Statistical modeling of trends in infant mortality after atmospheric nuclear weapons testing

The global fallout from atmospheric nuclear weapons testing in the 1950s and 1960s caused by far the greatest exposure of mankind to ionizing radiation. Surprisingly few epidemiological studies of the possible health effects of atmospheric testing have been conducted. Here, long-term trends in infant mortality rates in the United States (U.S.) and five major European countries (EU5) were examined: The United Kingdom, Germany, France, Italy, and Spain. Bell-shaped deviations from a uniformly decreasing secular trend were found beginning in 1950, with maxima around 1965 in the U.S. and 1970 in EU5. From the difference between observed and predicted infant mortality rates, in the period 1950–2000, the overall increase in infant mortality rates was estimated to be 20.6 (90% CI: 18.6 to 22.9) percent in the U.S. and 14.2 (90% CI: 11.7 to 18.3) percent in EU5 which translates to 568,624 (90% CI: 522,359 to 619,705) excess infant deaths in the U.S. and 559,370 (90% CI: 469,308 to 694,589) in the combined five European countries. The results should be interpreted with caution because they rely on the assumption of a uniformly decreasing secular trend if there had been no nuclear tests, but this cannot be verified. It is concluded that atmospheric nuclear weapons testing may be responsible for the deaths of several million babies in the Northern Hemisphere.