TY - JOUR
T1 - Two genetic analyses to elucidate causality between body mass index and personality
AU - Arumäe, Kadri
AU - Briley, Daniel
AU - Colodro-Conde, Lucía
AU - Mortensen, Erik Lykke
AU - Jang, Kerry
AU - Ando, Juko
AU - Kandler, Christian
AU - Sørensen, Thorkild I.A.
AU - Dagher, Alain
AU - Mõttus, René
AU - Vainik, Uku
N1 - Publisher Copyright:
© 2021, The Author(s), under exclusive licence to Springer Nature Limited.
PY - 2021
Y1 - 2021
N2 - Background/objectives: Many personality traits correlate with BMI, but the existence and direction of causal links between them are unclear. If personality influences BMI, knowing this causal direction could inform weight management strategies. Knowing that BMI instead influences personality would contribute to a better understanding of the mechanisms of personality development and the possible psychological effects of weight change. We tested the existence and direction of causal links between BMI and personality. Subjects/methods: We employed two genetically informed methods. In Mendelian randomization, allele scores were calculated to summarize genetic propensity for the personality traits neuroticism, worry, and depressive affect and used to predict BMI in an independent sample (N = 3 541). Similarly, an allele score for BMI was used to predict eating-specific and domain-general phenotypic personality scores (PPSs; aggregate scores of personality traits weighted by BMI). In a direction of causation (DoC) analysis, twin data from five countries (N = 5424) were used to assess the fit of four alternative models: PPSs influencing BMI, BMI influencing PPSs, reciprocal causation, and no causation. Results: In Mendelian randomization, the allele score for BMI predicted domain-general (β = 0.05; 95% CI: 0.02, 0.08; P = 0.003) and eating-specific PPS (β = 0.06; 95% CI: 0.03, 0.09; P < 0.001). The allele score for worry also predicted BMI (β = −0.05; 95% CI: −0.08, −0.02; P < 0.001), while those for neuroticism and depressive affect did not (P ≥ 0.459). In DoC, BMI similarly predicted domain-general (β = 0.21; 95% CI:, 0.18, 0.24; P < 0.001) and eating-specific personality traits (β = 0.19; 95% CI:, 0.16, 0.22; P < 0.001), suggesting causality from BMI to personality traits. In exploratory analyses, links between BMI and domain-general personality traits appeared reciprocal for higher-weight individuals (BMI > ~25). Conclusions: Although both genetic analyses suggested an influence of BMI on personality traits, it is not yet known if weight management interventions could influence personality. Personality traits may influence BMI in turn, but effects in this direction appeared weaker.
AB - Background/objectives: Many personality traits correlate with BMI, but the existence and direction of causal links between them are unclear. If personality influences BMI, knowing this causal direction could inform weight management strategies. Knowing that BMI instead influences personality would contribute to a better understanding of the mechanisms of personality development and the possible psychological effects of weight change. We tested the existence and direction of causal links between BMI and personality. Subjects/methods: We employed two genetically informed methods. In Mendelian randomization, allele scores were calculated to summarize genetic propensity for the personality traits neuroticism, worry, and depressive affect and used to predict BMI in an independent sample (N = 3 541). Similarly, an allele score for BMI was used to predict eating-specific and domain-general phenotypic personality scores (PPSs; aggregate scores of personality traits weighted by BMI). In a direction of causation (DoC) analysis, twin data from five countries (N = 5424) were used to assess the fit of four alternative models: PPSs influencing BMI, BMI influencing PPSs, reciprocal causation, and no causation. Results: In Mendelian randomization, the allele score for BMI predicted domain-general (β = 0.05; 95% CI: 0.02, 0.08; P = 0.003) and eating-specific PPS (β = 0.06; 95% CI: 0.03, 0.09; P < 0.001). The allele score for worry also predicted BMI (β = −0.05; 95% CI: −0.08, −0.02; P < 0.001), while those for neuroticism and depressive affect did not (P ≥ 0.459). In DoC, BMI similarly predicted domain-general (β = 0.21; 95% CI:, 0.18, 0.24; P < 0.001) and eating-specific personality traits (β = 0.19; 95% CI:, 0.16, 0.22; P < 0.001), suggesting causality from BMI to personality traits. In exploratory analyses, links between BMI and domain-general personality traits appeared reciprocal for higher-weight individuals (BMI > ~25). Conclusions: Although both genetic analyses suggested an influence of BMI on personality traits, it is not yet known if weight management interventions could influence personality. Personality traits may influence BMI in turn, but effects in this direction appeared weaker.
U2 - 10.1038/s41366-021-00885-4
DO - 10.1038/s41366-021-00885-4
M3 - Journal article
C2 - 34247202
AN - SCOPUS:85110546567
VL - 45
SP - 2244
EP - 2251
JO - International Journal of Obesity
JF - International Journal of Obesity
SN - 0307-0565
ER -