To “Live long and prosper,” a greeting popularized by Mr. Spock from the TV series Star Trek, is unfortunately not the reality for many Americans. New evidence suggests that your zip code provides a strong indication for your life expectancy and that across America the differences in life expectancy are  as big as 20 years.

According to a new report in JAMA Internal Medicine, inequalities in life expectancy among counties are large and growing. In over 3,000 U.S. counties from 1980 to 2014, researchers found that, positively, life expectancy at birth for both men and women increased from 73.8 years to 79.1 years overall. However, there’s a significant gap in life expectancy between counties with the lowest life expectancy and those with the highest: 20.1 years. They call this growing gap “geographic inequality.”

Axious reports on the best and worse counties and the gaps are stark:

Best and worse places: Central Colorado counties had the highest life expectancy (87 years) in 2014. North and South Dakota counties encompassing Native American reservations had the lowest– 66 years.

Where things got worse: Life expectancy grew 5.1 years overall but in some places it actually fell between 1980 and 2014. 8 of the 10 counties with the largest declines in life expectancy are in Kentucky.

Differences in the risk of death for children and teens between different geographic areas also shrank – good news for kids. This means kids in different zip codes increasingly have a better shot at life. However, the gap widened and increased among older adults. Aging Americans increasingly face a different life expectancy depending on where they live.

Three factors combined explain almost 75 percent of county-level variation in life expectancy: socio-economic and race/ethnicity, behavioral and metabolic, and access to health care. When broken down, we learn that personal behaviors such as smoking and lack of exercise as well as body composition explain 74 percent of life expectancy variations. That is followed by poverty, level of education, unemployment, and race explaining about 60 percent of county-level variances and access to health care factors explaining another 27 percent. We should not take this to mean that your race or your employment status will guarantee how long you live though. Your behaviors can actually mitigate these risk factors.

The papers concludes that the policy implications are important. Public health experts have said that our zip codes are better predictors of our health than even our genetic code. Public policy can be targeted at not just mitigating health issues (such as providing a healthcare card), but other factors that affect your health such as your education and employment. It’s critical then that we see national economic growth which lowers poverty levels by lifting people up by getting them to work to provide for themselves and their families.

Poverty-tackling programs can come alongside those efforts and provide a temporary safety net to catch those who have fallen. However, promoting good health and behavioral changes is also a strong partner in this fight. Let’s be careful, though, about trying to accomplish this by those pesky mandates such as banning sugar and trans-fat bans that take choice away from Americans, and which are often proven to be ineffective, if not downright counterproductive, a few years later.