Genetic obesity risk fails to predict short-term weight loss, study finds

Kanagawa University of Human Services-led researchers found short-term BMI reductions after an eight-to-12-week low-carbohydrate diet plus resistance-training program. BMI moved downward across the program while a derived efficiency score predicted BMI change more strongly than the genetic risk score alone.

Obesity affects more than 640 million people worldwide and prevalence continues to rise in developed and developing nations. Medical consequences of obesity include type 2 diabetes, hypertension, dyslipidemia, cardiovascular disease, cerebrovascular disease, malignant neoplasms, and mortality risk that increases with higher BMI.

Lifestyle factors recognized in obesity risk include physical activity, diet, and sleep. Standard weight management efforts often combine dietary guidance with increased physical activity, while outcomes vary across individuals even with structured programs.

Multiple influences beyond daily habits have been suggested, including developmental influences through maternal exposure during fetal development, endocrine-related diseases, medication effects, psychological effects, socioeconomic influences such as infrastructure and public transportation networks, and racial influences.

Genetics are also considered a possible source of that variation with factors that can modify the magnitude of environmental effects on obesity.

Diet and exercise vs. gene scores

In the study, “Genetic Risk Scores for Obesity and the Effectiveness of a Diet and Exercise Intervention Program: A Historical Cohort Study,” published in Obesity, researchers retrospectively analyzed data to examine genetic risk scores for obesity and to evaluate an efficiency score for predicting outcomes of a low-carbohydrate diet and resistance-training program.

Analyses compared genetic risk scores calculated from 75 SNPs identified in a previous Japanese genome-wide association study with an efficiency score generated using data envelopment analysis, then assessed both measures as predictors of change in BMI and body fat percentage using linear regression.

Participants came from RIZAP Corporation locations in the Kanto region of Japan. Eligibility included males and females aged 20–69 years who completed an eight-to-12-week program combining a low-carbohydrate diet with resistance training. Individuals with certain conditions were excluded as having potential sources of natural weight loss that could bias results.

Analytic inclusion totaled 145 participants, including 121 females and 24 males, with a mean age of 46.0 years. Primary outcome focused on weight loss measured through BMI. Secondary outcomes included change in body weight and body composition from baseline to completion.

The program

Diet targets set carbohydrates at about 50 g daily and protein at 1–2 g per kg of body weight, with no specific fat intake restrictions. Energy content used 4 kcal per g for carbohydrates and protein and 9 kcal per g for fat.

Resistance training covered 16 supervised sessions delivered at one or two sessions per week during the intervention period. Each 50-min session included bench presses, lat pulldowns, dead lifts, crunches, leg raises, and squats, using two to three sets of 10–15 repetitions for each exercise with 1-min rest intervals between sets. Trainers adjusted resistance loads based on participant performance and feedback to maintain a challenging stimulus, described as generally aiming for moderate-to-vigorous intensity appropriate for muscle hypertrophy and strength gains.

Genetic analyses came from saliva collected using a DNA kit after recruitment. Genetic risk score used 75 SNPs selected from a Japanese genome-wide association study. The range for the score ran from −2.194 to 2.334, with higher values indicating a stronger genetic predisposition to obesity.

Weight loss patterns and predictors

BMI fell from 27.6 kg/m² at baseline to 23.8 kg/m² at program end. Weight fell from 71.6 kg to 61.9 kg. Body fat percentage fell from 36.1% to 28.4%.

Systolic blood pressure fell from 123.5 mmHg to 117.0 mmHg. Diastolic blood pressure fell from 79.7 mmHg to 75.3 mmHg. Estimated bone mass fell from 2.6 kg to 2.5 kg. Muscle mass fell from 41.5 kg to 40.4 kg. Mean efficiency score measured 0.81.

BMI change tracked closely with body weight change, a correlation coefficient (r) of 0.98 (with r numbers near 1 indicating two measures moved together).

Body fat percentage change correlated with body weight change, with r=0.81. Body fat percentage change correlated with BMI change, with r=0.83.

Genetic risk score correlated with baseline BMI, with r=0.23. Genetic risk score correlated with baseline body fat percentage, with r=0.26. Efficiency score correlated with baseline BMI, with r=−0.97. Efficiency score correlated with baseline body fat percentage, with r=−0.78.

Models using genetic risk score explained little variation in outcomes. Genetic risk score did not significantly predict BMI change rate. Genetic risk score showed a statistically significant association with body fat percentage change rate, with low explanatory power.

Models using efficiency score explained more variation than models using genetic risk score. Adjusted R2 for BMI change rate reached 32.1% in a crude model and 34.3% after adjustment for age, sex, and program frequency. Adjusted R2 for body fat percentage change rate reached 7.4% in a crude model and 8.0% after adjustment.

Sex-stratified regression reported efficiency score prediction of BMI percentage change in males and females. Efficiency score predicted body fat percentage change in males, with a non-significant result reported in females. Genetic risk score showed non-significant associations with outcomes in both sexes in the sex-stratified analyses.

Conclusions

Researchers conclude that weight loss occurred regardless of genetic risk score derived from cross-sectional BMI genome-wide association results. Efficiency scores showed greater explanatory power for predicting intervention outcomes than genetic risk score alone.

© 2026 Science X Network