Potentially eligible participants were identified through phone calls and emails in response to flyers posted throughout the community and advertisements on Craigslist. The study took place between January 2009 and December 2010. The University of Michigan Medical School Institutional Review Board approved the study protocol and all procedures.
Men and women aged 18 to 65 years of age who met DSM-IV  criteria for primary insomnia ≥ 6 months were eligible for screening. We limited the age of the sample to those 18 to 65 because the incidence and causes of insomnia are significantly different in both the younger and older populations, thus requiring unique studies to investigate the effect of chamomile in these populations. We further operationalized insomnia as 7-day sleep diary evidence of a total sleep time (TST) < 6.5 hours and sleep efficiency (SE) < 85% (total sleep time/time in bed*100) on 3 or more nights.
Exclusion criteria included unstable chronic medical conditions, e.g., chronic heart failure, asthma, cancer etc.; current diagnosis of a mood or anxiety disorder, e.g., panic disorder, anxiety, obsessive-compulsive disorder; lifetime history of bipolar or any psychotic disorder; any eating or substance use disorder; evidence of another sleep disorder, such as obstructive sleep apnea or restless legs syndrome; women who were pregnant, lactating, or less than six months post-partum; known allergy or sensitivity to chamomile or members of the ragweed family; and currently taking cyclosporine, warfarin, or any hypnotic medication.
Interested individuals were scheduled for a screening visit. After providing written informed consent, participants underwent a physical exam and medical history, and provided a list of concomitant medications to rule out medical-and substance-related causes of insomnia. Participants also completed a 4-item screener for restless legs syndrome, the Berlin Questionnaire  to rule out individuals at high-risk for sleep apnea syndrome, and the Primary Care Evaluation of Mental Disorders, Patient Health Questionnaire (PRIME-MD) a 26-item self-administered questionnaire, which screens for five of the most common psychiatric disorders in primary care: depression, anxiety, alcohol abuse/dependence, somatoform, and eating disorders . Women provided a urine pregnancy test. Physical exam and history were performed by nurse practitioners. However, the study's investigators, a physician and psychiatrist, for study eligibility, evaluated all data obtained from the screening visits. Participants with acceptable screening exams were invited to participate in the study.
Participants who continued to be eligible after sleep diary screening were randomly assigned to receive, M. recutita extract, Chamomile High Grade Extract, (MediHerb, [Warwick, Australia]), 270 mg orally twice daily, or a matching placebo. Participants were instructed to take their study capsules between lunch and dinnertime and the second capsule around one hour before bed. Patients were seen at the study clinic 28 days after the baseline visit for their day 28 visit. At the baseline and day 28 visits participants completed their FSS, Beck Depression Inventory (BDI), State Trait Anxiety Inventory, Trait Subscale (STAI-T), Insomnia Severity Index (ISI), and the Pittsburgh Sleep Quality Index (PSQI) .
Objectives and Outcomes
Our primary objective was to compare the effects of four weeks of Chamomile High Grade Extract to placebo on sleep diary measures. Participants completed sleep diaries each morning for one week prior to starting the experimental agent and during the last week of the study (days 21 to 28). The sleep diary was a daily patient log that records bedtime, rise time, sleep onset latency; number and duration of nighttime awakenings, and sleep quality. Total sleep time and sleep efficiency (total sleep time/time in bed*100) were the primary dependent variables derived from the sleep diary, although sleep quality, sleep onset latency, sleep efficiency, number of nighttime awakenings, and wake after sleep onset were also analyzed.
The secondary objectives included: 1) evaluation of common daytime consequences of insomnia, including fatigue and 2) safety and tolerability as assessed by reports at patient visits and weekly phone or email contacts during the 28-day study period. Toxicities were graded based on National Cancer Institute Common Toxicity Criteria version 3.0 for Adverse Events .
Eligible participants were randomly assigned to receive a chamomile extract, manufactured by MediHerb (Warwick, Australia), 540 mg (6 tablets daily), or a matching placebo (6 tablets daily). The dose was chosen based on the manufacturer's recommendations and on doses used with chamomile extract that produced a sedative effect in rats . Each capsule contained 90 mg dry extract of chamomile flowering tops [6:1 (v/v) extraction solvent (ethanol 70%/30% water): flowering tops] standardized up to 2.5 mg of (-)-α-bisabolol and ≥ 2.5 mg of apigenin per tablet. The University of Michigan Investigational Drug Service (UM IDS) was responsible for dispensing all study medication. Based on high performance liquid chromatography (HPLC) and gas chromatogram mass spectroscopy (GC-MS) analysis conducted one year after the start of the study, 90 mg of chamomile extract contained 3.9 mg of apigenin and 1.8 mg of (-)-α-bisabolol (Integrated Biomolecule; Tuscon, AZ). Participants were told to take the study medication twice per day with water and to bring all unused tablets to the final (28 day) study visit.
Randomization, Blinding and Allocation
Eligible participants were randomized equally to either placebo or chamomile groups. The randomization code was computer-generated by the study biostatistician. The randomization list was then given to the research pharmacist who was not associated with the study. The research pharmacist dispensed the study medication, which was in packs provided by the manufacturer, and enclosed it in numbered boxes per the randomization scheme. All study participants as well as all study personnel who assessed outcomes, worked with study data, or administered tests or questionnaires were unaware of the randomization list or treatment assignment. At the final study visits participants were also asked to indicate if they thought they had received placebo or chamomile.
Statistical Methods and Sample Size
Baseline characteristics are reported, stratified by treatment group, using means and SDs for continuous variables, and counts and percentages for categorical variables. Balance between treatment groups on baseline characteristics was tested using independent samples t-tests for continuous variables and Fisher exact tests for categorical variables.
The effects of study drug versus placebo were evaluated with regression models where the dependent variables were the endpoint of interest at day 28, adjusting for treatment group and baseline value of the variable of interest. For categorical secondary variables, Fisher exact tests were first performed. Between groups effect sizes, reported as Cohen's d, were calculated using mean change between baseline and day 28 by group and the pooled standard deviation of the mean change of each treatment group. Analyses were conducted according to the intention-to-treat principle. Data were entered into and analysed with SPSS, Windows version 18 (SPSS, Chicago, ILL). For all analyses, two-sided tests and a significance level of 0.05 were used. No adjustments were made for multiple hypotheses testing as the secondary outcomes were viewed as hypothesis generating.
Since the effect of chamomile on sleep measures had not been studied previously, no such information was available to guide sample size considerations. Thus, our goal was to determine the effect size and variability of chamomile compared to placebo to use for future sample size calculations.