Two large studies link higher coffee consumption to reduced risk for death

1. Two large studies link higher coffee consumption to reduced risk for death*

For the first time, researchers also studied coffee's effect on white and non-white populations

*Video available for download: HD video soundbites of experts discussing their findings available at http://www.dssimon.com/MM/ACP-coffee.

Abstract: http://annals.org/aim/article/doi/10.7326/M16-2945

Abstract: http://annals.org/aim/article/doi/10.7326/M16-2472

Editorial: http://annals.org/aim/article/doi/10.7326/M17-1503

Free summary for patients: http://annals.org/aim/article/doi/10.7326/P17-9041

Study URLs go live when the embargo lifts

Two new studies published in Annals of Internal Medicine seem to confirm the mortality benefits of higher coffee consumption. The first study found that compared to non-coffee drinkers, those who consume the most coffee have a significantly lower risk for death. The study, conducted in 10 European countries, was the largest ever of its kind. The second study found that higher coffee consumption was associated with lower risk for death in whites and also in non-white populations. This finding is important because different races have different lifestyles and disease risks. The mortality benefit was the same for caffeinated and decaffeinated coffee.

Coffee is one of the most frequently consumed beverages worldwide. About 75 percent of adults in the United States drink coffee, and 50 percent drink it daily. Because of its ubiquitous consumption, understanding coffee's health effects is important.

Researchers from the International Agency for Research on Cancer and Imperial College London used data from EPIC (European Prospective Investigation into Cancer and Nutrition), a large multinational cohort or more than 520,000 men and women from 10 European countries, to compare all-cause and cause-specific mortality in coffee drinkers compared to non-coffee drinkers. They found that participants who reported drinking three or more cups of coffee per day seemed to receive the most benefit in terms of lowering the rate of death. This was particularly true for diseases of the digestive tract, but also for circulatory diseases.

Investigators at the Keck School of Medicine of the University of Southern California sought to determine how coffee consumption affected health across multiple races. The MEC (Multiethnic Cohort) study followed more than 185,000 African Americans, Native Americans, Hawaiians, Japanese Americans, Latinos, and whites for an average of 16 years. They found that higher coffee consumptions were associated with lower risks of death in whites, and also in non-white populations, including African Americans, Latinos, and Japanese Americans. Coffee drinking among non-white communities previously had little research. This study substantially increases the generalizability of previous findings across the racial/ethnic spectrum.

According to the author of an accompanying editorial, the studies show that a protective effect of coffee is biologically plausible. Polyphenols and other bioactive compounds in coffee have antioxidant properties, and coffee intake is associated with reduced insulin resistance, inflammation, and biomarkers of liver function. However, the author cautions that coffee itself is a heterogeneous exposure, and the presumed benefits may depend on components other than caffeine. Generally speaking, coffee intake up to 3 to 5 cups per day can be part of a healthy diet.

Media contact: For an embargoed PDF, please contact Cara Graeff at cgraeff@acponline.org. To speak with Marc Gunter, PhD, author of the EPIC study, contact him directly at GunterM@iarc.fr. To interview Wendy Setiawan, PhD, lead author of the MEC study, contact Zen Vuong at zvuong@usc.edu or 213-740-5277. To reach the editorialist, Eliseo Guallar, MD, DrPH, contact Stephanie Desmon at sdesmon1@jhu.edu.

2. Current performance measures for cervical cancer screening promote overscreening

Distinguishing between overscreening and appropriate screening and adding ranges to screening intervals dramatically change appropriate use

Abstract: http://annals.org/aim/article/doi/10.7326/L17-0140

URL goes live when the embargo lifts

Following current performance measures that require adherence to strict time intervals for cervical cancer screening may promote overscreening. Changing cervical cancer screening performance measures to distinguish between overscreening and appropriate screening and include time ranges rather than strict intervals could help to reduce the frequency of unnecessary procedures. The findings of a brief research report are published in Annals of Internal Medicine.

Current performance measures for cervical cancer screening are based on guidelines from organizations such as the American College of Obstetricians and Gynecologists and require strict adherence to screening intervals. Satisfactory cervical cancer screening involves at least one Pap test every 3 years for average-risk women between the ages of 21 and 64 or at least one Pap and HPV test every 5 years for average-risk women between the ages of 30 and 64. These performance measures have two main flaws: 1) they do not distinguish between overscreening and appropriate screening; and 2) they do not allow ranges of intervals in appropriate screening definitions.

To address these flaws, researchers at the University of Pittsburgh used Pennsylvania Medicaid administrative data for women aged 18 to 64 between 2007 and 2013 to determine how frequently screening practices were adherent to traditional performance measures and to alternative measures that distinguished between overscreening and appropriate screening and incorporated ranges of +/- 3 months and +/- 6 months to appropriate screening definitions. Distinguishing between appropriate screening and overscreening (i.e., creating a separate overscreening category) decreased appropriate screening by up to 70 percent. However, incorporating the ranges increased appropriate screening by up to 11 percent.

The authors conclude that changing cervical cancer screening performance measures to align better with clinical guidelines will help to reduce the frequency of unnecessary procedures and more accurately measure the quality of women's health care.

Media contact: For an embargoed PDF, please contact Cara Graeff at cgraeff@acponline.org. For an interview with the lead author, Natasha K. Parekh, MD, MS, please contact Allison Hydzik at hydzikam@upmc.edu or 412-647-9975.

###

Also new in this issue:

Sensitivity Analysis in Observational Research: Introducing the E-Value

Tyler J. VanderWeele, PhD; Peng Ding, PhD

Research and Reporting Methods

Abstract: http://annals.org/aim/article/doi/10.7326/M16-2607

URL goes live when the embargo lifts

Sensitivity Analysis for Unmeasured Confounding: E-Values for Observational Studies

A. Russell Localio, PhD; Catherine B. Stack, PhD; Michael E. Griswold, PhD

Editorial

Abstract: http://annals.org/aim/article/doi/10.7326/M17-1485

URL goes live when the embargo lifts