Coffee consumption might not be as arbitrary as people may think, according to a study from Lawrence Berkeley National Lab, or Berkeley Lab.

Utilizing a technique called quantile regression, in which data is separated into subgroups and then individually analyzed, Berkeley Lab staff scientist Paul Williams discovered a correlation between how much coffee people drink and the environments they live in. In other words, there is a positive feedback loop between genetics and the environment, according to a Berkeley Lab press release.

“Most of us … are willing to accept that our cholesterol levels, height and weight have strong genetic influences,” Williams said. “Drinking coffee being totally under our free choice? Turns out that’s not quite true.”

Genes that indicate a propensity for coffee consumption will be expressed more strongly in environments where people are more likely to drink coffee, according to the study. If an individual has a genetic trait with a preference for coffee and is in a community where drinking coffee is common, that person is more likely to drink coffee.

The study also notes that coffee intake was significantly different between people of European ancestry and people from other parts of the world. Additionally, male offspring were more likely to have a higher coffee intake than female offspring.

Williams said compared to standard regression, using quantile regression revealed a “synergism” between environmental factors and genetic traits.

“Standard regression assumes that the same line applies to all throughout the distribution, whether the 99th percentile or first percentile,” Williams said. “What if you do away with that restriction? Whether that influences the background (in) which genes operate, in fact it does.”

Williams’ research drew from data collected from the Framingham Heart Study, which the National Institutes of Health, or NIH, launched in 1948 to observe how cardiovascular disease is affected by genes and lifestyle. Williams’ study considered data from 4,788 child-parent pairs and 2,380 siblings — all of whom were related to an original group in Framingham, Massachusetts.

According to Williams, his work is part of a larger effort by the NIH to have outside scientists analyze its data.

Williams added in an email that since his previous research looked at more straightforward correlations, he wanted to see if coffee intake, a behavioral trait, would also be dependent on level of consumption. He said he believed people drank coffee for a variety of reasons, including socialization and health benefits, but that his research was more intended to see how genetics and the environment interact.

“Our genes affect behaviors,” Williams said. “They may not be behaviors specific to something like coffee, but might have to do with the types of foods we pursue.”