|Nutrition Evidence Library|
Kelemen LE, Kushi LH, Jacobs DR Jr, Cerhan JR. Associations of dietary protein with disease and mortality in a prospective study of postmenopausal women. Am J Epidemiol. 2005;161(3):239-249.
PubMed ID: 15671256
Prospective cohort study
B - Click here for explanation of classification scheme.
Research Design and Implementation Rating:
NEUTRAL: See Research Design and Implementation Criteria Checklist below.
· To investigate the associations of different protein sources with chronic disease and mortality in postmenopausal Iowa women
· Completed a 16-page questionnaire mailed in January 1986
· Women who were premenopausal at baseline · Reported a history of cancer other than skin cancer, known heart disease, or known diabetes · Left 30 or more food items blank · Had total energy intake less than 600 kcal/day or more than 5,000 kcal/day
· Women who were premenopausal at baseline
· Reported a history of cancer other than skin cancer, known heart disease, or known diabetes
· Left 30 or more food items blank
· Had total energy intake less than 600 kcal/day or more than 5,000 kcal/day
Description of Study Protocol:
· Eligible women between the ages of 55 and 69 years were randomly selected from the Iowa driver’s license registry.
· 15-year follow-up prospective cohort study from 1986-2000
Blinding used (if applicable)
Intervention (if applicable)
· Macronutrients were expressed as a percentage of total energy, and other dietary covariates were energy adjusted by the regression method.
· The distribution of potential confounding and risk factors were examined by quintiles of total protein intake.
· Continuous variables were categorized into quintiles and treated as indicator variables in statistical models following inspection of their relation with each outcome in univariable analysis.
· Risk ratios (RR) and 95 percent confidence intervals (CI) were calculated using Cox regression.
· Survival was modeled as a function of age using as the referent the lowest quintile of protein intake.
· The relation between dietary protein and each outcome was assessed with multivariable-adjusted nutrient density models.
Data Collection Summary:
Timing of Measurements
· In 1986, the questionnaire inquired about factors known or suspected to be related to cancer, including smoking, physical activity, postmenopausal hormone use, alcohol use and anthropometric measurements.
· Baseline diet was assessed with a semiquantitative food frequency questionnaire (SFFQ). The validity and reliability of the SFFQ was documented in reference 18.
· Supplementary questionnaires were mailed in 1987, 1989, 1992 and 1997 to establish vital status and change of address.
· Incident cases of cancer were ascertained through the State Health Registry of Iowa.
· Deceased non-respondents were identified through linkage with the National Death Index.
· CHD mortality, cancer incidence and mortality, all-cause mortality
· Dietary protein
· Known risk factors related to cancer
· Potential dietary confounding factors
Description of Actual Data Sample:
Initial N: 99,826 were randomly selected; N=41,836 (41.9%) responded
Attrition (final N): 29,017 for the analyses indicating 31% dropout rate
Age: Participants were between the ages of 55 and 69 years at baseline in 1986
Ethnicity: older Caucasian women
Other relevant demographics: Whether groups were significantly different on age and education were not described.
Anthropometrics: Whether groups were significantly different on waist/hip ratio and BMI were not described.
Summary of Results:
· 4,843 new cancers, 739 CHD deaths, 1,676 cancer deaths and 3,978 deaths from all causes were identified during 15 years of follow up.
· Among women in the highest quintile, CHD mortality decreased by 30% from an isoenergetic substitution of vegetable protein for carbohydrate (95% CI: 0.49, 0.99) and of vegetable for animal protein (95% CI: 0.51, 0.98), following multivariable adjustment.
· Although no association was observed with any outcome when animal protein was substituted for carbohydrate, CHD mortality was associated with red meats (RR=1.44, 95% CI: 1.06, 1.94) and dairy products (RR=1.41, 95% CI: 1.07, 1.86) when substituted for servings per 1,000 kcal of carbohydrate foods.
· Dietary proteins from animal and vegetable food sources appear to be differentially associated with mortality from CHD and all causes when substituted for carbohydrates in the diet.
· Long-term adherence to high-protein diets, without discrimination toward protein source, may have potentially adverse health consequences.
The dropout rate (31%) is high, and specific follow-up methods of handling withdrawals were not described in this study.
Copyright American Dietetic Association (ADA).