Still, research looking for a direct relationship between social media and well-being has not found much.

“There’s been absolutely hundreds of these studies, almost all showing pretty small effects,” said Jeff Hancock, a behavioral psychologist at Stanford University who has conducted a meta-analysis of 226 such studies.

What is notable about the new study, said Dr. Hancock, who was not involved in the work, is its scope. It included two surveys in Britain totaling 84,000 people. One of those surveys followed more than 17,000 adolescents ages 10 to 21 over time, showing how their social media consumption and life-satisfaction ratings changed from one year to the next.

“Just in terms of scale, it’s fantastic,” Dr. Hancock said. The rich age-based analysis, he added, is a major improvement over previous studies, which tended to lump all adolescents together. “The adolescent years are not like some constant period of developmental life — they bring rapid changes,” he said.

The study found that during early adolescence, heavy use of social media predicted lower life-satisfaction ratings one year later. For girls, this sensitive period was between ages 11 and 13, whereas for boys it was 14 and 15. Dr. Orben said that this sex difference could simply be because girls tend to hit puberty earlier than boys do.

“We know that adolescent girls go through a lot of development earlier than boys do,” Dr. Orben said. “There are a lot of things that could be potential drivers, whether they’re social, cognitive or biological.”

Both the boys and girls in the study hit a second period of social media sensitivity around age 19. “That was quite surprising because it was so consistent across the sexes,” Dr. Orben said. Around that age, she said, many people go through major social upheaval — like starting college, working in a new job or living independently for the first time — that might change the way they interact with social media, she said.