Unemployment reached 23 percent and the GDP shrank by as much as 14 percent, so it's hard to imagine a silver lining to the tumultuous years of the Great Depression. But could the general health of the U.S. population actually have improved when the nation's economic fitness took multiple nosedives? And, if a floundering economy improves longevity, what does this say about our current recession?

It turns out that the bleakest years of the Great Depression, as gauged by GDP and unemployment rate, saw the greatest gains in life expectancy and drops in mortality rates. And during the years that the economy perked up, the nation paid the price in terms of health, according to a study published last week in the Proceedings of the National Academy of Sciences.

To look at the relationship between economic and population well-being, social scientists José Tapia Granados and Ana Diez-Roux of the University of Michigan at Ann Arbor amassed U.S. Census Bureau data on mortality rates, life expectancy, unemployment and GDP for each year from 1920 to 1940. "What this [study] does is to look in detail at data that have now been available for some years but have not been looked at in detail," Tapia says.

First the authors compared the annual changes in life expectancy and mortality rate to annual changes in the unemployment rate and GDP growth for the 20-year period. When they plotted life expectancy over these years on a graph with the economic indicators, a negative correlation between life expectancy and economic growth became obvious. The greatest gains in years of life expectancy coincided with the years of economic recession, 1921 and 1938, and depression, from 1929 to 1933, whereas there were actually longevity setbacks in prosperous years, such as 1926 and 1936.

Similarly, there was a positive correlation between mortality rates and economic upturn. When the variables of age and sex were taken into account, the authors saw the same trend to varying degrees. Although mortality rates dropped or hit a plateau in recessionary years and climbed in years of economic recovery for all groups, the changes in life span were more pronounced for males, people of middle to old age and infants under one.

A trend that endures
"What they're finding for the Depression is what a number of us have found for mild recessions in recent years," says Christopher Ruhm, a professor of economics at the University of North Carolina at Greensboro's Bryan School of Business and Economics, who has studied the health impacts of recessions occurring in the past 50 years. "I don't know if I would have expected the results to hold for the Great Depression, but given that it did, it's not surprising."

This inverse relationship may seem counterintuitive because a strong economy suggests that more people would have access to health care. But, as Tapia points out, "people who have studied the effect of health care on population health are in general not inclined to think that health care has a big impact." Instead, the lifestyles and stresses that accompany a bullish market march up the death rates.

When they have incomes allowing them to indulge, "people tend to drink more, [and] tend to be overweight and obese during periods of economic expansion," Tapia says. In addition, the jobs in the Depression era caused stress, demanding long hours with little vacation and, in some cases, were inherently hazardous.

In keeping with these potential health risks, Tapia's study found that deaths due to cardiovascular disease, which was the major killer in the 1920s and '30s just as it is today, peaked with national affluence. Not surprisingly, the number of injuries related to automobiles, a luxury many people did without during the Depression, also swelled in the years of economic recovery. The authors did not examine the deaths due to dangerous work conditions, but Tapia says earlier studies have found workplace injuries also increased during high employment but, compared with heart attacks, they were a relatively minor cause of mortality, overall.

Deaths due to influenza and pneumonia also went down during recessionary years, possibly because people had less of the kind of close physical contact with each other that transmits the flu virus.

What about unemployment stress?

Although employment could lead to stresses and lifestyle differences that increase the risk of cardiovascular disease, the staggering unemployment during the Depression would have probably introduced intense stress. "When we're talking about a garden-variety recession," U.N.C. Greensboro's Ruhm says, "stress related to employment is greater than unemployment [stress]. I would have thought, [however,] that this collapse would have created a huge level of stress."

In fact, the number of suicides did increase during years of economic contraction. Out of the half dozen different causes of death that the researchers analyzed, it was the only one that was responsible for more fatalities during recessionary than prosperous years. "The stress of being jobless," Tapia says, "could be leading people to suicide." For every 100,000 people, the number of suicides rose from about 13 in 1927 and 1928 to about 17 during the years of the Depression, and dropped down to 14 in 1933, when the economic situation improved. But it was still a minor cause of death compared with the number of cardiovascular-related deaths, about 400 for every 100,000 people.

As low as the unemployment rate got, the situation did not actually spiral into starvation. "My feeling is that, in spite of many people having economic stress, probably nutrition didn't go to levels low enough at the national level as to cause generalized undernutrition," Tapia says.

The role of relief spending
One of the factors that could have spared the nation from widespread malnutrition was relief spending, says Price Fishback, a professor of economics at the University of Arizona's Eller College of Management in Tucson, who researches the effects of relief spending on population health. During these 20 years, there were local programs, although the extent of aid varied from city to city and state to state. And, when it did step in, "the federal government kind of totally overwhelmed everything," Fishback says, boosting direct financial relief to the unemployed and offering work programs.

Federal relief does not explain the inverse relationship between life expectancy and economic health that Tapia found, however, because life expectancy was on the rise starting in 1929, and the economic programs under the umbrella of President Franklin Roosevelt's New Deal did not begin until 1933. It is hard to say the part that state and local relief may have played in preventing malnutrition because Tapia and Diez Roux did not look at regional statistics. Fishback says that more work should be done to explore the effect of relief spending during this period on health indicators at the national and local levels.

The fact that Tapia sees an inverse relationship between economic and population health for the Great Depression similar to what has been found for lesser recessions suggests that this relationship applies even in very severe economic times. "A few months ago, I was asked if recent [health] results would hold up if we have a really severe collapse," Ruhm says. "These results suggest that even if it had got much more severe than it did, we would have seen health improvements."

Some complicating factors for today
The many safety net reforms and laws passed since the Depression era suggest that the peaks and valleys on a life expectancy and mortality rate graph would not be as sharp. Not the least of which is our welfare program. Today, we have economic insulators in place, such as unemployment insurance, that did not exist during the Depression. "If we look at countries with stronger social safety nets, you tend to see these effects being more muted," Ruhm points out.

Even though government support has expanded, Tapia says that social support has slipped in recent history. For one, the average size of the U.S. household is smaller now than in the 1920s and '30s. Also, a 2006 study in American Sociological Review found that the average person now has a smaller number of people in whom they could confide than folks typically did 20 years ago. Greater isolation among U.S. citizens could make us more vulnerable to economic stresses, and thereby to greater peaks and valleys in health, Tapia says, citing a body of research showing that people who are integrated in their communities tend to enjoy a greater degree of protection against premature mortality.

It could be years before we will know exactly how the nation's health is being altered by the current recession, although not quite as long as it took to get a better idea of the relationship during the Depression. Usually social scientists and economists have to wait five to 10 years after the period they want to examine for all the statistical data to become available. But, based on the recurring pattern seen for recessions over the past century, we already have a pretty good idea. Our health will have improved, at least somewhat, as, alas, our financial statements grow more pallid.