Timothy C. Hain, MD Page last modified: October 18, 2015
Mal de Debarquement or "MDD" (or MdDS, or MDD) is a type of vertigo and imbalance that occurs after getting off of a boat. The usual situation is that of a middle aged woman who has gone on a cruise. We use the abbreviation "MDD" here for clarity, because it has been used in the past. It has been pointed out that "MDD" can be confused with other disorders, and some prefer to use "MdDs" or various combinations of these letters.
To our knowledge, the first reference to the syndrome was made by Erasmus Darwin, in 1796. He wrote:
"Those, who have been upon the water in a boat or ship so long, that they have acquired the necessary habits of motion upon that unstable element, at their return on land frequently think in their reveries, or between sleeping and waking, that they observe the room, they sit in, or some of its furniture, to librate like the motion of the vessel. This I have experienced myself, and have been told, that after long voyages, it is some time before these ideas entirely vanish. The same is observable in a less degree after having travelled some days in a stage coach, and particularly when we lie down in bed, and compose ourselves to sleep; in this case it is observable, that the rattling noise of the coach, as well as the undulatory motion, haunts us. " (Darwin, 1796).
A 50-year-old woman went on her first ocean cruise. She had some motion sickness on the cruise, which responded to transdermal scopolamine. Immediately after returning from the cruise and getting onto solid ground, she developed imbalance and a rocking sensation, accompanied by fatigue and difficulty concentrating. This persisted for months. SHe felt better however when she was driving. Her description was “Imagine feeling like you are on rough seas 24 hours a day, 7 days a week.”
A related case ? An avid cross country skier noted that after she would come home from a day of skiing, she developed a sensation as if she was still skiing until she went to sleep. She was always fine in the morning.
Table1: Characteristics of MDD
Median Duration (mo)
Cha et al, 2008
Hain et al, 1999
3.5 years (6 mo to 10 years)
Brown and Baloh, 1987
Total or (average)
months to years
Table 1 summarizes the available literature about MDD. It is a disorder that mainly affects women (87%) in their mid 40's. Except in California where a large number of men are affected, (the site of Dr. Cha's study), almost all individuals (85%) with MDD are women. They typically have gone on a 7-day cruise. After getting off the boat, or "debarking" (debarquement), they immediately develop a rocking sensation, as if they are still on the boat.
Here we define MDD as rocking lasting for at least a month. This leaves the land-sickness group out.
|Age of onset of patients in Chicago Dizziness and Hearing database, as of 2014.|
The graph above is drawn on a much larger and more representative sample of patients than from our study of 1999, again mainly women (93/109), it shows that the distribution mainly includes females between the ages of 30-50.
As of 2014, we had encountered roughly 109 MDD patients in our dizzy clinic database out of an "n" of about 12,000 patients with assigned diagnoses. This suggests that only roughly 1/100 dizzy patients have MDD. As about 5% of the population develops "vestibular" dizziness per year, if we assume that our clinic sees MDD as a representative proportion of all dizzy patients, a very very rough estimate as to the prevalence of MDD is 0.05% (this is the same as 0.5/1000 persons). Or to put this into real-world terms, assuming 300 million persons in the United States, a 0.5 persons/1000 people prevalance works out to a total of 150,000 persons with MDD in the US. Compared to landsickness (which happens in roughly 50% of persons who go on boats), this is an unusual condition, and far less common than other dizzy conditions, say vestibular migraine (1% of population -- 3 million) or Meniere's disease (2/1000 people, 600,000 total in US or 0.2%). Note that this estimate could easily be wrong (probably too much) by an order of magnitude.
The figure above, from our 1999 paper, illustrates "worst case" durations -- it does not reflect the usual duration of symptoms. This is because the study group from our 1999 paper were persons with very severe MDD. A prospective study would be needed to answer this question properly.
Table 2: Features Distinguishing MDD from Land-sickness (LDS)
1 or more months
2 days maximum
About 90% female
Motion-sick on boat
Relieved by driving
There are several variants of motion induced sickness, that are not MDD. Table 2 lists the features that distinguish MDD from simple land-sickness. Land-sickness (LDS) is common, and between 41% and 73% of persons disembarking from seagoing voyages experience a brief unsteadiness syndrome (Gordon, Spitzer et al. 1995; Cohen 1996; Gordon, Shupak et al. 2000). Common LDS typically persists for 2 days or less. Persons with LDS are also likely to have sea-sickness, (Gordon, Spitzer et al. 1995) while persons with MDD generally are untroubled by sea-sickness. Males and females do not appear differ significantly in the incidence, intensity, or duration of land-sickness symptoms. (Cohen 1996). LDS, confusingly, is also termed "mal de debarquement" by some. Table 1 does not include reports or data concerning subjects who symptoms that last less than 1 month, i.e. potential land-sickness, except for the work of Cha, in whom the duration of patients with "classic" MDD could not be determined due to study design(Cha, Brodsky et al. 2008).
MDD also has some similarities to motion sickness (sea sickness, mal de mer). However, MDD is again easily distinguished by the relatively short duration of motion sickness and gender distribution. Persons with MDD reliably have relief of symptoms when in motion, such as driving a car, but experience recurrence of rocking once motion has stopped (Hain, Hanna et al. 1999; Cha, Brodsky et al. 2008). In motion sickness, many persons find driving very difficult. This is also often true for persons with vestibular disorders.
MDD also overlaps with a little studied group of patients called "rockers", who develop similar symptoms to MDD, without a preceding motion exposure. (Cha 2012) Often these patients develop head or trunk rocking, which is called "titubation" (Wassmer, Davies et al. 2003). In our clinical experience, the age, gender and pattern of medication responsiveness of this group are similar to those of MDD. Although titubation is associated with cerebellar disturbance, evidence of cerebellar damage is generally not found in "rockers".
We have also encountered a few patients, or "rockers", who developed this symptom after a well defined CNS lesion, generally a small stroke. This is usually in the motion processing area of the brain.
It is the general opinion that MDD is generally not caused by an injury to the ear or brain. At this writing (2014), the predominant opinion is that MDD is a variant motion sickness. While this theory doesn't explain why MDD seems to mainly affect women in their 40's it does seem to account for observations of naval personnel who have a similar land-sickness experience.
Some dizziness experts believe that MDD is caused by a variant of migraine. We do think that migraine increases risk of MDD. We have seen many patients who developed MDD when exposed to motion around the time of their period (which is also a high risk time for migraine). This might explain why some cruises are tolerated without distress, and others not.
Other theories about the cause of MDD it that is caused by inappropriately high weighting of somatosensory input for balance (Naichem et al, 2004). As somatosensory information and vestibular are unreliable on a boat, this is a difficult explanation to follow. We think that the internal model theory explanation (which follows) is the better one at this writing (2014).
A recent conjecture is that MDD is caused by adaptation to roll while rotating. In other words, if one is rocking side-side (roll), and also rotating the head, for long periods of time, one might develop an inappropriate cross-coupling between roll and rotation (Dai et al, 2009; 2014). Our position on this idea is that it could explain brief (2 hour) symptoms after getting off of a boat, and also offers an explicit hypothesis that might be tested formally (i.e. people who do a lot of head rotation on a boat would be more likely to get this than people who sit quietly). However, overall we are dubious that this conjecture is correct. Why would women only develop inappropriate cross-coupling ? Why wouldn't people get rid of this by simply walking around with their eyes open ?
Because the condition largely occurs in females, it may also have something to do with sex hormones, such as estrogen or progesterone. In fact, we have noted a pattern that if one asks, it is often the case that the woman who develops MDD was having their period while on the boat. MDD could also be genetic, related to two copies of the 'X' chromosome perhaps combined with other susceptibility factors. The "Norwalk" virus is common on cruise ships, and perhaps this syndrome is somehow related to this virus.
It seems unlikely to be a psychological disturbance -- although it is always difficult to entirely exclude psychological problems, the male:female ratio and other aspects of this disorder would make this unlikely.
Moeller and Lempert (2007) recently suggested that MDD is due to "deafferentation" or panic. We disagree with both of these ideas.
Cha et al (2012) recently reported changes in brain connectivity in persons with MDD. This study, however, was done in a powerful MRI scanner, which can cause temporary dizziness by itself. MDD seems to be associated with changes metabolism in the brain, in circuits related to vision, vestibular processing and emotional reactions. It would be interesting to see how much of this is due to MDD and how much is secondary to being dizzy. In 2015, in a study published in PLOS-1, Cha and Chakrapani reported that " Individuals with MdDS show brain volume differences from healthy controls as well as duration of illness dependent volume changes in (a) visual-vestibular processing areas (IPL, SPL, V3, V5/MT), (b) default mode network structures (cerebellar IX, IPL, ACC), (c) salience network structures (ACC and IFG/AI) (d) somatosensory network structures (postcentral gyrus, MCC, anterior cerebellum, cerebellar lobule VIII), and (e) a structure within the central executive network (DLPFC). " 80% of the 29 subjects in this study were female, almost all were middle aged, and almost all had symptoms for more than 1 month; thus they met the conventional criteria for this MDD. The authors themselves point out that the brain volume differences might be due to compensatory processes rather than causes of symptoms. They also note that the results were presented with "uncorrected p-values and it can be argued that many of the voxels seen in the contrasts could be seen by chance". Overall, these studies do not seem to have added much to our understanding of MDD.
There are some reports of MDD following use or withdrawal from serotonergic medications. The connection here is that serotonin may inhibit glutamate, an excitatory transmitter in the vestibular nucleus (Smith and Darlington, 2010.). This idea also provides an explanation why serotonergic medications may help MDD (see treatment section).
Note that the MDD of "Mal de Debarquement" has nothing to do with the other MDD of "Manic Depressive Disorder". The support group for MDD, being sensitive to this, has suggested that the proper name of the syndrome would best be "MdDs". Our own feeling is that there are many medical disorders that share initials with other disorders (PAN is an example -- periodic alcohol nystagmus and periodic alternating nystagmus), and one should not get too anxious about the initials
A plausible mechanism for the development of MDD is that it is due to formation of an inappropriate internal predictive model. We first proposed this idea in 2007 (Hain and Helminski, 2007). Internal models are sophisticated estimators that have been used to explain such difficult observations that one cannot "tickle" oneself (see the work of Wolpert (1995) and others). Examples of internal models are very easy to find in daily life -- suppose you pick up a suitcase, expecting it to be full, but it is empty ! Internal models are sophisticated methods of reacting to events even before they happen ! (Blakemore, 1998)
On a boat, one is faced with a difficult balancing problem, with components of rotation (pitch plane rotation -- about the axis between the ears), and linear movement (surge -- front-back movement of the boat). Both are somewhat predictable as the boat is large and it's interface between it and the ocean constrains it to low frequencies of movement.
Lets take an example -- when the boat pitches (rotates) forward, there is a small amount of pushing the person backward accompanied by a tilt of the visual world as the angle between upright and the boat surface becomes more acute. To stay upright in response to pitch, a person should not activate their ankle muscles much as inertia tends to keep their body upright. Vision is accurate on the deck but inaccurate inside. Thus vision is unreliable. Although there is rotation around the ankle joint, and thus somatosensory input, there should be no "righting" response from the person because the body is upright in space. Visual responses are correct on deck and incorrect inside, and thus a "rule" cannot be made. The rule then for pitch rotation of the boat, one should ignore somatosensory information signaling rotation . Thus for pitch of the boat, a selective "downweighting" of somatosensory information, or both somatosensory and visual information according to context, would be a reasonable adaptation (or rule).
For linear acceleration of the boat under the person, or "surge" as it is called in nautical contexts, inertia attempts to keep the person still in space, but due to shear force at the feet, the person becomes destabilized and rotates at the ankles. Then vision, vestibular and somatosensation senses are activated by the bodily rotation with respect to the boat, and an active response is needed to prevent a fall. Thus for surge of the boat, no relative sensory reweighting would be needed, although increased responses to all types of input might be helpful.
How does the brain figure which rule to apply ? We propose that people develop a predictive model of the boat motion, and use their prediction to select the rule to apply for boat motion (and avoid falling).
Normally, it seems likely that over a few days, people develop an internal model of periodic motion on the boat so that they predict and cancel out input (visual or somatosensory) that is phase-locked to pitch rotation, and enhance responses due to surge that is not. The internal model normally is disposed of once the person returns to terra firma, again over a period of hours to days. Persons with MDD are unable to dispose of this internal model, which is only useful when they are exposed to periodic motion (such as when driving a car).
We have encountered a few patients whose motion after-effects are specific to motion - -when they cross country ski, they have a motion after-effect of skiing, etc. Of course, this has to be an internal model problem.
The internal model theory explains most of the features of MDD.
What does the data say about mechanism ?
With respect to the hypothesis that MDD is caused by reweighting of visual, vestibular or somatosensory input, the data so far is contradictory. Nachum and associates used posturography to study young males aged 18-22 with motion sickness and land-sickness (they considered land-sickness to be equivalent to mal de debarquement in their paper -- see table 2 above). They reported that these young men developed increased reliance on somatosensory input after motion exposure, and reduced weighting of vision and vestibular input (Nachum, Shupak et al. 2004). While the accuracy of visual input depends on whether one is inside the boat or on the deck, semicircular canal input is accurate on boats, and somatosensory input is intermittently accurate. Accordingly, it is difficult to understand a rationale for this adaptation. An intrinsic problem with this study is that the study group were young men with motion sickness and transient land-sickness, not middle aged women with the month-or-greater MDD syndrome.
Stoffregen et al (2013) also studied a different group than the usual MDD sufferers -- 40 of average age 20.68 years, oddly enough, without reporting their gender. Like Naichum et al, they defined MDD to be landsickness, and thus they were studying something other than MDD, but calling it MDD. They defined "Low-MD" as symptoms for 30 minutes or less, and "High-MD" as symptoms for 120 minutes or more. Because this papers definition of MDD and subject population is so different than the clinical population in which the medical community diagnoses MDD, they were studying landsickness but they were calling it MDD, and there is little to be gained in considering their work further. It is interesting to note that their paper was published in Plos, an open source journal. While the PLOS journals are very accessible, their review process is quite different than the traditional one that involves experts in the field.
Well, at any rate, rather than the somatosensory weighting process suggested by Naichum, a more reasonable possibility is that individuals with MDD may develop an increased reliance on visual and vestibular information (and thus decreased somatosensory weighting). This occurs in normal subjects who are exposed to situations where somatosensory feedback is distorted (Peterka, 2002), and would also be a reasonable adaptation to boat pitch. Either adaptation might result in inaccurate land sensorimotor integration. Nevertheless, neither of these adaptations explain the rocking sensation of MDD or the characteristic improvement on driving a car.
The most recent mechanistic proposition for MDD is that of Dai and associates (2014). They proposed that MDDs was caused by maladaptation of the vestibulo-ocular reflex (VOR) to roll of the head during rotation, and reported that a 5 day long protocol attempting to readapt the VOR resulted in “substantial recovery on average for approximately 1 year” of 17 of 24 subjects. While these results are encouraging, this theory does not explain why patients with MDD are better while driving. Furthermore, it is difficult to see why ordinary movement through the environment should not recalibrate the VOR over a few days – the usual upper limit for the duration of land-sickness. At the time of this writing in 2015, the roll adaptation theory and treatment protocol needs more study.
The diagnosis of MDD is made by a combination of the history (rocking after prolonged exposure to a boat or other source of prolonged motion), improvement with driving, and exclusion of reasonable alternatives. A motion exposure of 2 hours is a bare minimum. The typical duration of exposure is a week.
Tests to exclude Menieres disease should be done, and if there is a history of plane flight, perilymph fistula should also be considered. A typical patient is a woman of appropriate age (see figure above), who has gone on a cruise and who is now rocking. Recently, the criterion has been expanded to require getting better on driving. This feature is extremely common in MDD, but extremely rare in inner ear disorders or Migraine.
In my practice, I get the following tests, in persons who are not "classic" -- i.e. not improved by driving and not on a cruise, see table 1:
- Audiogram and sweep OAE (expect normal, abnormal suggests other disorders)
- Rotatory chair test. This is usually shows strong and normal vestibular responses.
After MDD has started, most medications that work for other forms of dizziness or motion sickness are ineffective. Specifically, antivert (bonine, meclizine), dramamine, and scopolamine seem to be of little use. The author has tried out many other medications, and has also not found response to more unusual agents for dizziness such as betahistine, baclofen, or verapamil. The usual treatment strategy for MDD is to attempt to make the patient comfortable, while waiting for the MDD to end by itself (typically within 6 months, see table 1). Conventional vestibular suppressants that affect anticholinergic pathways such as meclizine and transdermal scopolamine are not helpful in MDD. (Hain, Hanna et al. 1999) Benzodiazepines, such as clonazepam, are of the most benefit. (Hain, Hanna et al. 1999; Cha 2012), and SSRI type antidepressants are also suggested as being potentially helpful (Cha 2012). There are also anecdotal reports of good responses to gabapentin, amitriptyline, and venlafaxine – all medications that are also helpful in migraine.
- Low doses of clonazepam, a benzodiazepine medication related to Valium (diazepam), are helpful in most persons with MDD. There is some worry that these medications may prolong the duration of symptoms (although this worry has not yet been tested by a research study). These medications are also addictive, which is worrisome.
- We have also had recent success with treatment with venlafaxine - -this is an antidepressant that is very useful in migraine. We use the same protocol as for treatment of migraine (top dose typically only 37.5 XL).
- Occasional patients have reported improvement from treatment with Neurontin (gabapentin). This is generally in very large doses (e.g. 2400 mg, although one would think that 900 would be plenty).
- Occasionally persons with rocking due to other causes respond to one of the SSRI type antidepressants, and this may also be worth considering. Paroxetine is the most common SSRI used in persons with dizziness. We have had some success with venlafaxine, which is an SNRI/SSRI.
- Somewhat similarly to venlafaxine, as of 2014, there are a few reports of a good response to Sinemet, which the brand name of a medication (carbidopa-levodopa) that increases dopamine. As dopamine is a precursor to many neurotransmitters, including norepinephrine, increased dopamine may be working downstream to increase the levels of these neurotransmitters. It seems unlikely to us that dopamine deficiency is the cause of MDD, as there is an immense population of people with dopamine deficiency (i.e. Parkinson's patients), that hugely exceeds the number of people with MDD or rocking.
- We have been told that non-steroidal anti-inflammatory medications have helped, but this does not seem to be a general pattern. Perhaps the mechanism here is quieting down migraine.
- Also anecdotally, phenytoin and carbamazepine (or oxcarbazepine) may be useful in reducing symptoms. Phenytoin has been reported useful in motion-sickness. A controlled trial of these medications may be in order if more evidence accumulates.
Medications to stop and procedures to consider stopping.
Dai et al (2014) reported successful treatment of MDD using a procedure involving optokinetic visual stimulation and tilting of the head about the front-back axis (roll). The procedures involves multiple short sessions over a week. The study was uncontrolled. It is too soon to tell whether or not this result can be replicated. As noted above, we are dubious that roll adaptation explains MDDs, and for this reason we are also dubious about the rationale for this treatment. Still, it seems harmless and given that feedback concerning it working seems to be positive, we think it reasonable to try. We have set this up 10/ 2015 in our dizzy practice in Chicago.
Repetitive transcranial magnetic stimulation over the dorsolateral prefrontal cortex was reported by Cha to be associated with “short-term symptom improvement”, in a pilot study of Cha in 2013 as well as helpful in 5 of 10 subjects in more recent studies of Guofa et al (2015) and Pearce (2015). More study is needed of this treatment modality for MDD. In TMS, generally any changes are temporary. For example, although TMS can be used to treat depression, one needs to do it over and over again every week. A treatment protocol that accomplishes long term improvement in this disorder that lasts months to years is crucial. See our "exciting stuff" page concerning our attempt to find a provider for TMS for MDDs patients in Chicago.
Medications taken prior and during boat travel might prevent development of MDD.
Physical therapy: The evidence for a positive role for physical therapy in MDD is somewhere between nonexistent and weak (Hain and Helminski, 2007). In our original study, 10/15 persons who had vestibular rehabilitation reported improvement, but the natural history of MDD is to improve, and one wonders what would have happened had they not undergone rehabilitation. In other words, this was an uncontrolled study, which sheds no light on whether PT is helpful. Cha commented in passing that "only rare patients seem to be cured by vestibular therapy" (Cha 2012). In fact, the only peer reviewed literature describing physical therapy treatment for MDD are two case reports (Zimbelman and Watson 1992; Liphart et al, 2014). Of course, it is not known how these cases would have done without PT. In other words, these were not controlled. In general, while many individuals with MDD undergo vestibular rehabilitation, again because of a lack of controls, it is not possible to determine whether they did any better than persons who were not treated (Hain, Hanna et al. 1999). Thus the efficacy of vestibular rehabilitation for MDD is unknown.
Motion sickness has been treated successfully with habituation (Dai, Raphan et al. 2011), and one might reasonably argue that MDD, being a motion sickness variant, might also respond to a similar approach. Habituation entails a down-weighting of motion input, and can reduce the long duration vestibular responses commonly associated with motion sickness susceptibility (Dai, Raphan et al. 2007). Although there are well developed self-directed motion habituation protocols such as the PUMA exercises (Puma 2010), there are presently no published reports of their efficacy in MDD (or motion sickness for that matter). Nevertheless, we are sympathetic to the general idea that things that make you feel worse (when you are dizzy) usually does result in some improvement (if you can stand it). The Puma protocol exercises are just so extremely stimulating that so far -- nobody has been able to tolerate them for more than a session or two. The Puma protocol exercises can be bought on the web in the form of a DVD from Dr. Puma's website.
The roll adaptation protocol of Dai et al (2014) could be reasonably viewed as a type of habituation. We are presently implementing this treatment in our practice in Chicago.
While we find this idea very doubtful, if MDD is indeed due to inappropriately high weighting of somatosensory input, vestibular physical therapy protocols that teach down-regulation of somatosensory input seem worth trying. Liphart (2015) reported results of "sensory reweighting therapy in a single atypical case. The single subject "felt she had improved 50%". A controlled trial of vestibular rehabilitation in a large number (i.e. 20) of MDD subjects could be helpful in clearing up this question (hope someone funds this !). Our guess is that results of vestibular rehabilitation treatment would not be any different than no treatment.
If MDD is instead caused by an internal oscillator developed to predict boat motion, one's treatment strategy should be aimed at manipulation of psychological variables rather than somatosensory integration. Patients need to ignore their aberrant internal signal, in the same way that most persons with tinnitus eventually develop an ability to ignore abnormal internally generated sounds. Treatments that decrease vigilance, obsessiveness, and anxiety as well as "tincture of time", would be the optimum strategy.
Going on another boat:
MDD is not very well studied. As of 2013, only 29 papers came up on a "Pubmed" search for "mal de debarquement". Not much has changed in 2015. Considering that many other obscure conditions have 1000's of papers written about them, this means that MDD has been generally ignored. There are many open questions. Here are a few:
- Controlled trial of medication(s) for MDD. Is clonazepam effective ? How about anticholinergics ? Antihistamines ? phenytoin ? oxcarbamazine ? gabapentin ? (our guess -- clonazepam is effective, none of the others are reliable or significant).
- How does MDD and migraine interact ? Do medications for migraine reduce probability of MDD ? (our guess - -certain ones -- venlafaxine - -might help)
- What factors affect duration of MDD ? Does ongoing motion exposure that alleviates symptoms (i.e. driving) slow down recovery ? (our guess- - yes)
- Role of physical therapy (if any) in MDD ? Compare vestibular physical therapy to strength training (for example). (our guess -- PT doesn't help)
We know of three ongoing research projects regarding MDD, two treatment projects in Oklahoma (Dr. Cha), another in Ohio(Clark)
We do not know of any projects involving medication.
There is a MdDs foundation, which maintains it's own website and encompasses a quite active group of volunteers. The author of this page, Dr. Hain, is loosely associated -- I sometimes provide some advice to this group.
Click here to return to index