Lichtman · The New England journal of medicine 1992 · controlled observational study · n=224

Discrepancy between self-reported and actual caloric intake and exercise in obese subjects.

Cited 1274 times in the scientific literature.

Level 4 - case-series / case-control

Controlled observational study comparing a small clinical subgroup to controls

PubMed 1454084 · doi:10.1056/NEJM199212313272701 · record verified 2026-08-29

What was done

Researchers evaluated potential mechanisms of self-reported diet resistance (failure to lose weight despite reporting <1200 kcal/day) among 224 consecutive obese individuals presenting for treatment. Ten subjects with a history of diet resistance (Group 1: 9 women, 1 man) underwent 14-day evaluations of actual caloric intake, total energy expenditure, and thermogenic components using indirect calorimetry and body composition analysis. Comparisons were made against control subgroups drawn from 80 subjects without a history of diet resistance (Group 2: 67 women, 13 men).

What was found

Total energy expenditure and resting metabolic rate in Group 1 were within 5% of predicted values for body composition, and thermic effects of food and exercise did not differ significantly between groups. Subjects in Group 1 underreported actual food intake by an average of 47 ± 16% (mean ± SD) and overreported physical activity by 51 ± 75%. Compared with Group 2, Group 1 subjects were significantly more likely to perceive a genetic cause for their obesity, use thyroid medication, and describe eating behavior as relatively normal (all P < 0.05).

Why it matters

This study establishes that self-reported resistance to dietary weight loss is primarily due to substantial underreporting of energy intake and overestimation of physical activity, rather than metabolic or thermogenic defects.

Limits

The detailed metabolic and caloric intake evaluations were conducted in only 10 diet-resistant subjects, of whom only one was male. All participants were treatment-seeking individuals at a clinical center, potentially introducing selection bias.

Cited by