Surely you’ve seen the headlines:
“Family Meals Help Teens Avoid Smoking, Alcohol, Drugs”
“Family meals may cut teens' drug use”
“The family that dines together, fends off the pitfalls of drug and alcohol abuse better”
The stories tend to crop up every September because of Family Day, a holiday invented by the National Center on Addiction and Substance Abuse (CASA) at Columbia University. CASA is responsible for the report “The Importance of Family Dinners,” which it revises to great fanfare every few years.
For a while, these headlines made me feel smug. I don’t have a teen, but I have a three-year-old, and my wife and daughter and I sit down for a home-cooked dinner seven times in a typical week.
After the smugness wore off, I got annoyed. I felt the way I imagine wine aficionados felt when the press began reporting that red wine is good for you: they are trying to take something I enjoy for its own sake and turn it into medicine. “Try these family dinners — they’ll put hair on your chest and keep your kids off the hard stuff,” they seemed to be saying.
So I turned skeptical.
The most recent CASA report, “The Importance of Family Dinners III,” was released in September 2006. I read it and came away wondering: Was anything I’d heard reported about this study really true?
For the study, CASA interviewed teens aged 12 to 17 and their parents, and produced a substance-abuse-risk score for the teens. Answering yes to questions like “Have you tried marijuana?” raised a teen’s substance-abuse score. Then CASA correlated the substance-abuse score with how often the teens were home for dinner. Here’s the key result:
|Dinners per week||Substance-abuse risk|
To give an idea of what these numbers mean: The teens who dined at home only 0 to 2 times per week were twice as likely to have tried pot and more than twice as likely to have tried cigarettes. Moms, start your Crock-Pots; this case is closed. Right?
Let’s turn back to my own family. My daughter is three. We eat dinner together every night. I can say with reasonable confidence that she has never used drugs and doesn’t hang out with those who do, although some of her friends eat a lot of Annie's cheesy bunnies.
The CASA study didn’t include 3-year-olds, of course. But it did include 12-year-olds, and the table compares 12-year-olds against 17-year-olds.
Presumably I don’t need to explain the various ways in which 17-year-olds are different from 12-year-olds, or convince you that 17-year-olds are both less likely to be at home at dinnertime and more likely to try drugs. Any time you’re reporting data on a variable that is known to be strongly influenced by age (cancer, for example), the right thing to do is report age-adjusted statistics.
In case this sounds like blather, here’s another way to think about why age-adjusting is important. Imagine a study that compared cancer rates at a nursing home and a preschool, found that the nursing home had 100 times the cancer rate, and concluded that something must be wrong with the nursing home.
Or try this: Teens who eat at home more often are shorter, and teens who eat at home less often are taller. But that doesn’t mean you can get Junior onto the basketball team by keeping him away from the family table.
Nothing in CASA’s report is presented as being age-adjusted. On the last page of the report is a statement that CASA used a statistical technique (analysis of variance, aka ANOVA) to verify that the reported differences still existed after adjusting for age. Great. But a representative of the firm that conducted the study for CASA admitted to the Wall Street Journal in 2005 that the difference in substance-abuse risk is more strongly correlated with age than with family-dinner attendance.
(I did not catch this statistical gaffe myself. I read about it in Cameron Stracher’s book, Dinner with Dad. In the book, Stracher, an overworked law professor, promises to be home for dinner five nights a week and finds, at least at first, that this leads to more family friction.)
So yes, there is indeed a gap in drug-abuse risk between the frequent home diners and the infrequent home diners — it’s just much smaller than CASA makes it out to be. Still, doesn’t this gap somehow prove that family dinners help prevent drug abuse, even if the effect isn’t as big as initially reported?
No. Which of these hypotheses sounds more likely to you?
Probably both are true, right? Elizabeth Planet, CASA’s director of special projects, agrees.
“The statistical relationships we observe do not demonstrate causality — indeed, there is probably some back and forth in the dinners/family relationships/teen attitudes/teen behavior,” Planet wrote to me in an email. “Our findings are consistent with other social-science research regarding the protective benefits of parental engagement and strong family relationships.”
That is not how the report is sold in September.
“If I could wave a magic wand to make a dent in the substance-abuse problem, I would make sure that every child in America had dinner with his or her parents at least five times a week. There is no more important thing a parent can do,” states CASA president Joseph Califano in the introduction to the report. (If he had an actual magic wand, wouldn’t he use it to just get rid of drugs altogether? Just wondering.)
A cottage industry has grown up around promoting the family dinner. Among the presentations given at the “Family Meals: Setting the Table” conference held last year in Philadelphia was one called “State of Dinner.” The presentation defined dinner as “Meal eaten in the evening, preferably cooked with Campbell’s products.”
Other sponsors of the event were Safeway, Coca-Cola, Kroger, and Del Monte. Another talk was entitled “Family Meals: Good for the Food Industry, Good for the Family, Good for Nutrition.”
CASA’s report is sponsored by the Nick at Nite and TV Land cable channels. In the introduction, Califano thanks Nick at Nite/TV Land president Larry W. Jones, who “understands that a revival of the family dinner in America will do more to curb kids from smoking, drinking, and using drugs than any law or public health campaign.”
Presumably Jones also understands that in a 2000 study of family meals published in the Journal of Nutrition Education, 52.8 percent of teens reported frequent television viewing during meals.
Could family dinners actually be bad for teens?
For teens with attention deficit hyperactivity disorder (ADHD), the answer may be yes. As Dr. Lily Trokenberg Hechtman writes in her book, ADHD in Adulthood :
The idea that all families should have a peaceful family dinner is an expectation that families with ADHD often come to with heartbreak. . . . [A] family dinner requires preparation. This often means leaving children unsupervised in the early evening when they are unmedicated.
Of course, Hechtman isn’t saying that families with ADHD should shun cooking. “Families with ADHD discover alternative patterns that can realistically work for them,” she writes. “They may allow family members to eat in smaller groups that are less overwhelming.”
Matthew Amster-Burton sniffs out the unexplained in the kitchen, the store, and the food world at large. He blogs at Roots and Grubs, podcasts at Spilled Milk, and is the author of the book Hungry Monkey.
Want more? Comb the archives.
A father’s legacy
The vegetarian-cooking pioneer
Barbecue, tamales, cocktails, and more
Good on everything