Short sleepers who are also obese screen positive for depression at a rate 48% higher than the general population of adequate sleepers — and the obesity-depression relationship itself doesn't run in the direction most people expect.

Key Numbers

• 23% of underweight adults screen positive for clinically significant depression (PHQ-9 ≥ 10) — more than double the rate among overweight adults (10.4%)
• 14.6% of obese adults screen positive, versus 11.2% among normal-weight adults
• Adults sleeping fewer than 7 hours average a PHQ-9 score of 5.0, compared to 3.52 among adequate sleepers — a 42% gap
• 17.5% of short sleepers hit the clinical threshold for depression, versus 10.1% of those sleeping 7–9 hours
• 15.4% of adults who are both obese and short sleepers screen positive for moderate-to-severe depression

The Obesity-Depression Gradient Isn't Linear — Underweight Adults Carry the Heaviest Burden

The standard narrative links obesity to depression. The data complicates that story immediately. Underweight adults (BMI < 18.5) have a depression prevalence of 23% — the highest of any BMI category, and more than twice the 10.4% rate among overweight adults. Obese adults do show elevated rates compared to normal-weight adults (14.6% vs. 11.2%), but the gradient across the four categories is U-shaped, not linear.

The underweight group is small — 92 individuals in the sample — so that 23% figure carries more uncertainty than the estimates for larger groups. Still, it's a striking outlier that the obesity-centric framing of the depression-weight relationship tends to erase entirely.

The overweight category actually has the lowest depression prevalence in the dataset — lower even than normal-weight adults. That finding alone should give pause to anyone drawing a clean line between higher BMI and worse mental health outcomes.

Short Sleep Is a Stronger Depression Signal Than Obesity Alone

Sleeping fewer than 7 hours on weeknights is associated with a 73% higher rate of clinically significant depression compared to adequate sleepers — 17.5% versus 10.1%. The average PHQ-9 score gap (5.0 vs. 3.52) is consistent with that prevalence difference: short sleepers aren't just slightly more symptomatic, they're scoring meaningfully higher across the distribution.

For context, a PHQ-9 score of 5 sits at the lower boundary of mild depression. An average score of 5.0 across an entire population subgroup means a substantial share of that group is scoring well above 5. The 17.5% clinical threshold rate confirms that: nearly 1 in 5 short sleepers crosses into moderate-to-severe territory.

Compare that to the obesity-alone signal: obese adults screen positive at 14.6%, which is elevated but still below the short-sleep rate of 17.5%. Sleep duration, at least in this dataset, is a sharper discriminator of depression risk than BMI category.

When Obesity and Short Sleep Overlap, Depression Rates Rise — But Not as Much as the Individual Signals Predict

Adults who are both obese and short sleepers screen positive for depression at 15.4%. That's higher than obese adults overall (14.6%) but — and this is the counterintuitive part — it's lower than the short-sleep population as a whole (17.5%).

One arithmetic explanation: the obese short-sleeper group (n = 518) is a subset of the short-sleep population. If depression rates are particularly high among non-obese short sleepers, the combined group's rate could land below the short-sleep average. The data doesn't break that out directly, but the numbers are consistent with that structure.

What the data does show clearly: stacking obesity and short sleep doesn't produce a dramatically amplified depression signal beyond what short sleep alone predicts. The 15.4% figure is essentially in the same range as both parent groups. Whether that reflects a ceiling effect, overlapping mechanisms, or something about the specific composition of this subgroup is a question the cross-sectional snapshot can't resolve.

Open Questions

1. Among short sleepers, do non-obese adults actually have higher depression rates than obese short sleepers — and if so, what accounts for that pattern?
2. The underweight group's 23% depression prevalence is the dataset's highest — is that driven by a small number of high-severity cases, or is it a stable signal that holds across different NHANES cycles?
3. Sleep duration here is self-reported weekday sleep: does the depression-sleep association look different when measured against actigraphy-based sleep data or when weekend sleep is included?