While the U.S. Army began publishing annual suicide rates in 1843, presently the CDC only has confidence in annual suicides rates for U.S. civilians from 1900 onward due to a lack of uniformity and reliability in U.S. state-level tracking and reporting in the 19th century. Thus, Fig. 1 graphs U.S. Army, age adjusted U.S. civilian, age adjusted U.S. civilian male, and age adjusted U.S. civilian female suicide rates from 1900–2020. Overall, U.S. Army annual suicide rates followed trends in the civilian population, but did so in a more dramatic fashion. That is, increases in U.S. civilian rates corresponded to more drastic spikes in U.S. Army rates, while decreases in U.S. civilian rates corresponded to steep declines in U.S. Army rates. As an extreme example, the precipitous decline in suicide rates during WWII saw U.S. Army rates even fall below the U.S. civilian age adjusted female rates for 1944–1945, an occurrence never reported before or since.
The elevated U.S. Army and civilian suicide rates in the first half of the 20th century that decreased in two successive phases during World War I (1914–1918) and World War II (1939–1945), appear to stabilize by midcentury possibly indicating a dramatic shift in suicidal behaviors. While the most recent (2006–2020) increase in U.S. Army suicide rates is significant when compared to U.S. Army rates over past few decades (1950–2000), the recent increase in suicide rates is present to a more gradual degree in U.S. civilians as well. Nevertheless, the recent increase in U.S. Army rates above overall U.S. civilian, U.S. civilian male, and U.S. civilian female rates is not unprecedented or unique. For example, during the 120 years graphed in Fig. 1, the U.S. Army annual rate was higher than the age adjusted U.S. civilian female annual rate in 119 of the years (99.1%), U.S. civilian annual rate 88 of the years (75.8%), and U.S. civilian male rate annual 51 of the years (42.5%).
In an effort to increase similarities in populations for more precise comparisons, Fig. 2 graphs the overall U.S. Army suicide rate against the U.S. civilian male age-specific death rates (ASDR) by civilian male age ranges from 1900–2020. Specifically, the study used the CDC reported male ASDR for suicide in male ages ranges of 15–24, 25–34, 35–44, 45–54, and 55–64 as these are the age ranges that most closely resemble the U.S. Army population. Unfortunately, the U.S. Army only started reporting annual suicide rates for female soldiers towards the later decades of the 20th century, despite women officially serving in the U.S. Army since 1948.
Until such time as the U.S. Army provides more complete historic rates for women, the authors acknowledge the limitations and statistical problems associated with being unable to accurately separate male and female solider suicide rates. Nevertheless, the authors elected to use the overall U.S. Army rate as an approximation of the U.S. Army male rate because historically women tend to constitute a small percentage of the U.S. Army population. The resulting Fig. 2 seems to show U.S. Army suicide rates paralleling U.S. male civilian rates in a more focused fashion to what was displayed in Fig. 1, as U.S. Army rates following trends in the civilian male rate are easier to identify when the civilian male population is subdivided into the age ranges selected. However, U.S. Army rates appear to break from the U.S. civilian male paralleling trend from about 1975 to roughly 1990, when U.S. Army rates decreased and stabilized below the U.S. civilian male rate.
Intriguingly, data shown in Fig. 2 also seem to exhibit a new, previously unidentified trend; that is a long-term compression or narrowing of the different annual rates of suicide between the civilian male age ranges as well as with that of the U.S. Army. For example, starting in 1900, the suicide rates are relatively stratified and spaced out, ranging from 7.7 for 15–24 age range to 40 for 55–64 age range for civilian males, and a rate of 29.8 for the U.S. Army. The standard deviation between all six rates graphed for the year 1900 is 10.8. However, by the end of the graph (2020), the rates appear to be converging on a much narrower range of rates with 22.4 for 15–24 age range to 27 for 55–64 age range for civilian males and a rate of 36.4 for the U.S. Army. The standard deviation between all six rates graphed for the year 2020 is 4.1, or an 81.6% decrease in standard deviation from that calculated for 1900.
The trend of U.S. Army and civilian male rates narrowing is first observed in approximately 1950 (standard deviation 12.4) with the narrowing of the suicide rates continuing until 1975 (standard deviation 4.3), for a reduction of standard deviation between suicide rates of 65.3%. From 1975 onward, the U.S. Army rate decreased below the civilian male rates and seemingly broke the paradigm of suicide rates compression. However, the trend of U.S. Army and civilian male suicide rates from 1950–1975 narrowing does not look to be a historical anomaly as the phenomenon appears to begin again from 2001 (standard deviation 4.7) to 2019 (standard deviation 2.9), for a reduction of standard deviation between suicide rates of 38.2%. Thus, the U.S. Army and U.S. civilian male rates appear to presently be in midst of the second, albeit less pronounced, compression of annual suicide rates. If this compression will continue is a cause for further study as U.S. Army and U.S. civilian rates may be diverging again given the increase in U.S. Army suicide rates of 36.4 for 2020 and 36.18 for 2021, which were significantly above their U.S. civilian male peers in 2020 and could be in 2021 as well.7
To more closely examine the change in standard deviation over time, Fig. 3 graphs the standard deviation between the U.S. civilian male age specific death rates for the five civilian male age ranges from Fig. 2 as well as the standard deviation for these civilian male age ranges with the U.S. Army rates included from 1900–2020. The graph appears to show the steady decline in standard deviation between suicide rates for both the civilian male age ranges selected and with the U.S. Army rates included over the course of 120 years. Given that adding the U.S. Army – i.e., one additional rate per year – to the civilian male standard deviation rate should only minimally affect the overall standard deviation, it is again worth noting the gap between the civilian male and civilian male plus the U.S. Army lines from approximately 1975 to 2006. The gap indicates that greater divergence between the civilian male and U.S. Army annual suicide rates for that period. The historic trend of U.S. Army rates following the fluctuations in the U.S. civilian male population appears to resume from 2007 to 2019 before diverging again in 2020.