Do likes pull harder on people who are already struggling?
Yes. In a study of more than 17 million posts from 7,736 Twitter/X users, people with depression ramped up their posting more sharply after getting likes than people without depression did. The pattern showed up in all three separate groups of users the researchers studied.
That result is the opposite of what most psychiatrists have been taught. For decades, lab studies have shown that depression dulls the response to reward. Give someone with depression a small win in a computer task, a bit of money or a point on a screen, and they usually push the button less eagerly than someone who is not depressed. Doctors call this blunted reward learning, and it is one of the standard explanations for why depression drains the pleasure out of ordinary life.
Social media appears to break that rule. A like is a tiny piece of social approval that arrives at unpredictable times, and it seems to grip the people who are struggling most.
What the data show
The researchers at Princeton did not run a lab experiment. They watched what people actually did online. For each user, they lined up the number of likes received on one day against the amount that person posted the next day, then looked at how steeply posting climbed as likes climbed. That slope is a measure of how strongly a person is reinforced by social feedback.
Users with a depression diagnosis or higher scores on depressive symptom questionnaires had a significantly steeper slope. More likes on Monday meant a bigger jump in posting on Tuesday, and the effect was stronger for them than for users without depression. Because the finding repeated across three independent datasets rather than appearing once, it is much harder to dismiss as a fluke of one sample.
Dr. Kumar’s Take
I find this study unsettling in a useful way. We have spent years telling patients that depression flattens their response to reward, and then handing them a phone that delivers a reward every few minutes. This paper suggests the two facts do not cancel out. Social validation may be one of the few reinforcers that still lands hard when everything else feels muted, which would explain why some of my patients describe scrolling as the only thing that “does anything.”
I want to be careful here. This is observational work built on public posting data, so it shows a pattern, not a cause. We cannot say from this that likes make depression worse, or that posting less would make anyone feel better. What we can say is that the standard lab picture of blunted reward in depression does not carry over cleanly into real online life, and that gap matters.
How strong is the evidence?
The scale is the strength. More than 17 million posts is far beyond what any lab study of reward learning can collect, and the behavior is real behavior rather than button presses for points. Replicating across three separate datasets is the part that convinces me most, because a single dataset can carry quirks of who signed up and when.
The limits are real too. Depression was established through diagnosis records or symptom questionnaires depending on the dataset, and neither is a live clinical assessment at the moment of posting. Twitter/X users are not a random slice of the population, so the findings may not transfer to other platforms or to people who mostly lurk. And a day-to-day link between likes and posting cannot tell us which way the arrow points over months or years.
What this means day to day
If you have depression, the practical read is not “delete everything.” It is that your own feedback loop may be steeper than you assume, and that noticing the loop is the first step to controlling it. Posting more after a good response is normal. Posting much more, and feeling flatter on the days the response does not come, is the pattern this study describes.
Practical takeaways
- Notice whether a strong response to one post pulls you into posting more for the rest of the day, since that escalation is the exact pattern this study measured.
- If your mood tracks the response to your posts, try turning off like counts or notifications for a week and see whether the pull weakens.
- Do not treat reduced posting as a treatment for depression, because this study shows an association in behavior and was never designed to test whether cutting back helps symptoms.
- Bring your social media use up with your doctor or therapist the way you would bring up sleep or alcohol, as a habit that interacts with your mood rather than a side issue.
Related studies and research
- Cutting back on social media eased loneliness in anxious teens
- Mindfulness-based cognitive therapy for major depression in students: dose-response, inflammation and BDNF
- Internet-based CBT self-help: long-term depression benefits proven
- Wim Hof method for women with high depressive symptoms: a randomized trial
FAQs
Does this mean social media causes depression?
No. This study looked at how people already diagnosed with depression respond to likes, not at whether likes brought the depression on. The design tracks one day against the next within the same person, which is good for spotting a feedback loop and poor for settling questions of cause. It is entirely possible that depression changes how people respond to social approval, rather than social approval creating depression. Sorting that out needs studies that follow people from before symptoms start.
Why do lab studies say depression blunts reward when this study says the opposite?
The rewards are not the same thing. Lab tasks usually offer small amounts of money or abstract points on a fixed schedule, delivered by a machine to a person sitting alone. A like is social, it comes from a specific person, it arrives unpredictably, and it says something about how you are seen. Depression involves a heavy focus on social standing and rejection, so a reward made of social approval may reach a system that money and points do not. That is a plausible reading of this result, though the study measured behavior rather than brain response.
Should someone with depression quit social media entirely?
There is no evidence here to support that as a blanket recommendation. This study measured a reinforcement pattern and did not test any intervention, so it cannot tell you whether quitting improves symptoms. For many people, especially those who are isolated, online contact is a real source of connection, and cutting it off removes support along with the loop. A more measured approach is to change the feedback rather than the contact, by muting counts and notifications while keeping the conversations.
How would I know if likes are driving my behavior?
Look for escalation and for the crash. The signal in this study was that a good day of likes led to noticeably more posting the next day, so watch whether your own output climbs after a post lands well. The companion sign, which people usually feel more clearly, is the dip on days when a post gets little response. Writing down your posting count and your mood for two weeks makes the link visible in a way that memory does not, because memory tends to keep the good days and drop the flat ones.
Bottom line
Across more than 17 million posts from 7,736 users, people with depression escalated their posting more steeply after receiving likes than people without depression, and the pattern held in three independent datasets. That overturns the tidy assumption that depression simply blunts the response to reward. Social approval delivered on a screen appears to be a strong reinforcer for exactly the people who are struggling most, which is worth knowing before we tell patients that nothing feels good anymore.
