Does long-term use of “antipsychotic” drugs cause more disability & more psychosis?

This sounds like a weird question – everyone knows that psychosis is often very disabling, and antipsychotic drugs are widely recognized for their effects in reducing psychosis in at least most people, and most often taking effect in just a few days.  And when people become psychotic again, it’s often understood that it’s because they “weren’t taking their meds.”

But what if it’s trickier than that?  What if “antipsychotic” drugs make things better in the short term, but make long term problems worse?  How would we even know?

In a recent letter to the Psychiatric Times, psychiatrist Sandy Steingard outlined some of the ways we can know that there definitely is a problem with the long term use of antipsychotics.  (Note that while she addressed a limited number of studies, that’s just because there actually are very few studies which look at really long term outcomes.)

She started her letter by writing about the Wunderink study, which found dramatically higher rates of recovery among the group that had been randomly selected 7 years earlier to receive a trial in getting off antipsychotic drugs, compared to those maintained on the drugs as usual.

It should be stated that the results of the Wunderink study are not perfectly clear in all respects.  For example, of the members of the group that guided discontinuation of the drugs, most had resumed taking at least some drugs over time, though the dosages on average were much smaller than those of the “treatment as usual” group.  So some have argued that the positive effects might have come from lower than average doses, and they argue that the study should not be taken to indicate that any use of drugs is detrimental long term.

But Sandy brought up more evidence, and the case against long term use of antipsychotics became more convincing as she continued.

She then discussed the Northwick Park study, where people were randomized to receive either antipsychotics or placebo over a 2 year period.  While members of the group on placebo was more likely to relapse into psychosis, they were also more likely to be employed.

And in case you are thinking that it might be worthwhile to have drugs interfere with employability if the tradeoff is reduced relapse, consider that the Wunderink study found that while those taking less drugs were more likely to relapse in the first two years, they were actually less likely to relapse in following years, such that there was no advantage to taking more drugs in forestalling relapse overall.

Then Sandy described a study done by  Gleeson and colleagues which attempted to see if helping people “adhere” better to taking medications as prescribed would help people have better outcomes.  In the first year, better adherence to drugs seemed to be helping reduce relapse, but after 30 months, overall relapse rates were similar in both groups, while those doing better at taking their drugs were less likely to be working, in other words, they were more disabled.

Finally, Sandy described what is possibly the most damming evidence of all.  This evidence comes from a 20 year outcome study done by Harrow and colleagues. It’s a naturalistic study, which means it just follows what people did, so some have argued that the much superior outcomes for those that those who came off drugs during the study period were achieved because these were the people who had recovered, rather than being because the drugs themselves impaired recovery.

But one way to sort out which is which is to look at one came first, the recovery or quitting the drugs.  To do this, we can compare those who stayed on drugs over the 20 year period, with those who got off drugs within 2 years and then stayed off them.  Here’s Sandy’s summary on that:

At 2 years, 74% of individuals in group 1 [those who stayed on drugs over the whole period] had psychotic symptoms, as did 60% of those in group 3 [the group that quit drugs by 2 years and stayed off]. Although these differences are not statistically significant, the lines diverge at year 4.5 and continue to diverge over the next 15 years. At 4.5 years, 86% of group 1 have psychotic symptoms compared with 21% of group 3. By year 20, the difference is 68% compared with 8%.

68% with psychotic symptoms when staying on drugs compared with 8% psychotic symptoms for those who quit drugs (that are called “antipsychotic”) – that’s about exactly the opposite of what the public has been led to expect!  But it’s precisely what we might expect if these drugs impair recovery from psychosis rather than promote it.

So where does that leave us?

Some psychiatrists who are aware of these studies, such as Torrey and Pierre, attempt to defend the drugs by maintaining that while some people can do well by getting off the drugs, others “clearly” benefit from continuing to take them long term.

My question is, how exactly can anyone, psychiatrist or not, know that a medication is “clearly” offering a long term benefit to a particular individual, when we can’t compare the person’s actual history with what would have happened to them had they gotten off the medications?

It seems to me that the closest we can do is to compare groups of people who get off or who are assisted in getting off medications, with people who stay on medications, and see who does better. As outlined above, this kind of evidence strongly suggests, even though it doesn’t absolutely prove, that staying on medications is likely unhelpful compared to getting off.

Of course it remains possible that medications are of benefit to some people in the long run, it’s just that this is not at all clear, and the appearance of clarity can be achieved only by ignoring the facts. This kind of ignoring is unfortunately all too common, perhaps because it is very disquieting for mental health professionals to consider the possibility that in most or possibly even in all cases, long term antipsychotic drug use is more damaging than helpful.

People often think it is “proven” that a particular individual needs to stay on antipsychotic drugs when that individual makes a number of attempts to get off and each attempt results in a relapse. But this in fact is not such proof. I know one person who reported about 20 such attempts before she got off successfully, and while that number is a bit high, many others also have stories of a string of failures before success. If we quit being sure this was impossible, we might put more energy into helping people succeed in getting off drugs and having a better chance at recovery.

The state of mind we call psychosis is often scary and even destructive, and it makes sense to believe that when nothing else seems to work, it may be helpful to reach for the kinds of drugs we call “antipsychotic, at least for awhile.  But if we take the long term studies seriously, I think we will work hard wherever possible to find ways of helping that don’t involve using these drugs, and when they are used, we will then work hard to support people in attempting to come off them safely.

Such efforts may not always succeed, but I believe both disability and long term psychosis could be reduced dramatically if we also very dramatically reduce use of the so-called “antipsychotics.”

In other words, yes it is sometimes risky to avoid using “antipsychotics” or to come off once on them, and sometimes and in certain situations starting to use or staying on them may be better than not, but we have to balance the risks of not using them with the risks of using or continuing to use them, and we are starting to see that those risks can be extremely high.  It’s not just the risk of “side” effects, things like weight gain and diabetes and permanent movement disorders and/or akathisia, but it’s also a likely increase in the very things the drugs were meant to reduce, the disability and the psychosis.

Discussion about this issue needs to happen throughout our mental health system, so that people can start to make informed choices.  Writing this column is one way I’m trying to get the dialogue going:  what steps will you take?  I’d love to see your ideas in the comment section . . .

* * * * *