Borderline or Bipolar: Can 3 Questions Differentiate Them?


Treatments for borderlinity and bipolarity are quite different. What if just 3 items from a standard screening questionnaire could increase your diagnostic certainty by 30%?

Table. Probability of bipolar disorder



Treatments for borderlinity and bipolarity are quite different. Which approach should you consider for a patient with impulsive risk-taking, episodes of irritability and hostility, fractured relationships, substance use problems, and severe depressions with brief phases of remission (maybe too good?) in between?

The Prisoner’s Dilemma paradigm separates the two,1 but that’s not practical as a clinical tool. What if you could pluck just 3 items from a standard bipolar screening questionnaire and increase your diagnostic certainty by 30% when faced with this common differential? That may be possible, based on a new study from Nassir Ghaemi and colleagues, led by Paul Vöhringer.2

Of course, replication studies will be needed before we can declare a new diagnostic approach is at hand. But in the meantime, I hope you might be curious: what 3 items from the good old Mood Disorders Questionnaire (MDQ)3 were so discriminating?

Vöhringer et al2 obtained an MDQ from 260 patients whose diagnosis was then established by structured interview (the usual gold standard in this kind of study). Then they analyzed the individual MDQ items looking for those that discriminate well between bipolar disorders and borderline personality disorder. They found 3, a “clinical triad,” that had remarkable statistical power:

1. Elevated mood: “You felt so good or so hyper that other people thought you were not your normal self or you were so hyper that you got into trouble?”

2. Increased goal-directed activities: “You were much more active or did many more things than usual?”

3. Episodicity of mood symptoms: “If you checked YES to more than one of the above, have several of these ever happened during the same period of time?”

(If you’re looking at the standard MDQ, these are questions 1.1, 1.9, and 2.)

If all 3 of these questions are endorsed, does that mean the patient has bipolar disorder, not borderline personality disorder? No, but the probability of bipolar disorder is much higher. How much higher? Ah, that depends on how certain you were before you looked at these test results. Say that based on your clinical history you think the probability of bipolar disorder for a given patient is 50% (a solid “maybe”). If these 3 MDQ items are positive, the probability of bipolar disorder is 83%. That’s a significant increase in certainty, no?

Wait a minute: how did you arrive at that pre-test probability of 50%? That was your clinical hunch, based on your complete examination, including the DSM criteria: all 11. If you can’t remember them all, you need a mnemonic. (You can’t part your hair well if your comb is missing a third of its teeth.) Try DIGFAST, that’ll get you 7.4 The last 4 are the DSM’s A criteria: elevated, expansive, or irritable mood; in the context of increased energy. But you’re not done yet. You need the non-manic bipolar markers, another 10 or 11 items depending on how you count: family history, age of first depression, post-partum depression, all those other statistically associated features.5 This is getting too much for a routine interview, isn’t it? Why don’t you use a questionnaire to gather all this? Here’s one.

Of course, there is much more to determining the probability of bipolar disorder. Your differential includes not just borderline personality disorder but substance use, medical illness, medications (like prednisone or interferon), and even “normal.” So your exam queries for all these possibilities as well. But for a narrow differential of bipolar vs borderline, this new clinical triad may help. Here’s a Table that converts your clinical hunch (0% to 100% chance) to a probability when the triad is positive (derivation follows the method of Phelps and Ghaemi, 20126). Notice that the greatest value of the test is when you are most uncertain, ie, when bipolar disorder is a 50/50 probability.

This new study has some weaknesses, as pointed out by the authors, including the fact that the patients were seen in mood disorder clinics, so bipolar disorders were common and borderline personality disorder less so. A replication with a different sample is needed. Until then, consider this more a thought exercise then a technique. But it’s tantalizing, isn’t it?


1. Phelps J. Borderline or bipolar: objective data support a difference. Psychiatric Times. July 2016. Accessed December 5, 2016.

2. Vöhringer PA, Barroilhet SA, Alvear K, et al. The International Mood Network (IMN) Nosology Project: differentiating borderline personality from bipolar illness. Acta Psychiatr Scand. 2016;134:504-510.

3. Hirschfeld RM, Williams JB, Spitzer RL, et al. Development and validation of a screening instrument for bipolar spectrum disorder: the Mood Disorder Questionnaire. Am J Psychiatry. 2000;157:1873-1875.

4. Phelps J. Bipolar diagnosis: navigating between Scylla and Charybdis. Psychiatric Times. 2007. Accessed December 5, 2016.

5. Phelps J, Angst J, Katzow J, Sadler J. Validity and utility of bipolar spectrum models. Bipolar Disord. 2008;10(1 pt 2):179-193.

6. Phelps J, Ghaemi SN. The mistaken claim of bipolar 'overdiagnosis': solving the false positives problem for DSM-5/ICD-11. Acta Psychiatr Scand. 2012;126:395-401.


Related Videos
© 2024 MJH Life Sciences

All rights reserved.