Your Instagram can predict how lonely you are

Instagram is often seen as a place where people carefully curate their lives, sharing happy moments, flattering photos and an idealized version of themselves. But new research from the University of California San Diego suggests that social media isn't all smoke and mirrors. Even when people may be presenting their best selves online, their everyday behavior can reveal clues about how they really feel.

In a study published in npj Mental Health Research, the authors analyzed Instagram data, finding that different types of loneliness can be detected through different patterns of Instagram behavior. 

For example, people experiencing emotional loneliness—the feeling that they lack a close, meaningful connection, even when surrounded by others—tend to have Instagram accounts with more late-night posts, particularly between 11 p.m. and 5 a.m., more negative language in their captions and profile photos showing sadder or angrier facial expressions.

They also tend to make more references to family and friends, perhaps reflecting a longing for closer connections.

Across multiple measures of loneliness, researchers also found that study participants reporting greater loneliness tended to have fewer faces in their photos and choose profile photos with their heads tilted upward rather than facing the camera directly. Another indicator was the use of more past-focused language, such as "was" and "had," potentially reflecting a tendency to dwell on the past. 

Instagram is one of the world's largest social media platforms, with more than 3 billion monthly active users worldwide, making it a major part of how people connect and share their lives online.

The authors found that Instagram behavior was useful in predicting how lonely students said they felt, but much less effective at predicting the size of the real-life support networks they reported.

“Instagram really reflects how lonely people feel, but not necessarily how many people they can turn to in real life,” said Sherry (Yueyi) Jiang, PhD, lead author of the study, who earned her doctorate in experimental psychology from the UC San Diego School of Social Sciences in 2025.

The research offers new insight into loneliness, a growing public health concern associated with increased risks of depression, anxiety and premature death. It also raises questions about whether technology could eventually help identify people experiencing loneliness before it develops into more serious mental health challenges.

Not All Loneliness Is the Same

The researchers distinguish between two types of loneliness: emotional and social.

Emotional loneliness involves feeling that you're missing a close, meaningful connection or don't have someone you can turn to in difficult times. It can involve a deep sense of emptiness and emotional distress, even when someone is surrounded by others.

This form of loneliness is particularly concerning because previous research has found that emotional loneliness is more strongly associated with depression and anxiety. 

The other type referenced in the study—social loneliness—involves feeling disconnected from a broader circle of friends or community. Someone experiencing social loneliness might have a best friend or partner but still feel like they don’t have a group to belong to.

“Loneliness isn't a one-size-fits-all experience, and neither are the solutions,” said Piotr Winkielman, a professor of psychology at UC San Diego. “Someone who feels disconnected from a broader community may benefit from more opportunities to socialize, while someone who lacks a close emotional connection may need a very different kind of support.” 

The study found that these two types of loneliness were associated with different patterns of Instagram behavior.

Emotional Loneliness: Late Nights and Negative Language

Students reporting greater emotional loneliness tended to have Instagram accounts with:

• Fewer photos featuring human faces, a finding associated with loneliness more broadly.
• More past-focused language, such as "was" and "had," potentially reflecting a tendency to dwell on the past.
• More posts late at night, particularly between 11 p.m. and 5 a.m.
• More negative language in their captions.
• Profile photos showing sadder or angrier facial expressions.
• More references to family and friends, perhaps reflecting a longing for close connections.

Social Loneliness: Less Engagement With Others

Students reporting greater social loneliness tended to have Instagram accounts with:

• Fewer photos featuring human faces.
• Profile photos with their heads tilted upward rather than facing the camera directly, a pattern also associated with other measures of loneliness.
• More past-focused language, such as "was" and "had," potentially reflecting a tendency to dwell on the past.
• Less frequent liking and commenting on other people's posts.
• Less posting on weekends.
• Fewer words expressing care or appreciation, such as "thank" and "care."
• Fewer communication-related words, such as "said" and "say," potentially reflecting fewer references to interactions with others.

Looking Beyond the Public Instagram Profile

Unlike many previous studies that relied on publicly available social media posts, the UC San Diego researchers analyzed the full Instagram histories of 415 undergraduate students, including information that is typically private. All participants provided informed consent, and their data were analyzed with privacy protections in place.

The students, who came from a variety of majors, volunteered to participate in exchange for course credit. They downloaded their Instagram archives and shared them with the researchers, providing access to photographs, captions, comments, likes and other activity dating back to when they first created their accounts.

More than 60% of the students had private accounts, making this study particularly distinctive. Rather than focusing on influencers with stylized content intended to build a personal brand, the researchers were able to examine how everyday Instagram users interact with friends and share their lives online, including activity that outsiders normally can't see.

The research team used machine learning to identify patterns across these different types of information and compared them with students' responses to established measures of loneliness. Students also answered questions about how many friends and relatives they regularly interacted with and could turn to for support.

“What's remarkable is that we can identify patterns associated with something as personal and difficult to observe as loneliness,” said Jiang. “By looking at photos, language and everyday activity together, we can pick up on subtle clues that might otherwise go unnoticed.”

The researchers found that Instagram activity provided moderately accurate predictions of students' reported loneliness, but was much less effective at identifying the size and availability of the real life support networks they reported.

Could Instagram Help Identify Loneliness?

The researchers say their findings suggest that digital behavior could eventually help identify people experiencing loneliness, including those who might be reluctant to talk about their feelings.

Winkielman compares the potential to wearable health technologies that can detect irregular heartbeats and alert users to possible health concerns.

"What if your Instagram account could suggest it's time to talk to someone?" he said.

Such technology could potentially help universities, counselors and health professionals identify early warning signs and offer support tailored to different forms of loneliness.

Privacy presents another significant challenge. Any future applications would require informed consent, safeguards against misuse and careful consideration of the consequences of incorrectly identifying someone as lonely.

However, the researchers emphasize that their findings represent patterns across groups, not reliable indicators for diagnosing individuals. The study also cannot establish whether loneliness changes how people use Instagram, whether Instagram use contributes to loneliness, or both.

Still, the study offers a new way of understanding the relationship between people's digital lives and their emotional well-being.

The study's other coauthors are Kevin Lewis, professor of sociology; Barton W. Palmer, professor emeritus of psychiatry; and Fei Dai, a research assistant in the Department of Psychology, all at UC San Diego.  

Read the full study, “Everyday Instagram Behavior Maps Distinct Facets of Loneliness in Young Adults,” published in npj Mental Health Research.