Caffeine & Coffee : Health Benefits

So, is coffee good for you? Or is it bad for you?
How about decaf?

It seems that there are mixed messages out there, about health effects from things such as coffee or caffeine.

What does the evidence tell us?

The evidence base is extremely compelling --

Here is a link to a study by Lopez-Garcia et al. from Annals of Internal Medicine in 2008, involving over 100 000 people, over 18 years of follow-up. Such massive studies with long follow-up time are incredibly informative:

http://www.ncbi.nlm.nih.gov/pubmed/18559841

In the study, it shows clearly that coffee drinkers have lower overall mortality rates, mainly due to a reduction in rates of cardiovascular disease. It showed a modest reduction in mortality rates associated with decaffeinated coffee as well. The authors conclude that the potential health benefits from coffee are due to components in the coffee other than caffeine.

The graph on the top is copied from the paper, and summarizes the results from the study. You can click on the graph to expand it. Actually, the data as presented on the graph suggests that people are healthier who have 2-3 cups of coffee per day, compared to those who have just 1.
Perhaps the less desirable effects from just a single cup per day are due to the more pronounced "jolt" of caffeine that would happen in a person who is not used to drinking as much coffee. Perhaps this effect would offset the health benefit, at low doses. At higher doses, the body may become more tolerant to caffeine's anxiogenic effects.


Unfortunately, this study did not look at mental health effects from drinking coffee, or from caffeine intake.

In another large epidemiological study of over 1400 people followed over 21 years, it was found that those who consumed coffee in mid-life had a substantially lower risk of developing dementia. This association was true after adjustment for demographic differences, lifestyle, and vascular disease.
http://www.ncbi.nlm.nih.gov/pubmed/19158424

This case-control study from Archives of Neurology in 2007 showed an inverse association between coffee intake and development of Parkinson disease (i.e. this relationship suggests that coffee could protect the brain from disease such as Parkinson's). However, smaller case-control studies such as this one are much weaker than the large prospective studies cited above:
http://www.ncbi.nlm.nih.gov/pubmed/17420321

The following 2008 article from Sleep Medicine Reviews presents a more cautionary account of caffeine's effects, particularly with respect to causing a dependence phenomenon, and to disrupting sleep quality. However, a lot of the data cited in this article is absurd, such as showing that a dose of caffeine immediately before sleep causes sleep disruption! Bedtime doses of coffee are not a realistic reflection of most people's caffeine intake!
http://www.ncbi.nlm.nih.gov/pubmed/17950009

Here's an interesting reference to an article showing that the tendency for caffeine to cause sleep disturbance is a heritable trait. That is, some people might be vulnerable to caffeine-induced insomnia, whereas others could sleep well regardless of their caffeine intake:
http://www.ncbi.nlm.nih.gov/pubmed/17969472

Here's a case report of a person with schizoaffective disorder improving upon discontinuing heavy caffeine intake:
http://www.ncbi.nlm.nih.gov/pubmed/18455857


In summary, it appears that coffee drinking is not harmful. In most cases, it may in fact be good for you. However, pregnant women should minimize their use of coffee. For some people, caffeine may cause or exacerbate anxiety or insomnia, especially at higher doses.

Buspirone

Buspirone is another of those medications that was introduced in the 80's, and was marketed for the treatment of anxiety. Most of the published studies on buspirone were done around 1990.

While many antidepressants simply increase the amount of serotonin or other neurotransmitters by blocking neurotransmitter re-uptake into neurons, buspirone works by directly stimulating one of the target receptors for serotonin, called the 5HT-1A receptor.

As with many new drugs, there was a wave of enthusiasm, which eventually faded. At this point buspirone is rarely prescribed.

In my opinion, it could be a useful and well-tolerated adjunct, to try in the following situations:

1) to treat generalized anxiety disorder
2) to augment antidepressants (i.e. to add to an antidepressant which isn't working well enough)
3) to treat antidepressant-induced jaw or tooth grinding (bruxism)
4) to treat aggressive or self-injurious behaviour; it may be particularly helpful in elderly patients with dementia, or in mentally handicapped patients
5) to treat migraine (a common comorbid problem among depressed or anxious patients)
6) to help with opiate withdrawal
7) to help quit smoking

Side effects are usually mild and subside with time; they include dizziness, nausea, sweating, or nervousness. About 10% of people in the clinical trials of buspirone discontinued the medication due to side effects.

Buspirone is metabolized through the cytochrome P450 3A4 system in the liver; because of this its levels in the body can be substantially increased by other medications or grapefruit juice, so these types of interactions have to be considered when choosing a dose.

I've been curious to revisit the evidence base for buspirone; here is my review of the literature:

1) Using buspirone as an augmentation to antidepressants, for treatment of depression:

http://www.ncbi.nlm.nih.gov/pubmed/17628435
{a good, important study from NEJM in 2006: 565 depressed patients who had not remitted despite receiving high-dose citalopram, were given augmentation therapy with either bupropion SR or buspirone. That is, the bupropion or buspirone was added onto their daily regimen of citalopram, and the patients were followed over at least 7 weeks. Both groups did similarly well, with about 30% of both groups having a remission. The bupropion group did slightly better in a few ways. Unfortunately there was no placebo augmentation group}

2) Treating generalized anxiety:

http://www.ncbi.nlm.nih.gov/pubmed/8666569
{a small study from 1996 showing that buspirone helps reduce anxiety symptoms in patients who also have mild depression; but the reduction in anxiety symptoms (about 50%) is only modestly different from placebo (about 35%) }

http://www.ncbi.nlm.nih.gov/pubmed/3320034
{this study from 1987 had a one-year follow-up of 700 patients. But it was open-label (no randomization, no placebo group). It did show that the patients taking buspirone for treatment of generalized anxiety showed sustained improvement, and tolerated the medication well}

http://www.ncbi.nlm.nih.gov/pubmed/17984162
{this 2007 study compares the effect sizes of numerous different medication treatments for generalized anxiety; buspirone fares particularly poorly, with a "non-significant" effect size of 0.17; SSRI's and venlafaxine do slightly better, and the novel anticonvulsant pregabalin actually does best. Complementary and alternative medications had a negative effect on symptoms, in this analysis. However, this meta-analysis is limited by the fact that most of the buspirone studies were done over 10 years ago and most of the results are from short-term treatment.}

http://www.ncbi.nlm.nih.gov/pubmed/2211567
{one of the small randomized studies comparing buspirone with a benzodiazepine for treatment of anxiety; the study shows similar effectiveness. Given that buspirone is non-addictive, it makes buspirone a more attractive option}


3) Treating other anxiety conditions:

http://www.ncbi.nlm.nih.gov/pubmed/2407755
{one of the studies showing that buspirone is NOT effective for treating panic disorder}

4) Improving cognitive function in schizophrenia:

http://www.ncbi.nlm.nih.gov/pubmed/17628435
{this 2007 study had a good randomized design, and 6 months of follow-up; it claimed in the abstract that buspirone had a beneficial effect on cognition when added to antipsychotics in schizophrenia -- but if you take a look at the actual data in the article, the differences in buspirone vs. placebo groups are very small. So I'm not impressed.}

5) Treating migraine:

http://www.ncbi.nlm.nih.gov/pubmed/16109114
{a small 2005 study in a headache journal looking at a group of 74 patients with migraine over 6 weeks of treatment; it showed that low-dose buspirone (10 mg) reduces migraine frequency by about 40%, and reduced anxiety scores by about 20%, both of which a substantial difference compared to placebo. The improvement in anxiety did not depend on the improvement in headache, they appeared to be separate, independent effects.}

6) Treating acute heroin withdrawal:

http://www.ncbi.nlm.nih.gov/pubmed/15876901
{an interesting 2005 study showing that 45 mg per day of buspirone can reduce symptoms of heroin withdrawal over a 2-week period; looking at the results directly, it appears that the effect is very substantial, that the buspirone almost eliminated withdrawal symptoms}

7) Helping quit smoking:

http://www.ncbi.nlm.nih.gov/pubmed/1739365
{a small 1992 study from Archives of Internal Medicine showing that buspirone helps with nicotine withdrawal, and may help people quit smoking}

8) Treating ataxia:

http://www.ncbi.nlm.nih.gov/pubmed/8806320
{another interesting study from Lancet in 1996, showing that buspirone helps improve symptoms of cerebellar ataxia, a type of brain disease which causes impaired balance & coordination}

9) Treating aggressive behaviours:

http://www.ncbi.nlm.nih.gov/pubmed/2016248
{a small study suggesting that buspirone can help reduce aggression and anxiety in mentally handicapped adults, without causing sedation or cognitive side-effects}

10) Treating bruxism:

http://www.ncbi.nlm.nih.gov/pubmed/10665633
{a 1999 study of 4 cases of SSRI-induced bruxism improving with buspirone}

11) Treating tardive dyskinesia:

http://www.ncbi.nlm.nih.gov/pubmed/8102622
{a 1993 study showing some improvement in tardive dyskinesia (a movement disorder) after treatment with buspirone for 12 weeks. However there are a few other case reports in the literature of buspirone causing worsened symptoms of various movement disorders, such as dystonias or myoclonus (twitching); but the incidence of such side effects appears to be very low}


12) Animal studies:

http://www.ncbi.nlm.nih.gov/pubmed/17312776
{in this study a badger in a zoo (!) was suffering agitation and engaging in self mutilation; "environmental enrichment" initially helped, but the behavioural problems still recurred. Buspirone ended up helping substantially, over an 18 month period, with no side-effects}

http://www.ncbi.nlm.nih.gov/pubmed/15766212
{in another animal study, buspirone helped reduce self-injurious behaviour in a group of rhesus macaques, and it seemed to help more than fluoxetine, with fewer side-effects}

Ferritin & Iron

Ferritin levels in the blood correlate well with the amount of iron available in the body's "reservoir". If ferritin levels are low, the body has very low reserves of iron. (the converse may NOT be true -- if ferritin levels are high, the body may still have low iron reserves, because there are a variety of conditions, such as inflammatory states, that can cause ferritin levels to rise)

Anemia is a condition in which the body does not have enough iron-containing red blood cells, therefore the body cannot deliver oxygen to the tissues (including the muscles, heart, and brain) as efficiently. One of the most common symptoms of anemia, not surprisingly, is fatigue.

Sometimes, iron reserves can be low, without actually causing anemia. It is like a low water reservoir: water may still be flowing into people's homes despite the water levels being low.

Here is a 2003 study from the major journal, BMJ, which shows that iron supplementation improves fatigue in non-anemic women with low ferritin:
http://www.ncbi.nlm.nih.gov/pubmed/12763985

This study, from the major medical journal Lancet in 1996, shows that iron supplementation given to non-anemic girls with low ferritin improved their verbal learning and memory:
http://www.ncbi.nlm.nih.gov/pubmed/8855856

Low ferritin levels are associated with a disease called "restless legs syndrome" (RLS), which causes discomfort and insomnia at night, and which can often give rise to a substantial reduction in quality of life. I suspect there are many milder cases of RLS which could be contributing to insomnia, and therefore contributing to resulting anxiety and mood problems. Here are some studies showing the association, and demonstrating that iron supplementation can improve RLS:

http://www.ncbi.nlm.nih.gov/pubmed/15854860

http://www.ncbi.nlm.nih.gov/pubmed/8085504

http://www.ncbi.nlm.nih.gov/pubmed/19200780


In this recent study from the Journal of Pediatric Neurology, children with ADHD and low ferritin levels showed improvement in their ADHD symptoms after receiving iron supplements:

http://www.ncbi.nlm.nih.gov/pubmed/18054688
In conclusion, I believe it is very important to evaluate ferritin levels, particularly in women, since the levels are frequently low; low ferritin is associated with fatigue, restless legs, ADHD, and reduced cognitive function. It could be a contributing factor to mood disorders and other psychiatric problems.

Usually, low ferritin levels are easily remedied by iron supplementation. Most daily multivitamins contain iron, but the amount of iron in these is usually enough only to maintain your iron stores, not to build them up. Similarly, increasing iron-rich foods in the diet will help to maintain or build iron stores, but this could take a very long time. In order to build up your iron stores more quickly, higher doses of iron salts, such as ferrous sulphate, need to be taken daily for several months.

I recommend aiming for a serum level of at least 50 micrograms / litre (50 ug/L). Many labs give a normal range starting at 20 ug/L, and therefore you may not hear from your physician if the level comes back at 25. It is important to know that the average for women is at least 50, and the average for men is about 100. Exceptions include children, whose ferritin levels are a little lower, and women in advanced stages of pregnancy, who have average ferritin levels of only about 20.

If you do have low ferritin, then further investigation could be warranted, to assess for other causes of iron deficiency (e.g. chronic blood loss from the digestive tract, from heavy menstrual flow, or from the kidneys). So your decisions on this matter should be discussed and followed-up with a primary care physician.

Pathological & Therapeutic Crying

This post is in response to a previous comment:

...So, maybe if there was a medication that would stop you from crying, depression levels could be taken down?

----
Antidepressants can directly reduce crying, probably independent of other effects on emotion.

There is a condition called "pathological crying" which can occur after a stroke or other brain damage; in this condition the afflicted person may be weeping uncontrollably, with or without a subjectively sad or negative emotional state. SSRI antidepressants can help greatly with this, here is a reference:

http://www.ncbi.nlm.nih.gov/pubmed/10576464

Here is a reference showing that mirtazapine could be an alternative:

http://www.ncbi.nlm.nih.gov/pubmed/16239769

A similar antidepressant-induced "reduction in crying" may sometimes be one of the sources of relief in depression. But such an effect could be unwelcome if it leads to a subjective restriction of emotional range.
Are tears therapeutic? Most of us would agree that crying is often a relief, or even a necessary emotional outlet. Many psychotherapists would consider a patient's tearfulness in a session to be therapeutic.

In my opinion, tears can sometimes be a relief, and can sometimes be very therapeutic -- but sometimes tearfulness can feed a cycle of exhausting, out-of-control sadness or despair. And sometimes tearfulness can be so reflexive that almost any event or trigger in daily life can bring it on. Occasionally tearfulness may be a type of "defence" which prevents dialog (with self or others) about an underlying experience or emotion.

I would add, as a formal personal opinion, that tearfulness need not be a sign of "loss of emotional control" (though sometimes it could be). A person who is confident, stable, mature, and sensitive may feel quite at ease with tearfulness. Tearfulness, in this case, is a normal, and often healthy, emotional display. For some individuals, tearfulness happens more naturally, for others they may be more comfortable experiencing emotion without tearfulness.


What does some of the evidence have to tell us:
http://www.emotionalprocessing.org.uk/tears/is%20crying%20good%20for%20you.htm
(a website with an introduction to the subject)

http://www.ncbi.nlm.nih.gov/pubmed/18509370
(here's a case study showing that crying can be therapeutic -- even if it is the therapist who cries; this is a position I strongly support, for a variety of reasons, most of all because I believe in the "Rogersian" idea of genuineness and transparency, also I believe that crying need not be considered a form of weakness, but a symbol of sensitivity and compassion, whether the tear is shed by the patient or by the therapist. Too many therapists are, in my opinion, so defended by various tactics of emotional detachment, that they become aloof, and in some cases ironically afraid of emotional intensity--this may in some cases lead to dismissive, or ineffectively cold, postures towards tearful or suffering patients. Of course, if the therapist is not functioning, and is tearful due to emotional fragility or depression, then the tearful therapist needs to take a break and seek therapy himself or herself.)

http://www.ncbi.nlm.nih.gov/pubmed/17587475
(an interesting look--from an anthropological perspective-- at a phenomenon called "wailing", a type of crying & lamentation which is part of a group ritual of mourning in the Yemenite Jewish community; this article includes interesting perspectives about crying and its theoretical role in bereavement. It suggests that the idea of "healthy bereavement" has been heavily influenced by Freudian, and largely "male", ideas, viewing emotions in a kind of "hydraulic" way -- as forces to be directed, or cathartically released, and in particular pronouncing healthy grief as a process involving letting go of the relational bond with the deceased. In the "wailing" phenomenon the author suggests that the group crying, accompanied by lyricism and dialog, may act to build a kind of emotional or relational "cradle" where the bereaved person may maintain a continuing loving bond with the deceased, rather than aim to let the bond go)

Beta-Blockers

Here's a link to a very interesting study which shows that the beta-blocker propranolol can interrupt the consolidation of fear in humans:

http://www.nature.com/neuro/journal/vaop/ncurrent/pdf/nn.2271.pdf

This study suggests a novel use for beta-blockers, which could facilitate behavioural therapy for PTSD. The study demonstrates a variety of things:
1) as was well-known before, when people experience something fearful or traumatic, it sensitizes them to react more strongly to the same fearful stimulus in the future
2) when people re-experience a fearful or traumatic memory, this re-experience consolidates, or strengthens, the strong fearful reaction. This is consistent with the evolution of PTSD and other anxiety disorders, in which an expanding variety of daily events can trigger and consolidate the fear (e.g. a survivor of a bad traffic accident may constantly re-experience traumatic symptoms when hearing traffic noise, loud sounds, etc.--and may start to avoid these situations. Every time this happens, the anxiety disorder becomes more entrenched).
3) Fears can be "extinguished" by re-experiencing the feared stimulus repetitively, in a safe setting. But the fear can be "re-kindled" after extinction more easily than in non-traumatized people (this suggests a permanence to "emotional memory" that can be only temporarily over-ridden by psychological techniques)
4) If the consolidation phase of fear or traumatic memory could be interrupted, then a person might not develop ongoing post-traumatic symptoms at all. In this experiment, there is evidence that propranolol can interrupt this consolidation.
5) Propranolol may disrupt the "emotional memory" consolidation but not the "declarative memory"--the former process may occur primarily through the amygdala, whereas declarative memory is consolidated mainly in the hippocampus. So, the use of propranolol would not "erase the memory" of a traumatic event--the facts of the event would still be remembered normally--but it might reduce the painful, reflexive feeling of emotional trauma associated with the event.

The study does NOT show that "propranolol erases memories", as some of the news headlines seem to be proclaiming. It DOES suggest that adjunctive propranolol may greatly enhance the effectiveness of behavioural therapy. It requires that the person use propranolol while engaging in exposure therapy. So, for example, a possible technique for treating PTSD or panic (especially new-onset) might be to use a 40 mg dose of propranolol 1-2 hours before a therapy session. In the therapy session, the memories of the upsetting events could be discussed. The propranolol might interrupt the process of these upsetting memories getting further consolidated, might facilitate a behavioural therapy process which would help the person feel emotionally comfortable with their thoughts and memories. This process may occur because of direct beta-blockade in the amygdala, which may interrupt consolidation of emotional memory directly.

Despite this encouraging study, there are a number of negative studies looking at using propranolol similarly, for example:
http://www.ncbi.nlm.nih.gov/pubmed/18761097

http://www.ncbi.nlm.nih.gov/pubmed/19060728

I think the main thing to take from the first study is that propranolol may help, but probably only as an augmentation to enhance the effectiveness of behavioural therapy (or CBT) for treating post-traumatic stress or other anxiety disorders.

Beta-blockers are drugs used primarily in cardiology. Some beta-blockers, such as atenolol, act only peripherally, that is they do not enter the brain very much. Others, especially propranolol, can more easily enter the brain, and therefore can act in the central nervous system as well as peripherally.

In psychiatry, propranolol has been useful to treat performance anxiety, especially if there is a component of tremor (e.g. shaking hands) accompanying the anxiety. Many musicians use doses of propranolol to reduce tremor during performances. The anti-tremor mechanism is most likely peripheral beta-blockade (i.e. outside the brain), but the accompanying reduction of subjective anxiety may also be due to central beta-blockade (i.e. inside the brain). This is consistent with some studies which show that peripherally-acting beta-blockers reduce tremor as well as propranolol, but people subjectively prefer the propranolol.
(Reference: http://www.mdconsult.com/das/citation/body/121508141-4/jorg=journal&source=&sp=6333536&sid=0/N/6333536/1.html?issn= )

Beta-blockers have been studied in the treatment of panic disorder, decades ago. They don't work. Here's a link to one of the many studies showing this:
http://www.ncbi.nlm.nih.gov/pubmed/2651490

Yet, these old studies don't look at the possibility that the beta-blocker could work as an "augmentation" to psychological therapy. Many effective treatments do not work on their own, they work only in conjunction with something else.

Beta-blockers have also been used to treat irritability or rage problems. Here are a few references:

http://www.ncbi.nlm.nih.gov/pubmed/15764868
{one of the studies in the geriatric psychiatry literature, showing possible benefit for using propranolol to help agitated dementia patients}

http://www.ncbi.nlm.nih.gov/pubmed/2136070
{an example of a study showing some benefit of propranolol treatment for reducing rage outbursts -- however the study is of low quality}

http://www.ncbi.nlm.nih.gov/pubmed/3546964
{another study from the Mayo Clinic in 1985, showing some success using propranolol to treat patients with rage outbursts}

http://www.ncbi.nlm.nih.gov/pubmed/9196923
{a review paper from 1997, looking at various pharmacologic treatments for aggression; some of the research about beta-blockers is reviewed here}

In summary of the above studies, beta-blockers may help a bit for irritability, aggression, rage outbursts, and agitation, due to a variety of causes, but the evidence base is mainly from before 1990, and the studies are not very rigorous.

Beta-blockers also help diminish a very uncomfortable symptom called "akathisia". Akathisia is a state of external, and internal, restlessness, that can be caused by older antipsychotic drugs.

Beta-blockers are also useful in migraine prophylaxis. Migraine is associated with depression, so a beta-blocker could be a good therapeutic choice in someone with migraines as well as anxious, irritable, or agitated depression.

There were a few studies suggesting beta-blockers could cause or worsen depression, but many of these studies are weak. Here is a review:
http://www.ncbi.nlm.nih.gov/pubmed/16466322

In a more recent major JAMA review, beta-blockers were not found to be causative of depression or fatigue:
http://www.ncbi.nlm.nih.gov/pubmed/12117400

In my opinion, beta-blockers should be used cautiously in people who have or develop depressive symptoms, but I don't think they are contraindicated, since they may be beneficial overall if they help other symptoms. Also, if there are depressive effects, these may be dose-dependent, and may disappear just by reducing the dose.

---

Beta-blockers literally "block" beta-adrenergic receptors in the body. These beta receptors are normally stimulated by the catecholamines adrenaline and noradrenaline (also called epinephrine and norepinephrine), which are hormones secreted by the adrenal glands and by a small area of cells deep in the brain called the locus ceruleus. There is always a little bit of these hormones in circulation (in quantities in the order of parts per trillion, concentrations which would be achieved by adding a single drop of hormone to the volume of 1-10 olympic-sized swimming pools).**

Here is a reference showing resting adrenaline and noradrenaline levels in healthy subjects:
http://hyper.ahajournals.org/cgi/content/abstract/30/1/71

These tiny quantities of hormone are nevertheless enough to stimulate beta receptors; such stimulation is required to maintain or increase the output of the heart, also many other actions in the body, including in kidneys and muscle tissue.

http://www.psychosomaticmedicine.org/cgi/content/abstract/52/2/129
{An excellent study looking at peripheral catecholamine levels (norepinephrine and epinephrine) in groups of patients with anxiety, patients with pheochromocytoma (a disease causing huge increases in catecholamine levels), and normal controls; they found that peripheral norepinephrine levels correlate with anxiety, but NOT in the pheochromocytoma patients; this supports a theory that anxiety states cause central, and secondary peripheral, stimulation of catecholamine release--but the catecholamines themselves do not necessarily CAUSE the anxiety, but are a RESULT of it. Incidentally, Psychosomatic Medicine is another excellent journal worth following}

http://www.psychosomaticmedicine.org/cgi/reprint/66/5/757
{a study showing that norepinephrine levels in the brain correlate highly with blood pressure in normal controls; but do not correlate at all with blood pressure in people with PTSD, suggesting that in PTSD there is an abnormality in catecholamine regulation}

http://ajp.psychiatryonline.org/cgi/reprint/158/8/1227.pdf
{a study from The American Journal of Psychiatry showing that people with PTSD have levels of CSF norepinephrine almost twice as high as normal, and that the norepinephrine levels correlate with the severity of PTSD symptoms}

http://www-personal.umich.edu/~nesse/Articles/AdrenFunctPanic-ArchGenPsychiatry-1984.PDF
{a study from Archives of General Psychiatry in 1984, showing higher levels of plasma catecholamines in panic disorder subjects; but less responsiveness to further adrenergic stimulation in the panic subjects--this suggests that anxious subjects have chronically high catecholamines, and consequently are actually LESS sensitive to catecholamine changes}

http://www.csbmb.princeton.edu/ncc/PDFs/Locus%20Coeruleus/Aston-Jones%20&%20Cohen%20(ARN%2005).pdf
(an article about the role of norepinephrine released in the brain's locus ceruleus, and its importance for optimizing performance of tasks)

**For the math, let us assume that the resting concentration of epinephrine is 100 pMol, or 10^-10 moles/litre; a litre of water has about 55.5 moles of water, so the concentration can be expressed as one part in (55.5 / 10^-10) or one part in 555 billion. A drop of water has a volume of about 1/20 mL. So this concentration of epinephrine corresponds to an analagous concentration of one drop in (555 billion/20) mL, which is about 1 drop in 28 million litres. An olympic swimming pool has a volume of about 2.5 million litres (http://en.wikipedia.org/wiki/Olympic_size_swimming_pool). So this concentration corresponds to 1 drop in a volume of over 10 swimming pools.