Short sleepers who are also obese screen positive for depression at rates over 50% higher than adequate sleepers. But the relationship between weight, sleep, and mental health doesn't run in the direction most people expect.

Two findings from NHANES physical exam and questionnaire data stand out. The weight category with the worst depression outcomes isn't the heaviest. And sleep duration is a stronger depression predictor than BMI category, even though obesity gets most of the attention in that conversation.

Underweight Adults Have the Highest Depression Rates

Look at depression prevalence by BMI category and the first surprise is at the bottom of the scale, not the top. Underweight adults (BMI < 18.5) screen positive for clinically significant depression (PHQ-9 ≥ 10) at 23%, the highest of any weight category and more than double the 10.4% rate among overweight adults.

It's U-shaped, not linear. Obese adults show elevated rates compared to normal-weight adults (14.6% vs. 11.2%), but the overweight category actually has the lowest depression prevalence, lower even than normal weight.

Only 92 underweight individuals appear in the sample, so that 23% carries wider confidence intervals than the estimates for larger groups. Still, it's a striking outlier that the obesity-focused framing of weight and depression tends to overlook entirely. Anyone drawing a clean line between higher BMI and worse mental health should look at this table first.

Sleep Duration Is a Stronger Signal Than BMI

Adults sleeping fewer than 7 hours on weeknights screen positive for depression at 17.5%, compared to 10.1% among adequate sleepers. Average PHQ-9 scores reinforce the gap: 5.0 for short sleepers versus 3.52 for those getting 7 to 9 hours, a 42% difference.

A PHQ-9 average of 5.0 across an entire population subgroup is not mild. That score sits at the clinical boundary, and when the average is that high, a large share of the group is scoring well above it. Nearly 1 in 5 short sleepers crosses into moderate-to-severe territory.

Now compare sleep to weight. Obese adults screen positive at 14.6%, elevated relative to normal weight but still three full points below the 17.5% short-sleep rate. If you had to choose one variable to predict depression risk between BMI category and sleep duration, sleep wins.

When Both Risk Factors Overlap, the Signal Doesn't Multiply

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 lower than the short-sleep population as a whole (17.5%).

The obese short-sleeper group (n = 518) is a subset of all short sleepers. If depression rates are particularly high among non-obese short sleepers, the combined group's rate would land below the short-sleep average. The data doesn't break that out directly, but the numbers fit that interpretation.

Stacking obesity on top of short sleep doesn't produce a dramatically amplified depression signal. The 15.4% rate sits in essentially the same range as both parent groups. Whatever mechanism connects poor sleep to depression appears to operate largely independent of weight.

That leaves an underexplored question: among people who sleep poorly, are non-obese adults actually at higher depression risk than obese ones? If the underweight group's 23% prevalence holds across survey cycles, the answer might reframe how we think about the weight-depression connection entirely.