August 19, 2008

It’sa Getting Warmer—I’ma Gonna Killa Myself

Filed under: Health Effects

Our ongoing quest for researchers making bizarre connections between (fill in the blank) and global warming frequently takes us to far-flung recesses of the library (or, more likely, dusty corner cobwebs of the World Wide Web). For this installment, we have uncovered some novel “climatology” being practiced in the Journal of Affective Disorders, a psychology journal “…concerned with affective disorders in the widest sense: depression, mania, anxiety and panic.” In a 2007 paper, provocatively entitled “Global warming possibly linked to an enhanced risk of suicide: Data from Italy, 1974–2003,” authors Preti, Lentini, and Maugeri argue that “global warming” has raised male suicide rates throughout the Boot.

After writing these essays for a few years, you get to the point where you can anticipate the study design, results, and interpretation without even having to read the paper. In this case, we assumed that suicide rates in Italy have, despite some annual ups and downs, been generally increasing since 1974. Of course, many places on the globe (Italy included) are warming. We likewise assumed the authors found a general increase in suicide rates. So, ipso facto, the two variables are positively correlated—write it up!

Well, not so fast. First, take a look at the Italian mean temperature history over their period of record, 1974–2003 (Figure 1). OK so far—it’s generally been warming, with maxima in 1995 and 2003 (recall the great western European heat wave that year).

Figure 1. Average annual temperatures in Italy, 1974-2003 (source: Preti et al., 2007)

Now, look at the male and female annual suicide rates (Figure 2). Hmmm. If the authors had said they lost all of the data since 1996, then the scenario I outlined above would play out. But this is a plot of suicide rates (per 100,000), which have clearly, systematically, and dramatically declined from the late 1990s through the early 21st century. And yes, this was also a period of increasing temperatures.

Figure 2. Annual number of suicides in Italy per 100,000 people. Males (top), Females (bottom) (source: Preti et al., 2007)

To strengthen their case, which is (clearly) not obvious from these figures, the authors produced seasonal correlations between suicides rates and temperatures (both as smoothed departures from normal). Seasonally, in eight trials they found one statistically significant correlation (here we’re using the standard scientific metric of 0.05 to represent statistical significance, or a 1 in 20 chance of erroneously rejecting the null hypothesis). When the data were run monthly, they only found one significant relationship in 24 (a positive relationship for females in July). This is the number of significant correlations expected to occur by random chance.

At some point during the course of a research study, one simply has to give up and move on to the next topic. But apparently undeterred, the authors decided that the best course of action was to throw out some of the data. With the appropriate months summarily exorcised from the study, the authors finally uncover significant positive relationships between above normal male suicide rates and high temperatures in May through August and November.

There seems to be a debate in the literature on whether there really is a relationship between temperatures and suicides, part of which relates to studies linking male aggressiveness to heat (as an aside, there was a classic study linking Major League Baseball hit batsmen rates to game time temperature). Last year, Mississippi State professor Grady Dixon and colleagues could not find any such relationships in selected U.S. counties. But this is irrelevant to this particular study, which suggests that there is evidence in the record of an effect from global warming, despite the fact that suicide rates have declined during recent, warm years. For example, the suicide rates in 2000 and 2003, two recent years with European heat waves, are in no way unusual compared to the long-term record.

One problem here is that non-climatologists confuse high temperatures with global warming. To make the global warming connection, you need to 1) find an increasing trend in the variable of interest (in this case, Italian temperatures), 2) demonstrate that the trend is likely related to greenhouse gases and other anthropogenic factors, and 3) demonstrate that the trend in your variable of interest (suicides) in linked to your independent variable. If we extrapolate the current trend from Figure 2 forward in time, global warming will be causing negative suicides by the year 2012.

This seems to be a rare case of authors so hell-bent on finding a relationship that they ignored their own analysis in the process. For example, consider the following excerpt from their paper in response to a reviewer’s comment about increasing cancer rates and earlier onset of neurological diseases among young people:

This may, as well, increase the risk of suicide, which is itself enhanced in these illnesses, leading to a spurious correlation with anomalies of temperature that are interrelated with all of the other effects of multi-pollution. Moreover, the waste of natural landscapes and the deterioration in flora and fauna related to multi-pollution and its effects on the natural world might further reduce the sense of satisfaction and happiness in people, excursion in natural environments representing an opportunity to buffer the negative stress of urban life.

Hmmm, again. Declining opportunities to just stop and smell the roses brought about by global warming is causing people to cash it in in record numbers. Interesting concept. We could just as well have suggested that the mortality peak in the early 1990s was the result of Italians being distraught after blowing an early one to nil lead against Argentina and losing as the host nation in the semifinals of the 1990 World Cup. But again, that would be wild speculation.

No doubt some people are stressed about pollution and global warming. Thank goodness we’re here to set the record straight, to allay these rampant fears that the end is near, and to put everything in a proper perspective. But admittedly, even we have had suicidal thoughts on rare occasion (most recently during a viewing of “An Inconvenient Truth”), but cooler heads prevailed, and we’re still here to continue our search for both traditional and offbeat claims about the purported causes and impacts of global warming.


Dixon, G.P, A.N. McDonald, K.N. Scheitlin, J.E. Stapleton, J.S. Allen, W.M. Carter, M.R. Holley, D.D. Inman, and J.B. Roberts, 2007. Effects of temperature variation on suicide in five U.S. counties, 1991–2001. International Journal of Biometeorology, 51, 395–403.

Preti, A., G. Lentini, and M. Maugeri, 2007. Global warming possibly linked to an enhanced risk of suicide: Data from Italy, 1974–2003. Journal of Affective Disorders, 102, 19–25.

No Comments

No comments yet.

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

Powered by WordPress