. Unified theory regarding A/P and M/L balance in quiet stance. J. Neurophysiol 75, 2334–2343. [DOI] [PubMed] [Google Scholar]
  37. Workman CD, Fietsam AC, Sosnoff J, Rudroff T, 2021. Increased Likelihood of Falling in Older Cannabis Users vs. Non-Users. Brain Sci 11, 134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Yamagata M, Ikezoe T, Kamiya M, Masaki M, Ichihashi N, 2017. Correlation between movement complexity during static standing and balance function in institutionalized older adults. Clin. Interv. Aging 12, 499–503. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Yentes JM, Hunt N, Schmid KK, Kaipust JP, McGrath D, Stergiou N, 2013. The appropriate use of approximate entropy and sample entropy with short data sets. Ann. Biomed. Eng 41, 349–365. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES