For more than two decades, Boston economist Paul Greenberg has been calculating the costs of depression — the mood disorder, that is, not the economic downturn.
His latest study, now out in the Journal of Clinical Psychiatry, finds that major depression is costing the American economy $210.5 billion a year — boosted dramatically by the toll of the recent recession. And rates of depression particularly shot up among people over 50.
I asked Greenberg, head of the health care practice at the Boston economic consulting firm Analysis Group, to elaborate. Our conversation, lightly edited:
First, what would you most highlight from your latest findings?
There are many highlights but I’d focus on two. The first is that the costs of depression are large and growing. And the second is that costs of depression are borne in the workplace in a very dramatic way. There’s no employer that’s exempt from the costs of depression. And I think both the magnitude of costs generally, as well as the costs that are specific to the workplace, are worthy of further attention, further thought, further research.
What are a couple of the specific numbers that you find most striking?
Let’s start with the overall finding: We find the costs of depression to be approximately $210 billion per year. One of the interesting aspects is that only 40 percent of those costs are actually attributable to depression itself.
Could you explain that?
That means that 60 percent is attributable to elevated costs that, in the data, don’t show up as directly connected to depression, but they’re associated with depressed people to a greater extent than with non-depressed people.
But should those costs really count toward depression, when it’s technically another disorder that’s causing them?To be more concrete, on the mental-illness side, there are an awful lot of co-morbid anxiety disorders, a lot of co-morbid PTSD-associated costs — those are examples where the same person who suffers from depression tends to have a higher likelihood of also incurring costs in these co-morbid categories.
Fair enough. That’s part of the age-old question of to what extent is this cause and what’s effect. Take the example of someone who suffers from cancer. It could be that in some instances, there’s an elevated cost of depression when you suffer from cancer. That’s one causal pathway where the depression likely follows the physical disorder. But in another instance, it could be that back pain or sleep disorders or migraines – those are examples of elevated physical disorder costs that accrue to depressed patients, likely as a result, at least in part, of the depression.
Here’s why it matters. If we’re more successful at treating the depression, there’s little or no hope it will alleviate any of the cancer costs. But if we’re more successful at treating depression, there’s a great opportunity to alleviate some or even a large part of those back pain, sleep disorder and migraine kinds of costs that are currently co-morbid with depression.
To be clear, not everybody who suffers from migraine necessarily suffers from depression. But among the patients who suffer from depression, they tend to be more at risk for these other, co-morbid physical and psychiatric disorders. That’s what we’re capturing in total in the $210 billion.
So part of what’s new about the numbers is not just the sheer magnitude of the $210 billion, which is an enormously large number, but the part we found quite fascinating is to disentangle the costs of the depression from all the costs of the people with depression. The 40 cents on the dollar being attributable to the depression vs. the 60 cents on the dollar that are not — that, to us, was another important and interesting distinction.
Now I have to ask you a cynical question…
I see you had no outside funding for this study, but in the past there was funding from the pharma company Eli Lilly. These costs of depression look incredibly high and the upshot seems to be that we need more and better treatment. But could there be bias here?
It’s a fair question. Which is why we were very forthcoming in saying, not just for this study but even for past studies that we wrote up, exactly what the funding sources were. Look, the reality is that treatment quality is still fairly poor, so we need to do better on treatment. Whether that’s pharmaceutical therapy or some other kind of interventional strategy, the bottom line is we need to do better on treatment. So I’m not conflicted on that. I’m not conflicted on any of this — as you said, there’s no funding at all, this is just, in some sense, our own public service research initiative to continue where we started 10 and 20 years ago.
You found a particular rise in depression among older people. What’s happening?
This is where I just need to be clear that I’m the economist, not the physician, psychiatrist, epidemiologist. I can observe, and do observe, in the data there’s a large increase in costs that seem to be attributable to the 50-and-older sufferers of depression. I don’t know what underlies that.
One hypothesis — and it’s kind of a segue to another aspect of what this study has been about — we know from past research that there’s an important piece of this puzzle that has to do with depression in the workplace. What was new and different about this study is that we studied, in particular, the years 2005 and 2010.
2005 was like the peak of the bubble, the best times of all, 2005 and 2006. By 2010, the great recession has really established itself and manifested in a very dramatic way. One of the things we see is the consequences on employment patterns associated with whether or not you suffer from depression. What we found was that, while depression has always been known to be a significant workplace concern, where you are in the business cycle interacts very specifically with what kinds of outcomes ensue.
What I mean by that is that individuals who suffer from depression are highly employable in robust economic times, but when the economy starts to decline and individuals start getting laid off, unfortunately there’s a disproportionate adverse impact on depressed workers compared to their non-depressed counterparts.
That shows up in at least two different ways. One, they become unemployed to a greater extent, but two, there’s also the buffer of part-time work that’s often available to many people as a stopgap. That buffer of part-time work tends not to be nearly as widely available to depressed workers as their non-depressed counterparts.
Do we know why?
No. I can only point to the data that tease out that result. The key obvious next step would be asking why would that be? What’s the dynamic that makes a person employable as the economy is doing well but becomes not just less employable when the economy reverses but also less part-time employable compared to non-depressed workers?
So it’s a very dire picture, that the recession particularly clobbered workers with depression…
It probably did. You just have less of a support network or less room to maneuver when you suffer from an illness like depression.
You first published research on the costs of depression way back in 1992. Are you going to keep going?
I’m going to keep going. I didn’t plan it this way. When you’re first starting out, you don’t say, ‘I’m going to keep doing this study every 10 years and see what changes.’ But it just kind of works out this way. I’m very confident this will always be a topic of great importance in the United States and internationally. I have no doubt in another 10 or 20 years, as long as I’m able to and there’s interest, I’ll be forever interested in updating these findings. Poking at what’s new, and asking both what has stayed the same and also what’s changed.
Some numbers from the press release:
· The number of adult Americans with MDD [Major Depressive Disorder] rose from 13.8 million to 15.4 million between 2005 and 2010, a 9 percent increase.
· For every dollar spent treating depression, an additional $4.70 is spent on comorbid costs, and another $1.90 is spent on a combination of reduced workplace productivity and suicide-related economic losses.
· Among Americans age 50 and older, the incidence of MDD rose nearly 50 percent between 2005 and 2010. Other groups held steady or saw a slight decline.
· MDD had a dramatic impact on employment, especially in light of the 2008 economic downturn:
o Full-time employment among the MDD group fell 6.9 percentage points or 8.5 percent.
o ~300,000 fewer people with MDD were employed full-time.
o ~300,000 more people with MDD were employed part-time.
o 1.6 million more people with MDD were not employed at all.