Thanks to Epigenetics, the 1998 Canadian Ice Storm Imprinted in Babies' Genes

In January of 1998, a huge ice storm hit a small region of Canada, in eastern Ontario, southern Quebec, and across to Nova Scotia, covering everything in a layer of ice. The weight of that ice toppled power lines and utility poles, crushing the power grid. Three million people in the province of Quebec were without power for as many as 6 weeks. Commuters were stranded in subway cars, temperatures dropped to -20°C, and no one had any power to heat their homes or their water. The storm was responsible for 27 deaths, and cost $3 billion in total losses. As you can imagine, this had a significant impact on the population, including, apparently, the next generation.

This was one of Canada's worst natural disasters, which prompted researchers at McGill University to look at the effects of in utero exposure to maternal stress. They recruited 176 women who were pregnant during the ice storm, or who conceived within 3 months (when stress levels would still be elevated). They were asked about their stress levels using a 32 point questionnaire, which assessed the degree of objective hardship, loss, scope (e.g., the number of days without power), threat, and change (e.g., time in a shelter). Subjective distress was also measured. Thirty-six youths from that study agreed to participate in a follow-up, providing blood samples for epigenetic testing at the age of 13. Twenty-eight of these kids were already incubating during the ice storm, and 8 were conceived shortly after.

By looking at the DNA methylation patterns in these kids, the authors found that the mothers' elevated stress levels during the 1998 ice storm was correlated with specific DNA methylation pattern in 957 of the kids' genes. These genes are predominantly related to immune function. Interestingly, subjective stress, or the woman's perceived stress and hardship during the event, was not associated with DNA methylation patterns in their children. There was a link between maternal subjective stress and anxiety, depression, and aggression in the kids, but it was the objective stress level that had more of an effect on their cognitive ability, IQ, language development, and physical outcomes like obesity and insulin secretion (a risk factor for diabetes).

DNA methylation "heatmap" showing similar patterns across
individuals. Each column represents a person, and each row a
DNA methylation site. Areas of lowmethylation are shown in
green, and those of high methylation are shown in red.

This study is the first to show the epigenetic effects based on both objective and subjective maternal stress following a natural disaster. Pre-natal maternal stress is normally hard to look at in humans, because it's hard to separate out genetic and psychosocial factors. But events like natural disasters have an acute, independent, and randomly distributed onset, which mimics the experimental controls we would normally see with animal studies.

The authors hope that, as the longitudinal study continues, they will be able to look at the relationship between DNA methylation and actual gene expression (mRNA levels). They also hope that similar studies, with a larger sample size, will provide information about the cause-and-effect relationship of pre-natal maternal stress and DNA methylation. Without collecting DNA samples at birth, the researchers can't know for sure if these epigenetic modifications occurred in utero, or if they emerged in response to downstream postnatal stress.