U.S. vital statistics have measured suicide and all the other causes of death for quite some time. But physicians are always becoming better at categorizing causes of death, so the classification system is updated every decade or so.
As reported by the Times, data collected using the current ICD-10 system reveal an increase in suicide rates for most age groups under 65 between 1999, the first year of ICD-10 collection, and 2004. Rates among adults 45-54 rose by 20%, from 0.000139 to 0.000166.
Temporal fluctuations in mortality are certainly interesting; there is evidence that many causes of death are more prevalent during economic expansions, for example. Suicide, on the other hand, typically falls during expansions and rises during recessions. This is discussed in a 2005 article by Jose Tapia. Seen in this light, a rise in suicide between 1999 and 2004, two years of relative prosperity, is a little odd.
One wonders whether physicians need time to learn a new cause-of-death coding system.