J Am Acad Audiol 25:521-528 (2014)
The Influence of Caffeine on the Sensory Organization Test DOI: 10.3766/jaaa.25.6.2 Kathleen M. McNerney* Mary Lou Coad* Robert F. Burkard*f
Abstract Background: Clinicians often request that patients refrain from consuming caffeinated beverages 24 h before vestibular function testing. However, there is limited research regarding how caffeine may affect the results of these tests. The sensory organization test (SOT) evaluates how well an individual is able to maintain his or her balance during several different conditions that manipulate vestibular, visual, or somatosensory information. Purpose: This study evaluated whether caffeine consumption affects the results of the SOT in a group of healthy young adults. Research Design: individuals were evaluated under two conditions: (1) after consuming 300 mg of caf feine before testing, and (2) without consuming a caffeinated beverage for 24 h before testing. Regular caf feine intake and caffeine withdrawal symptoms were assessed in these individuals. Participants were stratified into a no/low or a moderate/high caffeine intake group through the use of a self-reported 1-week caffeine diary. Study Sample: Thirty healthy control participants (mean age = 23.28 yr; males = 9) without any history of vestibular or balance impairment participated in the present study.
Data Collection/Analysis: The NeuroCom SMART Equitest was used to administer the SOT, whereas paired f-tests, completed with IBM SPSS Statistics 20, were used to analyze the data for statistical significance.
Results: Analysis of the data revealed a statistically significant difference between the caffeine and no caffeine sessions during (1) condition 5 (C5): eyes closed, platform sway-referenced; and (2) the total composite score. Statistically significant differences were also noted for the vestibular and somatosen sory preference ratios. In general, the participants performed better (i.e., higher equilibrium/composite scores) during the caffeine session. When significant results were found, the participants were stratified by weekly caffeine intake into a no/low caffeine (LC) intake group versus a moderate/high caffeine (HC) intake group. After this stratification, a statistically significant difference remained for C5, the composite score, and the somatosensory/vestibular preference ratios for the LC intake group, whereas no statisti cally significant results were found in the HC intake group. In addition, further analysis revealed less of a change in the equilibrium score as the amount of weekly caffeine intake increased. Despite these sig nificant results, the mean differences were small in magnitude, and C5, the composite score, as well as the sensory analysis ratios, fell within normal limits for all participants during both sessions. Conclusions: The ingestion of caffeine did not produce a clinically significant effect in healthy young control participants. Future research is needed to determine if these same results occur in older adults, or in individuals with a history of vestibular impairment.
Key Words: Sensory organization test; computerized dynamic posturography; caffeine Abbreviations: C = caffeine; HC = high caffeine; LC = low caffeine; NC = no caffeine; SOT = sensory organization test
'Department of Rehabilitation Science, University at Buffalo, The State University of New York, Buffalo, NY; fDepartment of Otolarynqoloqy University at Buffalo, The State University of New York, Buffalo, NY Kathleen M. McNerney, Department of Rehabilitation Science, University at Buffalo, The State University of New York, 511 Kimball Tower, Buffalo NY 14214; Phone: 716-829-6799; Fax 716-829-3217; E-mail: [email protected] This study was supported by a New Investigator Research Grant from the American Academy of Audiology/American Academy of Audiology Foundation Research Grants in Hearing and Balance Program. This study was presented orally at the American Balance Society Meeting, March 7, 2012, Scottsdale, AZ.
521
J o u rn a l o f th e A m erica n A cad em y o f A udiology/Volum e 25, Number 6, 2014
omputerized dynamic posturography allows one to evaluate an individual’s ability to main tain his or her balance in a variety of dynamic conditions. The sensory organization test (SOT) is com posed of six conditions (C1-C6) th at progress in diffi culty from the first condition to the sixth condition. During C3-C6, the visual surround and/or force plat form is sway-referenced (i.e., moves in response to) to the individual’s movement. During sway-referencing, information from the visual and/or proprioceptive modal ity is unreliable and is no longer useful for maintaining balance (Shepard and Janky, 2008). Table 1 provides a description of the SOT conditions. A large portion of the North American population acknowledges th at they consume caffeine on a daily basis (Hughes and Oliveto, 1997). Caffeine, particularly in excess, can induce adverse effects such as nausea, vomiting, restlessness, irritability, headaches, anxiety, increased heart rate, and difficulty sleeping (Pennington et al, 2010). Furthermore, caffeine seems to have a greater effect on individuals who do not consume it on a daily basis, according to the Mayo Clinic (2014). In addition to the adverse effects of consuming caffeine, studies have also shown that often withdrawal symp toms accompany abrupt cessation of caffeine consump tion. These symptoms can include headache, increased fatigue and irritability, difficulty concentrating, n au sea, anxiety, decreased energy and alertness, anxiety, and depressed mood (Ozsungur et al, 2009). It is not uncommon for clinicians to request that patients refrain from consuming caffeinated beverages before undergoing a vestibular evaluation (Nashner, 1997). However, there is limited published research regarding the effects of caffeine on vestibular function. Felipe et al (2005) is one of the only published studies that evaluated the effects of caffeine on calorics as well as on tests of oculomotor function. The study evaluated 19 indi viduals who were referred for vestibular testing based on a history of vestibular symptoms. They observed that ingestion of caffeine did not affect the outcome of the tests (i.e., individuals who displayed abnormal results on the tests when they were instructed not to drink coffee also displayed abnormal results when they were permitted to consume their normal daily intake of caffeine).
C
Table 1. Six Conditions of SOT C1
Stand with eyes open
C2
Stand with eyes closed Stand with eyes open; visual surround is
C3
sway-referenced Stand with eyes open; force platform is
C4
sway-referenced Stand with eyes closed; force platform is
C5
sway-referenced Stand with eyes open; visual surround and force
C6
S22
platform are sway-referenced
The purpose of the present study was to evaluate whether caffeine affects the SOT. The present study will also evaluate whether there are negative effects of caf feine withdrawal on the ability to perform the task. For instance, if an individual normally consumes a large amount of caffeine per day, does asking the individual to avoid caffeine before undergoing the SOT compromise the results of the tests (i.e., result in fatigue, inability to concentrate, anxiety, headache, or decreased ability to perform the task)? Finally, we evaluated whether caffeine had an influence on the SOT learning effect, by requiring the participants to undergo three repetitions of the SOT during their initial visit. This step also ensured that any changes that were displayed between the caffeine (C) and no-caffeine (NC) sessions could not be attributed to an SOT learning effect, which has been shown to occur after repetitive administrations of the test (Wrisley et al., 2007; Berstein and Burkard, 2009). METHODS articipants consisted of 30 healthy young adults ages 19-27 yr (mean = 23.28 yr; SD = 1.95 yr), 9 of which were males, with no history of vestibular or balance disorders, or any other significant medical conditions that could have compromised the integrity of the experiment. We permitted an unequal number of males and females to be enrolled in the study, as pre vious research has shown no significant effects of gen der on the SOT (Hu et al, 1996; Lewis et al, 2009). Participants were tested on 2 separate days (with at least 1 day in between; however, most of the sessions were completed 1 wk apart). During the NC session, the participants were asked to refrain from drinking any caffeinated beverages for 24 hr before undergoing the vestibular evaluation. During the C session, partic ipants were asked to refrain from drinking any caffei nated beverages on the day of the test and were then asked to consume approximately 16 oz of coffee contain ing ^300 mg of caffeine (www.caffeine.informer.com/ the-complete-guide-to-starbucks-caffeine; McCusker et al, 2003). The coffee was brewed in our laboratory to en sure th at all participants received approximately the same amount of caffeine. The participants were perm it ted to put additives in their coffee (e.g., cream and sugar), if they did not prefer their coffee black. The ves tibular evaluation began approximately 30 min after the participant began consuming the coffee to allow for adequate absorption of the caffeine, as peak absorp tion is usually reached within 30-45 min (Denaro and Benowitz, 1991). The C/NC order of the testing sessions was counterbalanced across participants, and the study sessions were completed well within the half-life of caf feine (i.e., 4—6 hr) (Denaro and Benowitz, 1991). Partici pants were asked to keep a caffeine diary for 7 consecutive days. If their sessions occurred 1 wk apart (or >1 wk
P
Caffeine and the SOT/McNerney et al
apart), they were simply asked to bring their diary with them to their next visit. If a participant underwent the two sessions less than 1 wk apart, they were given a caf feine diary after their second visit, and asked to return it in 7 days. Participants were required to include the num ber and type of beverage, including brand (e.g., Starbucks, Dunkin’ Donuts coffee, Pepsi, Coke, Red Bull), as well as the total number of ounces they consumed of each beverage each day. A mean for milligrams per week of caffeine consumption was then calculated for each par ticipant using the “dose by product” data reported on http://www.wilstar.com/caffeine.htm and the Mayo Clinic (2014). The study enrolled 22 individuals who were con sidered to be light caffeine users (0-200 mg per day), 6 users who were considered to be moderate caffeine users (201-399 mg per day), and 2 individuals who were consid ered to be heavy caffeine users (>400 mg per day), accord ing to the International Coffee Organization. To assess the effects of caffeine withdrawal (during the session where they had refrained from drinking caffeine for 24 hr), participants were given a caffeine withdrawal ques tionnaire that assessed the types of symptoms they were experiencing (e.g., nausea, headache, anxiety), as well as the severity of these symptoms. The questionnaire con sisted of 14 symptoms, which have been reported by Ozsungur et al (2009) to be symptoms conveyed by indi viduals who had refrained from consuming caffeinated beverages or foods for 48 hr. Participants were asked to rate whether they were experiencing these symptoms on a scale of 0-10. At the top of the survey, a score of 0 indi cated that they did not currently experience a particular symptom; 1-3, a mild form of a specific symptom; 4-7, a moderate form of a certain symptom; and 8-10, a severe form of a particular symptom. For the purposes of this study, total scores between 1 and 47 were considered to be mild caffeine withdrawal; scores between 48 and 94, moderate caffeine withdrawal; and scores between 95 and 140, severe caffeine withdrawal. During their initial visit, participants completed three full repetitions of the SOT (six conditions of the SOT, with three trials per condition), as a learning effect has been shown to be present, which plateaus after the third repe tition (Wrisley et al, 2007; Bernstein and Burkard, 2009). The results from the third SOT repetition were used for statistical comparisons. The sessions were counter balanced, so the C session occurred first in 15 partici pants, and second in the remaining 15 participants. The NeuroCom SMART EquiTest (NeuroCom; Clackamas, OR) was used to collect the data. The NeuroCom soft ware provided norms for the equilibrium scores for each of the six conditions, as well as for the composite score and the sensory analysis ratios. Statistical analyses, which included paired /-tests, were computed with use of IBM SPSS Statistics 20. A Cohen’s d for paired tests (d = [Mdifference/SD&fference]) was used to calculate effect size. An effect size of 0.20 is considered to be
a small effect size; 0.50, a medium effect size; 0.80, a large effect size; and 1.30, a very large effect size (Nolan and Heinzer, 2011; Cohen, 1988). When a statistically significant difference was found, the participants were stratified into two groups to allow us to evaluate the effects of daily caffeine consumption on the results. Mean consumption levels, which were provided by the International Coffee Organization, were used to stratify study participants into a no/low (n = 22; caffeine 200 mg/day) caffeine intake group. Written consent was obtained from each participant, and the experimental protocol was approved by the Univer sity at Buffalo Health Sciences Institutional Review Board. RESULTS C affein e D iary
Participants consumed a mean of 1135.28 ± 1005.62 mg of caffeine per week. A total of three participants reported no caffeine intake, whereas the participants with the largest caffeine intake reported consuming 4358 mg of caffeine per week (i.e., 32 cups of coffee and 1 cup of cola). In the no/low (n = 22) caffeine (LC) intake group, daily caffeine intake ranged from 0 to 183 mg per day, whereas in the moderate/high (n = 8) caffeine (HC) intake group, daily caffeine intake ranged from 231 to 623 mg per day. C affein e W ithdraw al Q u estio n n a ire
Participants were required to complete a caffeine with drawal questionnaire during th eir NC visit, which allowed us to assess the type as well as the severity of the withdrawal symptoms they were experiencing. Of a possible 140 points, participants’ scores ranged from 0 to 59. A higher score indicated more severe caffeine withdrawal symptoms. The average score across all par ticipants was 10.53 ± 13.07. Seven participants reported that they were experiencing no symptoms as a result of refraining from drinking caffeine, whereas the highest symptom withdrawal score that we recorded was 59 (which was provided by the individual who also had the largest amount of caffeine intake [4358 mg/week]). Table 2 displays the results from the caffeine withdrawal ques tionnaire. In general, 29 of the 30 participants were found to be experiencing mild caffeine withdrawal symptoms, whereas one of the participants was found to be experienc ing moderate caffeine withdrawal symptoms. S en so ry O rgan ization T est
E qu ilibriu m Scores The individual equilibrium scores indicate the postural stability during each of the six Sensory Organization Test
5S3
Journal of the American Academy of Audiology/Volume 25, Number 6, 2014
Table 2. Results of the Caffeine Withdrawal Questionnaire Score*
No. of Participants
>10 10-20 20-30 30-40 40-50 50-60
20 4 3 2 0 1
*Of a possible 140 points.
was a statistically significant difference betw een the C (mean = 80.30; SD = 4.09; n = 22) and the NC (mean = 76.98; SD = 4.97; n = 22) sessions for C5 (t(21) = 3.53; p = 0.002; Mdlff = 3.32; SDdifr = 4.41; Cohen’s d = 0.75) as well as between the C (mean = 87.64; SD = 2.17; n = 22) and NC (mean = 86.55; SD = 2.50; n = 22) sessions for the composite score (t(21) = 2.9; p = 0.009; Mdiff = 1.09; SDdiff = 1.8; Cohen’s d = 0.62), w hereas no statistically significant differences were seen between the sessions for the HC intake group.
Sensory A n alysis R atios (SOT) conditions (C1-C6), whereas the composite score is a weighted average th at combines all six SOT conditions (Shepard and Janky, 2008). Figure 1 and Table 3 display the differences in m ean equilibrium scores between the C and the NC sessions for C1-C6 of the SOT, as well as for the composite score. Although there were minimal mean differences between the sessions (the greatest differ ence being 2.79 for C5), and all of the composite scores were w ithin norm al limits, statistical comparison of the resu lts via paired L tests revealed a significant differ ence between th e C (mean = 80.04; SD = 4.48; n = 30) and NC (m ean = 77.26; SD = 4.71; n = 30) sessions for C5 of th e SOT (t(29) = 3.45; p = 0.002; Mdiff = 2.79; SDdiff = 4.43; d = 0.62); as well as between the C (mean = 87.60; SD = 2.53; n = 30) andN C sessions (mean = 86.6; SD = 2.62; n = 30) for the composite score (t(29) = 3.075; p = 0.005; Mdiff = 1.0; SDdiff = 1.78; d = 0.56). Figures 2A and B display the SOT results after the participants were stratified into LC versus the HC intake groups. There was a greater effect (although m inim al in m agnitude) of caffeine for C5, C6, and the composite scores in the individuals who drank lower am ounts of caffeine versus those who drank larger am ounts of caffeine. In the LC intake group, there
Figure 3 displays the differences in sensory analysis ratios for the C versus the NC sessions. The som atosen sory, visual, and vestibular ratios indicate how well the participant can use inform ation from his or h er soma tosensory, visual, and vestibular systems, respectively. The preference ratio indicates how m uch the partici p ant relies on visual inform ation to m aintain his or h er balance, even when the visual inform ation is not correct (Ferber-V iart et al, 2007). Despite m inim al dif ferences between the ratios (i.e., a difference of 0.01 for the somatosensory, visual, and preference analyses, and a difference of 0.03 for the vestibular analysis), and th at all of the ratios fell within norm al limits, statistical anal ysis via paired i-tests revealed a significant difference between the C and NC sessions for the vestibular pref erence ratio (t(29) = 3.70; p = 0.001; Mdirr = -0.032; SD djff = 0.047; d = 0.68) as well as for the somatosensory preference ratio (t(29) = 2.07; p = 0.048; Mdirr = -0.008; SDdiff = 0.022; d = 0.36). W hen the participants were stratified into LC versus HC intake groups, the m ean differences in the sensory ratios between the C and the NC sessions were higher in the LC intake group (somatosensory = HC = 0.006
Figure 1 . Mean ± SD (C = solid error lines; NC = dashed lines) equilibrium score across conditions from all 30 participants in the C (diamonds) versus the NC (squares) session.
5S4
Caffeine and the SOT/McNerney et al
Table 3. Mean Equilibrium Scores and Composite Scores (±SD) for the Caffeine versus No-Caffeine Sessions and the Difference between the Two Sessions Condition
Caffeine
No Caffeine
Difference
Paired f-test
C1
94.94 ± 1.53
95.28 ± 1.55
- 0 .3 3
0.320
C2
93.32 ± 1.62
92.86 ± 2.23
0.47
0.136
C3
93.86 ± 1.61
93.68 ± 2.50
0.00
1
C4
91.40 ± 3.05
90.99 ± 2.78
0.41
0.391
C5
80.04 ± 4.48
77.26 ± 4.71
2.79
0.002*
C6
81.20 ± 5.91
79.52 ± 5 .1 8
1.68
0.067
C om posite score
87.60 ± 2.53
86.60 ± 2.62
1.00
0.005*
'in d ic a te s a statistically significant difference.
versus LC = 0.009; vestibular = HC = 0.018 versus LC = 0.037; preference = HC = 0.001 versus LC = 0.013), with the exception of the visual sensory ratio (HC = 0.014 versus LC == 0.005). T h ere w as a significant difference betw een th e C a n d NC sessions for th e v estib u lar preference ratio (t(21) = 3.76; p = 0.001; Mdiff = 0.040; SD diff = 0.049; d = 0.81) as w ell as for th e som atosensory p ref erence ra tio (t(21) = 2.09; p = 0.49; Mdiff = -0.010; SD dig- = 0.022; d = 0.45) in the LC intake group, whereas
no significant differences were noted in the HC intake group. L e a rn in g E ffect An SOT learning effect has been reported in previous research (Wrisley et al, 2007; B ernstein and B urkard, 2009), which shows th a t individuals will perform better on later adm inistrations of the te st compared w ith the first adm inistration. This effect has been shown to
A
Figure 2. Mean equilibrium scores in the C versus the NC sessions for the low (n = 22) (A) versus the high (n = 8) caffeine drinkers (B). Diamonds: C session. Squares: NC session.
5S5
Journal of the American Academy of Audiology/Volume 25, Number 6, 2014
Figure 3. Differences in sensory ratios between the C and the NC sessions. A statistically significant difference is indicated by an (*).
plateau after the third repetition. The present study addressed the SOT learning effect by asking the partic ipants to complete three repetitions of the SOT during their first visit (which meant th at half of the partici pants completed three repetitions of the SOT during the C session, and the remaining 15 participants com pleted three SOT repetitions during the NC session). The present study accounted for the learning effect, so that any changes th at were seen on the SOT between sessions could not be attributed to simply learning the task. Table 4 displays the improvement in equilibrium scores from the first to the third repetition of the SOT for all six conditions as well as the composite score, for all 30 participants. Similar to previous research, the learning effect was greater for the later conditions (i.e., C3-C6) as well as the composite score, compared with the earlier conditions (i.e., C1-C2). The results from statistical analysis of the learning effect can be found in Table 5. To determine the effects of caffeine on the SOT learning effect, we divided the participants into groups based on whether they performed the first three SOT repetitions during the C (n = 15) versus the NC (n = 15) session. Paired i-tests revealed no statisti cally significant differences in the learning effect for those who underwent the C versus the NC session first, for any of the conditions. Table 4. Improvement in the Equilibrium Score from the First to the Third Test Administration Condition
improvement from Repetition 1 to Repetition 3
C1 C2 C3 C4 C5 C6 Composite score ’ Indicates a statistically significant difference.
5S 6
0.12 0.10 1.70* 2.42* 5.83* 7.22* 4.07’
D IS C U S S IO N
he caffeine withdrawal questionnaire that was used in the present study may not have been very sensi tive, given that low caffeine users sometimes reported high caffeine withdrawal scores, and vice versa. We also had one individual report that they were experiencing symptoms of caffeine withdrawal, despite that they reported they did not consume any caffeine during the prior week. Of our 30 participants, 29 reported experi encing mild caffeine withdrawal symptoms based on our questionnaire. Therefore, the effects of caffeine with drawal on the SOT could not effectively be assessed in the present experiment. Future research would need to be completed on an equal number of individuals who were experiencing mild, moderate, and severe caf feine withdrawal symptoms to draw an accurate conclu sion on the effects of caffeine withdrawal on the SOT. The participants in the present study underwent three repetitions of the SOT during their initial visit to account for the learning effect. Therefore, it would seem that the statistically significant changes for the C versus the NC sessions th at were found for C5 and the composite score of the SOT, as well as for the sen sory analysis ratios, cannot be explained as a simple learning effect on SOT performance, as such an effect had presumably been achieved during the first three baseline SOT administrations. The results showed that all of the composite scores as well as the sensory anal ysis ratios were within normal limits for all of the participants, regardless of session or caffeine intake. Interestingly, if there was a difference between the C and NC sessions, in general, participants tended to perform better in the caffeine session (i.e., they had higher (better) equilibrium scores in the C ver sus the NC sessions). W hen participants were eval uated based on regular caffeine intake, the m ean data revealed a greater difference betw een the C and NC sessions for the LC intake group compared
T
Caffeine and the SOT/McNerney et al
Table 5. Mean (±SD ) Equilibrium and Composite Scores for Session 1 versus Session 3 of the SOT Condition
Repetition 3
C3 C4 C5 C6 Composite score
93.71 91.01 78.32 80 86.93
N o te :
(2.35) (3.11) (4.58) (5.38) (2.75)
Repetition 1 92.01 88.59 72.49 72.78 82.87
(2.76) (4.59) (9.42) (13.64) (4.93)
Statistical Analysis Paired Paired Paired Paired Paired
f-test f-test f-test f-test f-test
(29) (29) (29) (29) (29)
= = = = =
-3.45; -3.65; -4.21; -2.90; -5.43;
= 0.002; Mdiff = -1.70; SDdiff = 2.70; d = 0.63 0.001; Mdiff = -2.42; SDdiff = 3.64; d = 0.66 p
The Influence of Caffeine on the Sensory Organization Test DOI: 10.3766/jaaa.25.6.2 Kathleen M. McNerney* Mary Lou Coad* Robert F. Burkard*f
Abstract Background: Clinicians often request that patients refrain from consuming caffeinated beverages 24 h before vestibular function testing. However, there is limited research regarding how caffeine may affect the results of these tests. The sensory organization test (SOT) evaluates how well an individual is able to maintain his or her balance during several different conditions that manipulate vestibular, visual, or somatosensory information. Purpose: This study evaluated whether caffeine consumption affects the results of the SOT in a group of healthy young adults. Research Design: individuals were evaluated under two conditions: (1) after consuming 300 mg of caf feine before testing, and (2) without consuming a caffeinated beverage for 24 h before testing. Regular caf feine intake and caffeine withdrawal symptoms were assessed in these individuals. Participants were stratified into a no/low or a moderate/high caffeine intake group through the use of a self-reported 1-week caffeine diary. Study Sample: Thirty healthy control participants (mean age = 23.28 yr; males = 9) without any history of vestibular or balance impairment participated in the present study.
Data Collection/Analysis: The NeuroCom SMART Equitest was used to administer the SOT, whereas paired f-tests, completed with IBM SPSS Statistics 20, were used to analyze the data for statistical significance.
Results: Analysis of the data revealed a statistically significant difference between the caffeine and no caffeine sessions during (1) condition 5 (C5): eyes closed, platform sway-referenced; and (2) the total composite score. Statistically significant differences were also noted for the vestibular and somatosen sory preference ratios. In general, the participants performed better (i.e., higher equilibrium/composite scores) during the caffeine session. When significant results were found, the participants were stratified by weekly caffeine intake into a no/low caffeine (LC) intake group versus a moderate/high caffeine (HC) intake group. After this stratification, a statistically significant difference remained for C5, the composite score, and the somatosensory/vestibular preference ratios for the LC intake group, whereas no statisti cally significant results were found in the HC intake group. In addition, further analysis revealed less of a change in the equilibrium score as the amount of weekly caffeine intake increased. Despite these sig nificant results, the mean differences were small in magnitude, and C5, the composite score, as well as the sensory analysis ratios, fell within normal limits for all participants during both sessions. Conclusions: The ingestion of caffeine did not produce a clinically significant effect in healthy young control participants. Future research is needed to determine if these same results occur in older adults, or in individuals with a history of vestibular impairment.
Key Words: Sensory organization test; computerized dynamic posturography; caffeine Abbreviations: C = caffeine; HC = high caffeine; LC = low caffeine; NC = no caffeine; SOT = sensory organization test
'Department of Rehabilitation Science, University at Buffalo, The State University of New York, Buffalo, NY; fDepartment of Otolarynqoloqy University at Buffalo, The State University of New York, Buffalo, NY Kathleen M. McNerney, Department of Rehabilitation Science, University at Buffalo, The State University of New York, 511 Kimball Tower, Buffalo NY 14214; Phone: 716-829-6799; Fax 716-829-3217; E-mail: [email protected] This study was supported by a New Investigator Research Grant from the American Academy of Audiology/American Academy of Audiology Foundation Research Grants in Hearing and Balance Program. This study was presented orally at the American Balance Society Meeting, March 7, 2012, Scottsdale, AZ.
521
J o u rn a l o f th e A m erica n A cad em y o f A udiology/Volum e 25, Number 6, 2014
omputerized dynamic posturography allows one to evaluate an individual’s ability to main tain his or her balance in a variety of dynamic conditions. The sensory organization test (SOT) is com posed of six conditions (C1-C6) th at progress in diffi culty from the first condition to the sixth condition. During C3-C6, the visual surround and/or force plat form is sway-referenced (i.e., moves in response to) to the individual’s movement. During sway-referencing, information from the visual and/or proprioceptive modal ity is unreliable and is no longer useful for maintaining balance (Shepard and Janky, 2008). Table 1 provides a description of the SOT conditions. A large portion of the North American population acknowledges th at they consume caffeine on a daily basis (Hughes and Oliveto, 1997). Caffeine, particularly in excess, can induce adverse effects such as nausea, vomiting, restlessness, irritability, headaches, anxiety, increased heart rate, and difficulty sleeping (Pennington et al, 2010). Furthermore, caffeine seems to have a greater effect on individuals who do not consume it on a daily basis, according to the Mayo Clinic (2014). In addition to the adverse effects of consuming caffeine, studies have also shown that often withdrawal symp toms accompany abrupt cessation of caffeine consump tion. These symptoms can include headache, increased fatigue and irritability, difficulty concentrating, n au sea, anxiety, decreased energy and alertness, anxiety, and depressed mood (Ozsungur et al, 2009). It is not uncommon for clinicians to request that patients refrain from consuming caffeinated beverages before undergoing a vestibular evaluation (Nashner, 1997). However, there is limited published research regarding the effects of caffeine on vestibular function. Felipe et al (2005) is one of the only published studies that evaluated the effects of caffeine on calorics as well as on tests of oculomotor function. The study evaluated 19 indi viduals who were referred for vestibular testing based on a history of vestibular symptoms. They observed that ingestion of caffeine did not affect the outcome of the tests (i.e., individuals who displayed abnormal results on the tests when they were instructed not to drink coffee also displayed abnormal results when they were permitted to consume their normal daily intake of caffeine).
C
Table 1. Six Conditions of SOT C1
Stand with eyes open
C2
Stand with eyes closed Stand with eyes open; visual surround is
C3
sway-referenced Stand with eyes open; force platform is
C4
sway-referenced Stand with eyes closed; force platform is
C5
sway-referenced Stand with eyes open; visual surround and force
C6
S22
platform are sway-referenced
The purpose of the present study was to evaluate whether caffeine affects the SOT. The present study will also evaluate whether there are negative effects of caf feine withdrawal on the ability to perform the task. For instance, if an individual normally consumes a large amount of caffeine per day, does asking the individual to avoid caffeine before undergoing the SOT compromise the results of the tests (i.e., result in fatigue, inability to concentrate, anxiety, headache, or decreased ability to perform the task)? Finally, we evaluated whether caffeine had an influence on the SOT learning effect, by requiring the participants to undergo three repetitions of the SOT during their initial visit. This step also ensured that any changes that were displayed between the caffeine (C) and no-caffeine (NC) sessions could not be attributed to an SOT learning effect, which has been shown to occur after repetitive administrations of the test (Wrisley et al., 2007; Berstein and Burkard, 2009). METHODS articipants consisted of 30 healthy young adults ages 19-27 yr (mean = 23.28 yr; SD = 1.95 yr), 9 of which were males, with no history of vestibular or balance disorders, or any other significant medical conditions that could have compromised the integrity of the experiment. We permitted an unequal number of males and females to be enrolled in the study, as pre vious research has shown no significant effects of gen der on the SOT (Hu et al, 1996; Lewis et al, 2009). Participants were tested on 2 separate days (with at least 1 day in between; however, most of the sessions were completed 1 wk apart). During the NC session, the participants were asked to refrain from drinking any caffeinated beverages for 24 hr before undergoing the vestibular evaluation. During the C session, partic ipants were asked to refrain from drinking any caffei nated beverages on the day of the test and were then asked to consume approximately 16 oz of coffee contain ing ^300 mg of caffeine (www.caffeine.informer.com/ the-complete-guide-to-starbucks-caffeine; McCusker et al, 2003). The coffee was brewed in our laboratory to en sure th at all participants received approximately the same amount of caffeine. The participants were perm it ted to put additives in their coffee (e.g., cream and sugar), if they did not prefer their coffee black. The ves tibular evaluation began approximately 30 min after the participant began consuming the coffee to allow for adequate absorption of the caffeine, as peak absorp tion is usually reached within 30-45 min (Denaro and Benowitz, 1991). The C/NC order of the testing sessions was counterbalanced across participants, and the study sessions were completed well within the half-life of caf feine (i.e., 4—6 hr) (Denaro and Benowitz, 1991). Partici pants were asked to keep a caffeine diary for 7 consecutive days. If their sessions occurred 1 wk apart (or >1 wk
P
Caffeine and the SOT/McNerney et al
apart), they were simply asked to bring their diary with them to their next visit. If a participant underwent the two sessions less than 1 wk apart, they were given a caf feine diary after their second visit, and asked to return it in 7 days. Participants were required to include the num ber and type of beverage, including brand (e.g., Starbucks, Dunkin’ Donuts coffee, Pepsi, Coke, Red Bull), as well as the total number of ounces they consumed of each beverage each day. A mean for milligrams per week of caffeine consumption was then calculated for each par ticipant using the “dose by product” data reported on http://www.wilstar.com/caffeine.htm and the Mayo Clinic (2014). The study enrolled 22 individuals who were con sidered to be light caffeine users (0-200 mg per day), 6 users who were considered to be moderate caffeine users (201-399 mg per day), and 2 individuals who were consid ered to be heavy caffeine users (>400 mg per day), accord ing to the International Coffee Organization. To assess the effects of caffeine withdrawal (during the session where they had refrained from drinking caffeine for 24 hr), participants were given a caffeine withdrawal ques tionnaire that assessed the types of symptoms they were experiencing (e.g., nausea, headache, anxiety), as well as the severity of these symptoms. The questionnaire con sisted of 14 symptoms, which have been reported by Ozsungur et al (2009) to be symptoms conveyed by indi viduals who had refrained from consuming caffeinated beverages or foods for 48 hr. Participants were asked to rate whether they were experiencing these symptoms on a scale of 0-10. At the top of the survey, a score of 0 indi cated that they did not currently experience a particular symptom; 1-3, a mild form of a specific symptom; 4-7, a moderate form of a certain symptom; and 8-10, a severe form of a particular symptom. For the purposes of this study, total scores between 1 and 47 were considered to be mild caffeine withdrawal; scores between 48 and 94, moderate caffeine withdrawal; and scores between 95 and 140, severe caffeine withdrawal. During their initial visit, participants completed three full repetitions of the SOT (six conditions of the SOT, with three trials per condition), as a learning effect has been shown to be present, which plateaus after the third repe tition (Wrisley et al, 2007; Bernstein and Burkard, 2009). The results from the third SOT repetition were used for statistical comparisons. The sessions were counter balanced, so the C session occurred first in 15 partici pants, and second in the remaining 15 participants. The NeuroCom SMART EquiTest (NeuroCom; Clackamas, OR) was used to collect the data. The NeuroCom soft ware provided norms for the equilibrium scores for each of the six conditions, as well as for the composite score and the sensory analysis ratios. Statistical analyses, which included paired /-tests, were computed with use of IBM SPSS Statistics 20. A Cohen’s d for paired tests (d = [Mdifference/SD&fference]) was used to calculate effect size. An effect size of 0.20 is considered to be
a small effect size; 0.50, a medium effect size; 0.80, a large effect size; and 1.30, a very large effect size (Nolan and Heinzer, 2011; Cohen, 1988). When a statistically significant difference was found, the participants were stratified into two groups to allow us to evaluate the effects of daily caffeine consumption on the results. Mean consumption levels, which were provided by the International Coffee Organization, were used to stratify study participants into a no/low (n = 22; caffeine 200 mg/day) caffeine intake group. Written consent was obtained from each participant, and the experimental protocol was approved by the Univer sity at Buffalo Health Sciences Institutional Review Board. RESULTS C affein e D iary
Participants consumed a mean of 1135.28 ± 1005.62 mg of caffeine per week. A total of three participants reported no caffeine intake, whereas the participants with the largest caffeine intake reported consuming 4358 mg of caffeine per week (i.e., 32 cups of coffee and 1 cup of cola). In the no/low (n = 22) caffeine (LC) intake group, daily caffeine intake ranged from 0 to 183 mg per day, whereas in the moderate/high (n = 8) caffeine (HC) intake group, daily caffeine intake ranged from 231 to 623 mg per day. C affein e W ithdraw al Q u estio n n a ire
Participants were required to complete a caffeine with drawal questionnaire during th eir NC visit, which allowed us to assess the type as well as the severity of the withdrawal symptoms they were experiencing. Of a possible 140 points, participants’ scores ranged from 0 to 59. A higher score indicated more severe caffeine withdrawal symptoms. The average score across all par ticipants was 10.53 ± 13.07. Seven participants reported that they were experiencing no symptoms as a result of refraining from drinking caffeine, whereas the highest symptom withdrawal score that we recorded was 59 (which was provided by the individual who also had the largest amount of caffeine intake [4358 mg/week]). Table 2 displays the results from the caffeine withdrawal ques tionnaire. In general, 29 of the 30 participants were found to be experiencing mild caffeine withdrawal symptoms, whereas one of the participants was found to be experienc ing moderate caffeine withdrawal symptoms. S en so ry O rgan ization T est
E qu ilibriu m Scores The individual equilibrium scores indicate the postural stability during each of the six Sensory Organization Test
5S3
Journal of the American Academy of Audiology/Volume 25, Number 6, 2014
Table 2. Results of the Caffeine Withdrawal Questionnaire Score*
No. of Participants
>10 10-20 20-30 30-40 40-50 50-60
20 4 3 2 0 1
*Of a possible 140 points.
was a statistically significant difference betw een the C (mean = 80.30; SD = 4.09; n = 22) and the NC (mean = 76.98; SD = 4.97; n = 22) sessions for C5 (t(21) = 3.53; p = 0.002; Mdlff = 3.32; SDdifr = 4.41; Cohen’s d = 0.75) as well as between the C (mean = 87.64; SD = 2.17; n = 22) and NC (mean = 86.55; SD = 2.50; n = 22) sessions for the composite score (t(21) = 2.9; p = 0.009; Mdiff = 1.09; SDdiff = 1.8; Cohen’s d = 0.62), w hereas no statistically significant differences were seen between the sessions for the HC intake group.
Sensory A n alysis R atios (SOT) conditions (C1-C6), whereas the composite score is a weighted average th at combines all six SOT conditions (Shepard and Janky, 2008). Figure 1 and Table 3 display the differences in m ean equilibrium scores between the C and the NC sessions for C1-C6 of the SOT, as well as for the composite score. Although there were minimal mean differences between the sessions (the greatest differ ence being 2.79 for C5), and all of the composite scores were w ithin norm al limits, statistical comparison of the resu lts via paired L tests revealed a significant differ ence between th e C (mean = 80.04; SD = 4.48; n = 30) and NC (m ean = 77.26; SD = 4.71; n = 30) sessions for C5 of th e SOT (t(29) = 3.45; p = 0.002; Mdiff = 2.79; SDdiff = 4.43; d = 0.62); as well as between the C (mean = 87.60; SD = 2.53; n = 30) andN C sessions (mean = 86.6; SD = 2.62; n = 30) for the composite score (t(29) = 3.075; p = 0.005; Mdiff = 1.0; SDdiff = 1.78; d = 0.56). Figures 2A and B display the SOT results after the participants were stratified into LC versus the HC intake groups. There was a greater effect (although m inim al in m agnitude) of caffeine for C5, C6, and the composite scores in the individuals who drank lower am ounts of caffeine versus those who drank larger am ounts of caffeine. In the LC intake group, there
Figure 3 displays the differences in sensory analysis ratios for the C versus the NC sessions. The som atosen sory, visual, and vestibular ratios indicate how well the participant can use inform ation from his or h er soma tosensory, visual, and vestibular systems, respectively. The preference ratio indicates how m uch the partici p ant relies on visual inform ation to m aintain his or h er balance, even when the visual inform ation is not correct (Ferber-V iart et al, 2007). Despite m inim al dif ferences between the ratios (i.e., a difference of 0.01 for the somatosensory, visual, and preference analyses, and a difference of 0.03 for the vestibular analysis), and th at all of the ratios fell within norm al limits, statistical anal ysis via paired i-tests revealed a significant difference between the C and NC sessions for the vestibular pref erence ratio (t(29) = 3.70; p = 0.001; Mdirr = -0.032; SD djff = 0.047; d = 0.68) as well as for the somatosensory preference ratio (t(29) = 2.07; p = 0.048; Mdirr = -0.008; SDdiff = 0.022; d = 0.36). W hen the participants were stratified into LC versus HC intake groups, the m ean differences in the sensory ratios between the C and the NC sessions were higher in the LC intake group (somatosensory = HC = 0.006
Figure 1 . Mean ± SD (C = solid error lines; NC = dashed lines) equilibrium score across conditions from all 30 participants in the C (diamonds) versus the NC (squares) session.
5S4
Caffeine and the SOT/McNerney et al
Table 3. Mean Equilibrium Scores and Composite Scores (±SD) for the Caffeine versus No-Caffeine Sessions and the Difference between the Two Sessions Condition
Caffeine
No Caffeine
Difference
Paired f-test
C1
94.94 ± 1.53
95.28 ± 1.55
- 0 .3 3
0.320
C2
93.32 ± 1.62
92.86 ± 2.23
0.47
0.136
C3
93.86 ± 1.61
93.68 ± 2.50
0.00
1
C4
91.40 ± 3.05
90.99 ± 2.78
0.41
0.391
C5
80.04 ± 4.48
77.26 ± 4.71
2.79
0.002*
C6
81.20 ± 5.91
79.52 ± 5 .1 8
1.68
0.067
C om posite score
87.60 ± 2.53
86.60 ± 2.62
1.00
0.005*
'in d ic a te s a statistically significant difference.
versus LC = 0.009; vestibular = HC = 0.018 versus LC = 0.037; preference = HC = 0.001 versus LC = 0.013), with the exception of the visual sensory ratio (HC = 0.014 versus LC == 0.005). T h ere w as a significant difference betw een th e C a n d NC sessions for th e v estib u lar preference ratio (t(21) = 3.76; p = 0.001; Mdiff = 0.040; SD diff = 0.049; d = 0.81) as w ell as for th e som atosensory p ref erence ra tio (t(21) = 2.09; p = 0.49; Mdiff = -0.010; SD dig- = 0.022; d = 0.45) in the LC intake group, whereas
no significant differences were noted in the HC intake group. L e a rn in g E ffect An SOT learning effect has been reported in previous research (Wrisley et al, 2007; B ernstein and B urkard, 2009), which shows th a t individuals will perform better on later adm inistrations of the te st compared w ith the first adm inistration. This effect has been shown to
A
Figure 2. Mean equilibrium scores in the C versus the NC sessions for the low (n = 22) (A) versus the high (n = 8) caffeine drinkers (B). Diamonds: C session. Squares: NC session.
5S5
Journal of the American Academy of Audiology/Volume 25, Number 6, 2014
Figure 3. Differences in sensory ratios between the C and the NC sessions. A statistically significant difference is indicated by an (*).
plateau after the third repetition. The present study addressed the SOT learning effect by asking the partic ipants to complete three repetitions of the SOT during their first visit (which meant th at half of the partici pants completed three repetitions of the SOT during the C session, and the remaining 15 participants com pleted three SOT repetitions during the NC session). The present study accounted for the learning effect, so that any changes th at were seen on the SOT between sessions could not be attributed to simply learning the task. Table 4 displays the improvement in equilibrium scores from the first to the third repetition of the SOT for all six conditions as well as the composite score, for all 30 participants. Similar to previous research, the learning effect was greater for the later conditions (i.e., C3-C6) as well as the composite score, compared with the earlier conditions (i.e., C1-C2). The results from statistical analysis of the learning effect can be found in Table 5. To determine the effects of caffeine on the SOT learning effect, we divided the participants into groups based on whether they performed the first three SOT repetitions during the C (n = 15) versus the NC (n = 15) session. Paired i-tests revealed no statisti cally significant differences in the learning effect for those who underwent the C versus the NC session first, for any of the conditions. Table 4. Improvement in the Equilibrium Score from the First to the Third Test Administration Condition
improvement from Repetition 1 to Repetition 3
C1 C2 C3 C4 C5 C6 Composite score ’ Indicates a statistically significant difference.
5S 6
0.12 0.10 1.70* 2.42* 5.83* 7.22* 4.07’
D IS C U S S IO N
he caffeine withdrawal questionnaire that was used in the present study may not have been very sensi tive, given that low caffeine users sometimes reported high caffeine withdrawal scores, and vice versa. We also had one individual report that they were experiencing symptoms of caffeine withdrawal, despite that they reported they did not consume any caffeine during the prior week. Of our 30 participants, 29 reported experi encing mild caffeine withdrawal symptoms based on our questionnaire. Therefore, the effects of caffeine with drawal on the SOT could not effectively be assessed in the present experiment. Future research would need to be completed on an equal number of individuals who were experiencing mild, moderate, and severe caf feine withdrawal symptoms to draw an accurate conclu sion on the effects of caffeine withdrawal on the SOT. The participants in the present study underwent three repetitions of the SOT during their initial visit to account for the learning effect. Therefore, it would seem that the statistically significant changes for the C versus the NC sessions th at were found for C5 and the composite score of the SOT, as well as for the sen sory analysis ratios, cannot be explained as a simple learning effect on SOT performance, as such an effect had presumably been achieved during the first three baseline SOT administrations. The results showed that all of the composite scores as well as the sensory anal ysis ratios were within normal limits for all of the participants, regardless of session or caffeine intake. Interestingly, if there was a difference between the C and NC sessions, in general, participants tended to perform better in the caffeine session (i.e., they had higher (better) equilibrium scores in the C ver sus the NC sessions). W hen participants were eval uated based on regular caffeine intake, the m ean data revealed a greater difference betw een the C and NC sessions for the LC intake group compared
T
Caffeine and the SOT/McNerney et al
Table 5. Mean (±SD ) Equilibrium and Composite Scores for Session 1 versus Session 3 of the SOT Condition
Repetition 3
C3 C4 C5 C6 Composite score
93.71 91.01 78.32 80 86.93
N o te :
(2.35) (3.11) (4.58) (5.38) (2.75)
Repetition 1 92.01 88.59 72.49 72.78 82.87
(2.76) (4.59) (9.42) (13.64) (4.93)
Statistical Analysis Paired Paired Paired Paired Paired
f-test f-test f-test f-test f-test
(29) (29) (29) (29) (29)
= = = = =
-3.45; -3.65; -4.21; -2.90; -5.43;
= 0.002; Mdiff = -1.70; SDdiff = 2.70; d = 0.63 0.001; Mdiff = -2.42; SDdiff = 3.64; d = 0.66 p
