What is a 'Symptom'?

When I studied psychopathology at Hopkins, I was taught that a symptom is an observable manifestation of an underlying pathological (disease) process, and could be distinguished from a 'sign' of the process as in the distinction between 'the patient says he feels hot' versus 'I measured the temperature and it is one degree above normal' or as something akin to what might be answered when the patient is asked "So what is bothering you and brings you to see me today?" versus what the MD discerns that the patient might not know about. The answer to that question might or might not be a symptom. For example, the patient might say "I came to you because I need a script for Adderall." And that would not be a symptom. Or, if you have started to down-regulate the dose of an opioid analgesic, the patient might say "I'm not getting the same pain relief when I take the drug," and it would not be 'tolerance' but rather your own down-regulation of the dose that makes the difference.

So, when we look at the natural history of drug involvement, what are we to make of a cannabis user who says "I have to take more of the drug to get the same effect." Presumably, if the user is getting the drug from an official MM vendor, dosage is more or less held constant. Or is it? And if the user is getting it on the street, the dosage is even more unknown. So, is that report a substantiation of a 'symptom' of a drug dependence process? To hard to tell.

For this reason, we use the term 'clinical feature' when we are uncertain about whether the user report is a symptom or whether it's something else (an 'experience'?).

Tolerance is a clinical feature of the neuroadaptational facet of drug dependence, and we ask about tolerance in our assessments of drug dependence, but the answer to that single question doesn't make it into a 'symptom' or 'symptom report'. It is a report on a drug experience, that in context we think of as a hypothesized clinical feature of what might be an underlying syndrome.

But suppose the user reports this 'tolerance experience' but nothing else remarkable is going on? Then is it a symptom? For sure not ! But is it still a 'clinical feature' for drug dependence. Yes, it is, but not 'of drug dependence as observed in that user'.

We have some papers on the profile of symptoms observed in users who qualify for the drug dependence syndrome case definition. In that context, when the user clearly qualifies as a case of drug dependence, we are well within standard psychopathology scientific practice to say that we are studying the profile of symptoms in a set of users found to qualify as cases of the drug dependence syndrome.

But when we are studying emergence of clinical features, one by one, in a set of newly incident users, irrespective of whether they have become cases of drug dependence, then we are not studying the symptoms of drug dependence in those patients. Hence, we shift to the term 'clinical features'.

Many years ago, we used the vocabulary terms 'experiences' and 'problem experiences' and even 'problems' in order to avoid what then was called 'over-medicalization' of illness experience. (See Austrian philosopher Ivan Illich.) About the time of the NIMH ECA studies, we were allowed to speak of symptoms and clinical features even when talking about drug experiences.

[What's the problem with 'problem'? Well, consider when you ask an LSD user whether the drug has caused them to have hallucinations or illusions, and they enthusiastically say they have had that happen, and then you ask "Was that a problem for you?" The response often is a puzzling look, shift of eyebrows, etc. (That's why they were taking the LSD in the first place. It wasn't a problem for them. It was what they were seeking in the LSD experience.) Not a problem. Just an experience.]

Until the pendulum swings again, I urge you to think carefully about whether you are studying 'symptoms' as can be seen in suspected or confirmed cases, but more of the time, you can save yourself trouble with reviewers by speaking of these experiences as 'clinical features' of the syndrome or as 'clinical features associated with' the syndrome. That is, the object of study is anchored as something that is found in the context of a syndrome and helps to define the syndrome via its 'running together' with other clinical features. And then the methods of study, if limited, become interpretable as a mixture of signal and noise, where the signal would be an experience that truly is a 'symptom' of an underlying pathological process.

But for the most part, reviewers tend to go along with 'clinical features' (and many would accept, without thinking, the use of the term 'symptom' if they never have studied psychopathology seriously). But you can be disciplined about your use of the term 'symptom' and show that you are a serious student of psychopathology.

Unless our trainees are studying known cases for whom the pathological process is well-characterized, I ask them to use the term 'clinical feature' or 'experience' in lieu of 'symptom' and that seems to satisfy even the most serious students of psychopathology one encounters in the peer review community. 

Even once in a while, a reviewer will insist on use of the term 'symptom' and I find that a note to the editor of the journal is sufficient to address that kind of reviewer misbehavior. [Proper reviewing behavior would be to raise a minor word choice issue and suggest an alternative, but not to insist on one's preferred word choice as a condition of accepting a manuscript for publication. But proper reviewing behavior is another topic. I won't post a blog on that topic unless the reading audience files comments to ask for it. I'm enough of a pontificator as it is, but you'll have to encourage me to pontificate on that subject, which is not really about neuropsychiatric epidemiology.]