One, almost classic, example for abrupt reversals in life expectancy relates to the changes associated with the alcohol reforms of the Union of Soviet Socialist Republics in the Gorbachev era, which reversed a negative trend in life expectancy into a positive one following a reduction in the supply of alcohol (increase in life expectancy from 1984 to 1987: men, 3.2 years; women, 1.3 years [5, 6]). When the restrictive alcohol policy was abandoned and alcohol became widely available, the trend in life expectancies reverted once again (decrease in life expectancy between 1987 and 1994: men, 7.3 years; women, 3.3 years [5, 7]). Obviously, these figures represent associations between availability and consumption of alcohol with life expectancy rather than a proof of causality, however, at least for the first reversal, it would be hard to find other explanations, as the increases in life expectancy occurred during a time of economic crisis with no plausible alternative explanations. For the negative trend reversal, attribution of causality is less clear, as this change took place in a period of massive privatization (e.g. [8]) and economic decline, with resulting unemployment (e.g. [9, 10]), community destabilization, subsequent psychological stress (e.g. [11]), and increasing inequality [12]. Some of these variables were included in the most comprehensive econometric modeling on the Russian experience of rapidly decreasing life expectancy [7]. Moreover, alcohol-attributable causes of death were especially impacted in both reversals for life expectancy [5, 7]. Finally, more recent examples of the effects of alcohol policies on mortality confirmed the importance of alcohol policies in the Russian case [13, 14]. In terms of mechanisms, the underlying policy interventions to reduce mortality were mainly alcohol availability restrictions and taxation [5, 13], the latter only for more recent interventions.

The second example concerns drug policy, and specifically policies for regulating prescription opioids (POs) in the United States (US). Overall, larger quantities of POs are used in the US than in any other country (in the latest available statistic for 2011–2013, Canada was a distant second, with about 60 % of the consumption per capita of the US [15]). The rapid increase of PO use and misuse commenced in the mid-1990s, in part by allowing family doctors to prescribe short-acting opioids like oxycodone for relatively common disease categories such as chronic pain [16, 17]. With increased availability of POs, non-medical use increased proportionally along with their associated harm, such as overdose deaths (albeit with some lag) [16, 18–20], contributing to a reversal of all-cause mortality of middle-aged white non-Hispanics in the US [21]. While the mortality trend for the general population was steadily decreasing prior to 1999, between 1999 and 2013 mortality increased by 9 % in white non-Hispanics, but continued to decrease in black non-Hispanics and Hispanics. Again, this is an association, but the increase in mortality for white non-Hispanic middle aged Americans was largely accounted for by increasing death rates from drug and alcohol poisoning deaths, suicide, and chronic liver diseases and cirrhosis, all directly or indirectly associated with substance use and substance use policies (drug and alcohol poisoning deaths per definition; suicide [22]; liver cirrhosis [23]; causal links for illicit drugs in general [24]). POs played the major role in this mortality mix, as prescription overdose deaths have become the most prevalent form of overdose death in the past decade in North America, accounting for approximately 40 % of the total drug poisoning deaths [25–27]. The rise in heroin overdose deaths in recent years can also be partly attributed to use initiated by way of previous PO use [28].

The third example, also arising from illicit drug policies, relates to the systemic consequences of substance use policies (for a definition see [29]). A recent paper showed that, after six decades of gains in life expectancy for Mexico, the trend stagnated for the period following the year 2000 and, for men after 2005, it actually reversed [30]. This reversal of trends in life expectancy was mainly caused by an unprecedented rise in homicide rates, in large part linked to illicit drugs and the war on drugs, i.e. linked to gang wars and/or conflicts between drug gangs and police or the army [31]. In general, enforcement of prohibitive drug laws has been shown to impact adversely on drug market violence in a systematic review of the evidence [32], and alternative regulatory models will be required if drug supply and drug market violence are to be meaningfully reduced [33].