Author

Dwyer (1996)⁴

Conclusions

*Comprehensive review of literature showed a linear relationship between weight & health risk.
*States disease virtually guaranteed when BMI is 2-3 times "normal".

Concern

Any risks related to low or decreasing BMI were not addressed.

Authors

Manson, Willett, Stampfer, Colditz, Hunter, Hankinson, Hennekens, & Speizer (1995)⁵

Conclusions

*A positive linear relationship between weight/mortality exists.
*Statistically significant health risks for women when BMI > 27.
*Women should attempt to maintain BMI at or below 21.

Concerns

*Confounding variables including hypertension, diabetes mellitus, hypercholesterolemia, & weight cycling were not controlled in initial data analysis.
*The sample did not represent the target population (American women).
*Self-reported heights/weights were used.
*Increased health risk was demonstrated for BMIs over 27, but BMIs at or below 21 were recommended.

Authors

Troiano, Frongillo, Sobal, & Levitsky (1994)⁶

Conclusions

Meta-analysis of research with over 30 years of follow up:
*Many longitudinal studies do not support a positive linear relationship between weight & adverse health risk.
*Some studies demonstrate a negative relationship.
*Moderately increased BMIs are not associated with increased mortality.
*Weights at or slightly below current recommendations are associated with increased health risks.
*A "J" shaped relationship between weight and health exists.

Concerns

*Risk associated with weight loss may be explained by smoking and/or other comorbid conditions; however, studies that controlled for these variables have not demonstrated the removal of associated health risks