AcuSleep in Mild Traumatic Brain Injury (TBI) (AcuTBI)

The proposed research will examine innovative and effective interventions to improve veterans' sleep and quality of life (QoL), while exploring the potential of integrating complementary alternative treatment modality, such as acupuncture, when compared to established sleep education model. The findings will also allow better understanding of the complex dimensions of sleep disturbance and its association with other co-existing symptoms (e.g. pain, depression, anxiety and post-traumatic stress [PTS]) using mild traumatic brain injury (TBI) as a model and exploration of the mechanism(s) that underlie sleep improvements.

Objectives:

In the mild traumatic brain injury (TBI) population, the treatment of sleep difficulties is complicated by multiple coexisting symptoms, especially pain, depression, and anxiety, and the risk of sleep medication side effects. Complementary alternative medicine (CAM), such as acupuncture, can potentially provide efficacy in treatment of sleep difficulties with fewer side effects.

The specific aim of the proposed research project is to address the following research questions:

1. Do specific sleep outcomes improve with real versus sham acupuncture treatment in mild TBI patients?
2. Are changes in sleep outcomes associated with changes in autonomic function, pain, depression, or anxiety in mild TBI patients?

Research plan:

Study design: a clinical trial evaluating treatment effects of insomnia in mild TBI patients with subjects randomly assigned to one of the two groups a) real acupuncture; b) sham acupuncture.

Hypotheses:

1. Primary Sleep outcomes, as measured by lower global scores on the Pittsburgh Sleep Quality Index [PSQI], will improve with real acupuncture versus sham acupuncture.
2. Other sleep outcomes, as measured by higher SE, total sleep time [TST], and lower wake time after sleep onset [WASO]) recorded by wrist actigraphy will improve with real acupuncture versus sham acupuncture.
3. Secondary Quality of Life outcomes, as measured by higher VR-36 Health Status Survey sub-scales, will improve with real acupuncture versus sham acupuncture.
4. Acupuncture-based improvements in primary outcomes in mild TBI patients will be associated with changes in measures of autonomic function, pain, depression and anxiety.

Methods:

Subject Population: Veterans at the Atlanta Veteran Affairs Medical Center (VAMC) TBI Clinics, age 18 to 55, with mild TBI over 3 months post injury who have sleep difficulties > 4x per week and PSQI >8. Total of 66 subjects will be targeted for recruitment with a sample size of 50.

Procedures to Be Used: The hypotheses will be tested in a 2 group Randomised Control Trial (RCT) with assignment to sham acupuncture, or real acupuncture. Subjects will receive an intensive 60-minute sleep hygiene education session with research assistant/respiratory therapist and a package containing information on sleep hygiene prior to randomization. The subjects will be asked to practice these rules for 4 weeks at home. There will be 5 weeks of intervention and 4 weeks of follow-up. Both subjective and objective sleep outcome measures along with relevant co-variates will be assessed at baseline, post-intervention and after follow-up.

Outcome measures: Well-validated and reliable outcome measures of sleep using both objective (actigraph) and subjective (PSQI) measures, pain (Brief Pain Inventory), depression (Beck Depression Inventory), anxiety (Beck Anxiety Inventory), PTSD (PTSD Check List-Military), quality of life (Veterans Rand-36 Item Health Survey) will be used. Quantitative analysis of autonomic nerve system modulation will be investigated with measures of normalized spectral heart rate variability.

Statistical analyses: Data analysis will proceed from univariate descriptive statistics through multivariate analyses. Continuous variables will be summarized using the usual sample summary statistics (overall and within each group). Categorical variables will be represented by frequencies. For all analyses, we will evaluate model assumptions before conducting any tests of hypotheses.

Baseline characteristics between the two groups will be compared using unpaired t-tests or Chi-square tests to evaluate the equivalency of the treatment groups with respect to injury type, time from injury, severity of sleep difficulties, pain, anxiety, depression, PTSD, autonomic function in addition to the demographics. If there are significant differences between the two groups on any of the variables, these variables will be included as control variables or covariates in the analyses described below. Since we believe that headaches and pain can alter an individual's sleep, we will control for headache and pain as well.

We will conduct an intent to treat analysis of the differences in outcomes measured at post treatment between the two arms using simple t-tests, performed at a 5% significance level. Using the intent to treat analysis, all subjects will be included in the analysis as randomized. We will also make comparisons within each group of the difference in outcomes measured at baseline and post treatment using paired t-tests. Changes in sleep and QoL outcomes in the treatment group that receives real acupuncture and sleep education will be compared to changes in the treatment group receives sham acupuncture and sleep education, after adjusting for significant baseline values and subjects' demographic characteristics, will be analyzed using one way analysis of covariance. Pearson's correlations will be investigated to quantify an association between changes in sleep outcome measures and change in HRV, pain, headaches, depression, anxiety, and PTSD scores. Logistic regression models will be developed to determine if certain variables can be considered predictors of an individual achieving normalized sleep. For covariates, reference cell coding will be used for categorical variables and the influence of each observation will be examined when employing the logistic regression models.

Attrition rates will be compared across treatment arms. Given the longitudinal nature of this study, parameter estimates from mixed linear models will be developed. The justification for mixed linear models is that fact that mixed linear models do not use case-wise deletion approaches to missing data. In addition, last observation carried forward method will be used to also handle missing data. The parameter estimates from these models will be compared to parameter estimates from mixed linear models; thus, providing a clear understanding of the impact of attrition and missing data on our results.

Clinical Relevance:

The research will improve our scientific understanding of sleep difficulties in mild TBI, may also promote innovative integration of conventional therapies with CAM modalities to enhance treatment effects, reduce medication side effects, advance patient quality of care, and save healthcare costs.

• Brain Injuries, Traumatic
• Sleeplessness

• Other: Acupuncture
  Treatments will be performed by the PI, who is a licensed physician acupuncturist and has practiced acupuncture independently for 5 years with about 4000 patient/visit treatment history. A total of 10 acupuncture treatments represents a reasonable approach to optimal duration of treatment. Acupoint selection will be based on a combination of standardization and individualization for best treatment effects. At least 5 standardized body and auricular acupoints will be selected. Standardized sterile, disposable acupuncture needles will be applied for 20 minutes.
• Other: Sham Acupuncture
  Treatments will be performed by the PI, who is a licensed physician acupuncturist and has practiced acupuncture independently for 5 years with about 4000 patient/visit treatment history. A total of 10 acupuncture treatments represents a reasonable approach to optimal duration of treatment. Acupoint selection will be based on a combination of standardization and individualization for best treatment effects. At least 5 standardized body and auricular acupoints will be selected. Standardized sterile, disposable acupuncture needles will be applied for 20 minutes. Each sham or real acupuncture needle will be applied through a tube as sham needles have blunt tip and telescopic shaft, the visual effect and percutaneous sensation of sham needle mimic the real needle penetration.

• Active Comparator: Arm 1
  sham acupuncture
  Intervention: Other: Sham Acupuncture
• Experimental: Arm 2
  acupuncture
  Intervention: Other: Acupuncture

Inclusion Criteria:

1. Age: 18 to 55; and,
2. Meeting the diagnosis criteria for mild traumatic brain injury as listed above at the time of injury; and,
3. At or over 3 months post injury; and,
4. With untreated sleep complaints (Pittsburg Sleep Quality Index [PSQI] > 8, sleep difficulties at least 4x per week in the past month, average sleep duration of 6 hours or less); and,
5. Has the capacity to give informed consent.
6. Agree to attend 13 clinic visits

Exclusion Criteria:

1. Same sleep complaints present prior to traumatic brain injury; or,
2. Diagnosis of obstructive sleep apnea (OSA) by prior sleep study; or,
3. With a non-daytime work-schedule; or,
4. With prior acupuncture experience for the treatment of sleep difficulties or with acupuncture treatment within 3 months; or,
5. History of bleeding diathesis or currently on anticoagulation with INR over 2.5; or,
6. Severe depression with Beck Depression Score of 29 and above; or,
7. Moderate and severe alcohol users.
8. Does not have a permanent address