J a p a n e s e    C u l t u r e

Upon hearing of Watter's book (Watters, 2010; ウォッターズ. 2013) and the allegations he made regarding the lack of rigour in introducing new antidepressant drugs (SSRIs) to Japan, and knowing how easy it would be to portray the Japanese as depressed compared to Westerners due to their lack of a need for positive self regard (lack of a need to boast: see Heine, Lehman, Markus, Kitayama, 1999), I became alarmed at the possibility that there may be a link between SSRI use and suicide in Japan.

Fortunately there is a study (Nakagawa., Grunebaum., Ellis, Oquendo, Kashima, Gibbons, & Mann, 2007) that appears to prove the opposite: SSRIs were concluded to reduce suicide among the Japanese as well. Phew

The main graphs for men and women from that study is shown above top. I have added some arrows drawn in by hand (without any calculation) to illustrate the my understanding of the authors' conclusions.

The first author claims on a Paxil web page paxil.jp/documents/da/ev/ev004.php (from where the above image is taken) that while there is an overall correlation between SSRI antidepressant use and suicide (illustrated by my red arrows for both men and women) when one looks at the correlation between SSRI use and suicide in each age cohort, one finds a negative correlation between SSRI use and suicide rate as shown by the gently downward sloping blue arrows. The author claims that this is an illustration of "Simpson's Paradox," and that his data shows that SSRIs reduce suicide if one compares like with like.

(Simpson's Paradox is real, as many of the illustrations in the wikipedia page linked demonstrate. I have called Simpson's paradox on the data that suggests that Japanese working mothers have more children, arguing that this is in fact due to the high birth rate among poor women. I suggested, though have no evidence to demonstrate, that the tendency for dual income parents to have more children might be reversed if similar income groups were compared. )

So is this research fair to conclude that SSRI's reduce suicide rates among Japanese? On the face of it I was persuaded by the paper's conclusions since age does seem to be a very major predictor of suicide, and confounding factor in the relationship between SSRIs and sucide, so it would make sense to compare the relationship between SSRI use and suicide among similar age cohorts.

At the same time, I had assumed that the high instance of suicide among older people is due to the higher instance of health related suicide - the most common cause of suicide, rather than depression related suicide. What surprises me is that older people should be taking so much more antidepressants. In other words

1) If older Japanese people are simply more depressed than younger people, and therefore both taking more anti-depressants and committing suicide more, then Nakagawa et al.'s conclusions, and the suggestion that one should look at age cohorts, would seem to be entirely sound.

2) If older Japanese people are being diagnosed more as depressed, taking more antidepressants and therefore committing suicide more then that would be a tragedy.

How can one differentiate between the two hypotheses?

Prior to the introduction of antidepressants ("a cold of the heart") there were far fewer diagnoses of depression in Japan so it would be difficult and unfair to look at diagnoses of depression in the various age groups prior to the introduction of antidepressants. Of course there would have been fewer diagnoses since it was only after the arrival of the drugs that doctors started asking the questionnaires that would allow the prescription of the drugs.

However, if the arrival of anti depressant drugs should have been accompanied by an increase in suicide in older but not younger people then that might suggest reason (2).

After going through this reasoning, I went to see Japanese government data showing suicide rates for age cohorts, which I add below the first set of graphs.

Scarily, there does seem to be an event in about 1999 which spreads the suicide rate especially amongst males with many more older people suddenly killing themselves. The ratio of the suicide rates of young people 15-24, to that of 45 to 64 year olds jumps from three to five to one. In other words, prior to the arrival of antidepressants, the would be "confounding factor:" "old people are depressed and kill themselves more" was not nearly so true.

As pointed out by his New York Times article (Did Antidepressants Depress Japan?) and indeed Nakagawa et al.'s article shows that the introduction of antidepressants was also 1999.

That sudden dramatic post 1999 rise in death by suicide among older men represents in the first ten years of the decade represents more than 120,000 lost lives.

It also seems to me that if one answers a popular Japanese language self-check form for depression in a Japanese geezer (ojisan) way, being honest, as Japanese people are about ones declining vigour, then one is diagnosed with being "slightly depressed," merely by being honest, and perhaps therefore given drugs. This might provide a mechanism for why older people in Japan would be likely to be prescribed more anti-depression drugs.

There is also data to suggest that Japanese try harder in situations of adversity, so that would provide a mechanism for why those traditional Japanese might take their own lives when proscribed SSRIs (Heine, et. al, 2001).

At the very least Nakagawa's statement that, "After eliminating the effects of long-term linear trends, we found annual increases in antidepressant treatment were associated with annual decreases in suicide rates. " appears to be difficult to comprehend in view of the fact that 1999 was marked by a striking non linear increase in Japanese suicide rates, and it appears that that the "long term linear trends" were eliminated only for the period AFTER the introduction of SSRIs (1999-2003) and the contemporaneous step like increase in suicide among older males that continued and continues more than a decade later.

Bibliogaphy
Heine, S. J., Kitayama, S., Lehman, D. R., Takata, T., Ide, E., Leung, C., & Matsumoto, H. (2001). Divergent consequences of success and failure in Japan and North America: An investigation of self-improving motivations and malleable selves. Journal of Personality and Social Psychology, 81(4), 599.
Heine, S. J., Lehman, D. R., Markus, H. R., & Kitayama, S. (1999). Is there a universal need for positive self-regard?. Psychological review, 106(4), 766.
Nakagawa, A., Grunebaum, M. F., Ellis, S. P., Oquendo, M. A., Kashima, H., Gibbons, R. D., & Mann, J. J. (2007). Association of suicide and antidepressant prescription rates in Japan, 1999–2003. The Journal of clinical psychiatry, 68(6), 908. Retreived from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3804897/
Watters, E. (2010). Crazy like us: The globalization of the American psyche. Simon and Schuster.
ウォッターズ.E著 阿部 宏美訳 (2013)『クレイジー・ライク・アメリカ：心の病はいかに輸出されたか』紀伊國屋書店

Labels: culture, hazel markus, japanese culture, nihonbunka, 日本文化