The co-morbidity between bipolar disorder and borderline PD is quite high in some studies but the problem here is the reader must beg the question, just how much did the researcher do to tease the two disorders apart? They may have misdiagnosed.
I still, almost everyday, see people misdiagnosed with bipolar disorder that don't have it but have a cluster B personality disorder and tell me they know they have it because 1-a psychaitrist told them so and 2-"I have highs and lows."
To which I respond....
1-Bipolar Disorder misdiagnosis is actually quite high and some doctors diagnose everyone with it without spending the appropriate time or amount of information to scrutizine and 2- "Everyone has highs and lows, that doesn't mean you have bipolar disorder."
Another problem is some people have a sub-threshold bipolar disorder that's on the spectrum but doesn't meet a DSM-IV or V criteria of it. E.g. cyclothymia.
I've had the hardest time teasing cyclothymia and borderline PD apart.
Teasing apart actual bipolar I or II disorder and borderline isn't hard IMHO if you have enough information and the person is drug-free. Manic episodes don't last 1 hour or less. Depressive episodes don't last 1 hour or less. Average manic episode per Kaplan and Sadock lasts 3 months. (Though I caution that K&S didn't cite the specific reference for that source of information). If someone has bouts of mood problems that occur in a depression cluster, mixed cluster, or manic cluster that has literally lasted several days with no break whatsoever that strongly suggests bipolar disorder.
Someone having high emotional reactivity that lasts on the order of minutes to hours, especially in regards to impulsivity, interpersonal stressors and abandonment is likely a cluster B disorder.
Other softer signs of borderline over bipolar disorder: history of emotional/sexual/physical abuse or neglect as a child, very dysfunctional family structure as a child.
A problem, however, occurs when someone has both and one could have both disorders. In such cases the medication benefit will plateau and could stabilize the bipolar disorder but will likely not do much if anything for the borderline PD.
I mentioned in other threads that psychiatrists shouldn't be too liberal with diagnosing. This is especially true in this area. One study (that I can't find off-hand right now) showed that when psychaitrists used a SCID for diagnosing bipolar disorder it was accurate over 90% of the time but when using a very open format interview it accurately diagnosed it about 60% of the time. The SCID if you don't know uses very clear DSM-IV criteria. They were able to verify this because they followed the patients for about 5 five years (nuts I can't find this study and I have cited it often in court when trying to show the judge the person's prior dx of bipolar disorder is not accurate).
Here's one study though not the one I mentioned above about the problems with dx borderline PD vs bipolar disorder.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2849890/
Slightly off on a tangent but while I worked on a forensic unit a colleague of mine had a very large and strong female on her unit that frequently attacked others and was diagnosed with bipolar disorder. The psychiatrist became convinced it was all a combination of antisocial and borderline cause the patient only became enraged and attacked people if she didn't get something she wanted. If she got something she wanted she was very sweet and pleasant. I agreed with my colleague and commended her for having the spine to give the right dx despite a history of several prior dx of bipolar disorder and schizoaffective disorder.
When previous psychiatrists were asked why she was diagnosed with bipolar disorder despite the patient not really meeting a criteria of it the other doctors gave answer to the effect of -well I got to medicate her with something so I had to use this diagnosis even if I didn't believe in it.-
Incredible. What unseen hand forced this psychiatrist to misdiagnose? Why not just say what it is? Where is the evidence that intentional misdiagnosis and medicating a disorder that doesn't benefit from medications will benefit the patient?