By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Please be advised that we experienced an unexpected issue that occurred on Saturday and Sunday January 20th and 21st that caused the site to be down for an extended period of time and affected the ability of users to access content on Wiley Online Library. This issue has now been fully resolved. We apologize for any inconvenience this may have caused and are working to ensure that we can alert you immediately of any unplanned periods of downtime or disruption in the future.
In many European countries as well as in the USA, the leaflet, or even the packaging of indomethacin, contains a specific warning to refrain from activities requiring mental alertness and motor coordination, such as driving a car. In this placebo-controlled randomized study with a crossover design we attempted to find evidence for that warning.
Indomethacin 75 mg slow release or a visually identical placebo with similar flavour was taken orally twice daily for 2.5 days. It was suggested that indomethacin affects the motor coordination required to avoid obstacles successfully during walking and that this effect will be even stronger when simultaneously performing a cognitive task that puts mental alertness to the test. Nineteen healthy middle-aged individuals (60 ± 4.7 years, eight female) performed an obstacle avoidance task on a treadmill), combined with a cognitive secondary task. Biceps femoris (BF) muscle response times, obstacle avoidance failure rates and composite scores ((100 × accuracy)/verbal response time) were used to evaluate the data.
No differences between indomethacin and placebo were found on the outcome measures regarding motor coordination, avoidance failure rates (P = 0.81) and BF response times (P = 0.47), nor on the performance on the secondary cognitive task (P = 0.12).
Even though surrogate methods were used, the current study provides evidence to suggest that there might be no need to caution patients who experience CNS side effects after indomethacin use to avoid activities requiring quick and adequate reactions, such as walking under challenging circumstances and maybe also driving a car.
In many European countries as well as in the USA, the leaflet or even the packaging of indomethacin contains a specific warning to refrain from activities requiring mental alertness and motor coordination, such as driving a car. In this placebo-controlled randomized study with a crossover design we attempted to find evidence for the above mentioned warning.
What This Study Adds
The present study showed that tasks which demand maximum attention remain unaffected by a high dose of indomethacin despite the frequently reported CNS side effects such as dizziness, drowsiness and headaches. Hence, the current study provides evidence to suggest that there might be no need to caution patients who experience CNS side effects after indomethacin use to avoid activities requiring quick and adequate reactions, such as walking under challenging circumstances and maybe also driving a car.
The prevalence of pain in the elderly is high. It is estimated that over 50% of the elderly living at home suffer from pain, a percentage which increases to 80% in long term care patients [1, 2]. When paracetamol is insufficient to reduce the pain, non-steroidal anti-inflammatory drugs (NSAIDs) are often prescribed as analgesics. As a result of that, it is estimated that over 10% of all elderly have used a NSAID in the past 24 h .
Although NSAIDs have proven to be effective in several inflammatory diseases, their use is often accompanied with various adverse effects such as gastrointestinal, respiratory, renal and central nervous system (CNS) problems. More specifically, these CNS problems are observed in about 5% of patients using NSAIDs . Most commonly reported problems are dizziness, headaches, drowsiness, mood alteration and confusion . The extent of CNS side effects differs between the available NSAIDs with a relatively high prevalence with indomethacin . Therefore, in many European countries as well as in the USA, Competent Authorities and/or pharmaceutical companies put a specific warning in the leaflet or even on the packaging of indomethacin. This warning implies that if a user experiences CNS effects that affect physical or mental capabilities, he/she should be discouraged from engaging in activities such as driving a car or operating machinery. These activities require mental alertness and motor coordination for quick and adequate responses. However, the scientific basis for this warning remains unclear as published studies on CNS effects of indomethacin on such complex activities are ambiguous.
Tests measuring postural sway during quiet standing tasks can easily be used and serve as a moderately sensitive indicator of the CNS effects for a variety of drugs . Postural balance is an important function of the CNS and is often assessed using force plates to measure vertical ground reaction forces during quiet standing. Increases in postural sway during quiet standing are usually interpreted as impaired balance control and have been shown to be a reliable predictor for those at risk of falling . In the latest systematic review on NSAIDs and falls it was shown that an increased risk for accidental falls is probable when elderly individuals are exposed to NSAIDs . However, a recent randomized clinical trial demonstrated that a high dose of indomethacin did not affect postural sway during quiet standing in healthy elderly individuals, despite the fact that 40% of the participants reported CNS side effects .
As a more complex activity, walking is considered to require more mental alertness and motor coordination than ‘simply’ keeping balance during quiet standing. Particularly, walking becomes even more complex when circumstances, such as uneven or unknown terrain or having a conversation, make it more challenging. Hence, it might be more useful to assess complex walking abilities to test for CNS side effects of NSAIDs related to falls. Obstacle avoidance during walking, for instance, is a task known to be sensitive enough to detect differences between different age groups or fallers and non-fallers [10, 11]. In addition, previous research in the elderly showed that divided attention  or even low alcohol consumption  significantly increased the risk of hitting an unexpected obstacle during walking due to delayed and weaker avoidance responses. As CNS effects are a well-known consequence of alcohol consumption we expected that a high dose of indomethacin could hamper obstacle avoidance skills in a similar way.
In this placebo-controlled randomized study in healthy seniors we assessed the effect of a high dose of indomethacin on mental alertness and motor coordination to find evidence for the label warning of indomethacin affecting these skills. Therefore, the present study investigated (i) whether indomethacin affects motor coordination when obstacles have to be avoided during walking, (ii) whether this effect, if present, would be enhanced by the secondary cognitive task and (iii) if a high dose of indomethacin impairs the performance of the cognitive task.
A crossover design (Figure 1) was selected for this double-blind, randomized, placebo-controlled study. Intake of indomethacin or placebo for the first measurement was randomly assigned by a pharmacist and was kept concealed until all measurements and analyses were finished. Indomethacin 75 mg slow release or a visually identical placebo with similar flavour was taken orally twice daily. The choice for a slow release form limited the number of capsules the participants had to take, while still reaching the desired concentration in the blood. The participants started their intake in the morning 2 days prior to the measurement day, taking the final capsule on the morning of the measurements, in total five capsules in 2.5 days. To prevent gastro-intestinal problems, all participants were given esomeprazole 20 mg to be taken simultaneously with the capsule in the morning. The participants returned their empty capsule containers on the measurement day. The measurements were planned with a 1 week interval between each to wash out the medication completely.
A group of 21 healthy senior individuals (mean age 60 ± 4.8 years, 12 male) was enrolled in this study. We included volunteers aged between 50–70 years. Participants were excluded if they suffered from serious neurological, orthopaedic or cognitive impairment, hearing problems, had poor knowledge of Dutch or used medication that affected the locomotor system or interfered with indomethacin. All participants provided written informed consent in accordance with the Declaration of Helsinki. This study (Clinical Trials Registration Number: NCT00462111) was approved by the regional ethics committee of Arnhem and Nijmegen, and the Dutch Competent Authority.
Seven participants used prescribed medication besides the study medication and were all adapted to it (Table 1). A pharmacist decided that simultaneous use of indomethacin and these medications would have no consequences for either the user or the experiment.
Table 1. Medication used by participants in the study
n, number of participants using medication from this medication class.
Proton pump inhibitors
Blood glucose lowering drugs, excl. insulin
Metformin (A10BA02), tolbutamide (A10BB03)
Platelet aggregation inhibitors, excl. heparin
Acetylsalicylic acid (B01AC06)
β-adrenoceptor blocking agents, non-selective
β-adrenoceptor blocking agents, selective
Selective calcium channel blockers with mainly vascular effects
ACE inhibitors, plain
Angiotensin II antagonists, plain
Lipid modifying agents, plain
Simvastatin (C10AA01), atorvastatin (C10AA05)
Adrenergics and other drugs for obstructive airway diseases
Obstacle avoidance task
The participants were subjected to an obstacle avoidance task on a treadmill (walking velocity 3 km h−1, Figure 2A), wearing their own comfortable low-heeled shoes. A wooden obstacle (measuring 40 × 30 × 1.5 cm) with an embedded piece of iron was held by an electromagnet just above the treadmill surface and it was released by a trigger from the computer. The obstacle was only released when a regular walking pattern (= a maximum difference of 50 ms between two consecutive steps) was observed and until at least five normal strides for the trial had been completed. The obstacle was suddenly dropped on the treadmill in front of the left foot leaving only 150–550 ms to react. Stepping to the side was discouraged and any contact of the left foot with the obstacle was defined as a failure. Avoidance failure rates were assessed using visual inspection. Surface electromyography (EMG) data were collected from the biceps femoris muscle (BF, a prime mover involved in the avoidance reaction ) to assess avoidance response times . BF response times were determined as the time between obstacle release and the moment at which BF activity exceeded the average control stride activity by at least 2 SDs for more than 30 ms (for example, see Figure 2). See Hegeman et al.  for more details concerning the obstacle avoidance task.
The participants listened to the words ‘high’ or ‘low’ in Dutch, presented in either a high or low pitch. They were instructed to indicate verbally which tone was presented as quickly as possible. The stimulus was congruent when the word and pitch matched and incongruent when they did not. Hence, incongruent stimuli were most difficult. Both stimulus and response were recorded synchronously with the data of the obstacle avoidance task.
Both the obstacle avoidance task and the cognitive task were practised before the experiment started. Five obstacle avoidance practice trials were performed, and the auditory Stroop task was practised until the subject felt comfortable with it while in a standing position and during walking on the treadmill. Each measurement consisted of four series of 15 obstacle avoidance trials each. All participants performed the first and fourth series of the obstacle avoidance task without the cognitive task (single task condition). The second and third series were performed in conjunction with the auditory Stroop task (dual task condition). Hence, each condition (single and dual task) comprised 30 trials. The participants were instructed not to prioritize any of the tasks, but to try their hardest to perform both tasks as well as possible.
For the obstacle avoidance task failure rates (the number of failed trials divided by the total number of trials) were calculated for each participant in both single and dual task condition of each experimental intervention. Similarly to the failure rates, average BF response times were calculated for each participant as well.
Verbal response times (VRT) on the auditory Stroop task were calculated as the time between the stimulus onset and the response. VRTs are determined by both task characteristics (congruent or incongruent stimulus) as well as the strategy used and thus speed-accuracy trade-offs have to be considered . Quick verbal responses usually increase the risk of failures, whereas slower responses often improve the accuracy. Using a composite score ((100 × accuracy)/VRT)  takes both speed and accuracy into account. Accuracy was defined as the percentage correct responses given and when no response was given this was treated as a failure. Composite scores were calculated for the stimuli presented prior to, during and just after obstacle crossing, representing the timing of the secondary task with respect to obstacle release (OR, Figure 2B). The last stimulus and its response before obstacle release, was defined as prior to obstacle crossing. If the response was given during perturbed walking, the stimulus was defined as during obstacle crossing. The first stimulus during the recovery from avoiding the obstacle, was defined as just after obstacle crossing.
We used repeated measures anova with post hoc pairwise comparisons (SPSS® 12.0.1: SPSS Inc., Chicago, Illinois, USA) to test for differences in obstacle avoidance outcomes (failure rates and BF response times) and in the composite scores on the secondary task. Intervention (indomethacin or placebo) was set as within subjects factor and sequence (start crossover with indomethacin or placebo) as between-subjects factor. For all analyses α was set at 0.05. Means for each intervention are presented with their standard errors (SEM).
To calculate the sample size needed we retrieved information from previous research . It was shown that the average difference in BF response times after consumption of two alcoholic drinks was 20 ms (SD 21 ms) in a similar population . To be able to identify a similar difference of 20 ms in the mean BF response times between the indomethacin and placebo intervention, a sample size of 14 subjects would be needed in the present study (β = 0.9, α = 0.05).
Figure 1 shows a flow diagram displaying the progress of all participants through the trial. Indomethacin was randomly allocated to 11 participants for the first measurement. One participant dropped out before the first measurement because of severe nausea due to indomethacin. Another participant, assigned to placebo, declined further participation prior to the first measurement without stating a reason. Of the remaining 19 participants (60 ± 4.7 years, 11 male), eight reported side effects when exposed to indomethacin, one of them also reported a headache when exposed to placebo. Drowsiness (n = 7), headache (n = 4), and nausea (n = 3) were the most reported adverse effects. The overall performance of the participants with side effects did not differ from those without reported side effects.
Obstacle avoidance task
The analysis revealed that obstacle avoidance failure rates were similar between indomethacin and placebo (F1,17 = 0.06, P = 0.81), both in the single and dual task condition (F1,17 = 0.01, P = 0.91). In the dual task condition a significant doubling in failure rates was revealed (increase from 3% to 6% on average, F1,17 = 12.6, P < 0.01; Figure 3A), which was similar for indomethacin and placebo (F1,17 = 0.15, P = 0.70).
The analysis also showed that in the single and dual task condition (F1,17 = 0.10, P = 0.76) BF response times were comparable between indomethacin and placebo (F1,17 = 0.55, P = 0.47). For both interventions BF response times increased by 22 ms on average in the dual task condition (F1,17 = 67.4, P < 0.001; Figure 3B).
Secondary task performance
Statistical analysis revealed that there were no differences between the composite scores of the indomethacin and placebo intervention (F1,17 = 2.73, P = 0.12; Figure 4). This similarity remained present even when the task was most difficult due to incongruent stimuli during obstacle crossing (F2,16 = 1.14, P = 0.35).
The present placebo-controlled randomized study investigated the effect of indomethacin on activities that require both motor coordination as well as mental alertness: the avoidance of suddenly appearing obstacles during walking combined with a cognitive task. The results clearly show that indomethacin had no effect on the reaction times of both the primary obstacle avoidance task and the secondary cognitive task. This finding is in line with the absence of differences in failure rates, as reaction time is an essential component of the performance. Even in the most difficult situations no differences between indomethacin and placebo were found.
To the best of our knowledge, the present study is the first to assess the effect of indomethacin on complex attention demanding tasks relevant to falls. Previous research mainly focused on psychomotor functioning [17-19]. Linnoila et al.  found that a single dose of 50 mg indomethacin slightly impaired performance in driving-related attention and coordination tests such as a choice reaction time test, two-coordination test and a divided attention test. On the other hand, Bruce-Jones et al.  and Saarialho-Kere et al.  showed that tests of coordination, reactive skills, attention and psychomotor speed remained unaffected after administration of indomethacin. Recent work confirmed these findings and demonstrated that neither postural balance, nor manual reaction time was affected by a high dose of indomethacin in healthy seniors . In addition, the present study showed that tasks which are even more attention demanding remained unaffected by indomethacin despite the frequently reported CNS side effects such as dizziness, drowsiness and headaches. In theory, tachyphylaxis could have occurred and underlie the absence of psychomotor effects after administration of indomethacin . However, this seems unlikely for the present study since indomethacin was only administered for 2.5 days and, to the best of our knowledge, in the highest dose ever studied in similar research. Pullar et al.  reported tachypylaxis after administration of lower doses (25 and 50 mg) of indomethacin in rheumatic patients during 5 days, while in our study after 2.5 days steady-state was just reached.
Next to the effects of indomethacin on psychomotor functioning, previous research assessed the effects of several NSAIDs on driving performance as well [21-23]. Considering the increased risk of involvement in automobile crashes with NSAID use [21, 24, 25], it could be possible that CNS side effects attribute to this. However, both diclofenac and bromfenac (the latter has been withdrawn due to severe side effects ) were found not to impair driving abilities [22, 23]. In addition, McGwin et al.  considered it unlikely that the crash risk could be explained by the impact of NSAIDs on cognition or psychomotor performance.
The current study did not assess specific driving skills, but used a combination of tasks known to require motor coordination and mental alertness , and to be associated to accidental falls . Similar to the knowledge that driving a car requires full concentration in order to prevent crashes, maximum concentration was required to perform both study tasks simultaneously without hitting the obstacle. Based on the present results and our previous work  we therefore suggest that indomethacin is not expected to increase fall risk during walking in a senior population. Hence, the probability of an increased fall risk in elderly individuals exposed to NSAIDs demonstrated in our recent review  is very likely due to other (methodological) factors, such as confounding by indication.
The reported side effects after indomethacin exposure might have unblinded and thus influenced the participants. However, this seems unlikely since the performance of the participants with side effects was similar to the performance of the participants without side effects on all tasks. Moreover, the BF response times were often very fast and therefore the avoidance reaction may not have been initiated consciously [14, 27]. Hence, if CNS side effects were experienced by the participants it seems very unlikely that this would have affected the reaction times.
In conclusion, a high dose of indomethacin did not affect capabilities deemed essential for safe walking in healthy senior individuals. Hence, though with surrogate methods, the current study provides evidence to suggest that there might be no need to caution patients who experience CNS side effects after indomethacin use to avoid activities requiring quick and adequate reactions, such as walking under challenging circumstances or even driving a car.