Download Changes in Three Diet Quality Scores and Total and Cause

Changes in Three Diet Quality
Scores and Total and CauseSpecific Mortality
Mercedes Sotos Prieto, PhD
Shilpa N Bhupathiraju, Josiemer Mattei, Teresa T Fung, Yanping Li, An Pan,
Walter C Willett, Eric B Rimm, Frank B Hu
Department of Nutrition, Harvard T.H. Chan School of Public Health, Boston, MA
Mercedes Sotos Prieto, PhD
Changes in Three Diet Quality Scores and Total and
Cause-Specific Mortality
 FINANCIAL DISCLOSURE:
No relevant financial relationship exists
 UNLABELED/UNAPPROVED USES DISCLOSURE
None
DISCLOSURE INFORMATION
• Fats: MUFA,
SFA
• Mg
• Fruit,
Vegetables
• Nuts, meat
Single nutrients
or foods
Dietary
patterns
• Mediterranean diet
• Healthy eating pattern
• DASH pattern
• Vegetarian pattern
Diet quality
scores
AMED
AHEI
DASH
Cespedes EM, Hu FB.. Am J Clin Nutr 2015;101:899-900.
BACKGROUND
The Alternative Healthy Eating Index-2010 (AHEI) score
• Based on recommendations for food and nutrient consumption with
• Current scientific evidence of beneficial health effects
The Alternative Mediterranean diet (AMED) score
• Comprised of foods and nutrients characteristic of the Mediterranean
Pattern
The Dietary Approach to Stop Hypertension (DASH) score
• Developed from the DASH dietary recommendations aiming to reduce
blood pressure
AHEI: Chiuve et al., 2012. J Nutr. AMED Fung TT et al., 2009. Circulation. DASH: Fung TT et al., 2008 Arch Intern Med.
BACKGROUND
 Consuming the healthiest quality diet assessed by AHEI,
AMED and DASH:
8% to 22% of all-cause mortality
 Dietary Patterns Methods Project (DPMP): 4 dietary
patterns, 3 prospective cohorts
18–26% lower all-cause mortality
19–28% lower risk of CVD mortality
11–23% lower risk of cancer mortality
George SM et al, Am J Epidemiol 2014; Harmon BE et al, Am J Clin Nutr 2015; Reedy J et al. J Nutr 2014. Sofi F et al. Public Health Nutrition 2013
BACKGROUND
 Healthy dietary patterns as practical way for the
public to understand and act upon.
 Few studies have evaluated changes in diet quality
over time in relation to mortality risk.
 Diet and other lifestyles occur over the lifetime and
may influence health outcomes and mortality risk, it
is important to evaluate how much impact improving
the diet might have on subsequent mortality risk.
BACKGROUND
Health Professionals Follow-up Study (51,529 men, aged 40 to 75 years )
1986
1988
1990
MD+LS MD+LS MD+LS
1992
1994
MD+LS
MD+LS
1998
MD+LS MD+LS
Diet
Diet
Diet
1996
2000
........
2010
MD+LS
MD+LS
MD+LS
Diet
Diet
Nurses’ Health Study (121,701 women, aged 30 to 55 years )
1976
MD+LS
1978
1980
1982
MD+LS MD+LS MD+LS
Diet
1984
1986
1988
1990
1992
1994
MD+LS MD+LS
MD+LS
MD+LS
MD+LS
MD+LS MD+LS
Diet
Diet
Diet
Diet
Repeated measurements over time:
• Every Two Years: Medical records and lifestyle (MD+LS)
• Every Four Years: Detailed dietary habits (FFQ)
1996
.......
2010
MD+LS MD+LS
Diet
1
Evaluate the association between 12-y changes (19861998) in three diet quality indices assessed by the AHEI,
the AMED, and the DASH with total and cause-specific
mortality in the NHS and HPFS from 1998-2010
1986
1990
1986
1990
1994
1998
2002
2006
2010
12y change
1994
1998
2002
2006
2010
Diet
Subsequent 12-year risk of mortality risk
AIMS
2
We also examined shorter- and longer-term changes
in diet quality and total and cause-specific mortality.
8 y change
1986
1990
1986
1990
1994
1994
1998
1998
2002
2002
2006
2006
2010
2010
Diet
Subsequent 16y mortality risk
16 y change
1986
1990
1986
1990
1994
1994
1998
1998
2002
2002
2006
2006
2010
2010
Diet
Subsequent 8 y mortality risk
AIMS

Using 1986 as baseline for both cohorts, with follow
up until 2010

Exclusions:
 CVD history and cancer at baseline
 Missing information on diet and other lifestyle
covariates.
 Implausible total daily energy intake (i.e., men <800 or
>4200 kcal/day; women <500 or >3500 kcal/day).
 Participants who died before 1998.
 23,100 men in the HPFS and 44,501 women in the
NHS
METHODS
Components
AHEI-2010
AMED
DASH
Population
based median
0-9
Population
based quintiles
8-40
Vegetables (excluding potatoes) s/d
Fruit, s/d
Whole grains, g/d
Sugar-sweetened beverages, s/d
Nuts and legumes, s/d
Red and processed meat, s/d
Fish, s/d
Low fat dairy, s/d
Trans fat, % of energy
Long-chain (n-3) fats (EPA + DHA), mg/d
PUFA, % of energy
MUFA:SFA ratio
Sodium, mg/d
Alcohol, drinks/d
Scoring
Total range
A priori cutoff
0-110
S=serving
BACKGROUND
 Cox proportional hazard models with time-varying covariates and age as the
underlying time scale were used to estimate HR and 95% CI.
 Model 1: age, initial diet quality score, race, family history of MI, diabetes, and
cancer, aspirin use, multivitamin use, initial BMI category, menopausal status and
hormone use in women, and baseline and simultaneous changes in other lifestyle
factors: smoking status, physical activity, and total energy intake.
 Model 2: Model 1 + history of hypertension, hypercholesterolemia or type 2
diabetes, weight change
 20-percentile increase in each score was calculated from the median value of
each quintile
 Pooled analysis: inverse, variance-weighted meta-analysis model, accounting for
heterogeneity between studies
METHODS
During 758,683 person-years of follow-up we documented 9,772 deaths, including 2,292
CVD deaths and 3277 cancer deaths.
Nurses’ Health Study
Q1
Q3
Health Professional Follow-up Study
Q5
Q1
Q3
Q5
No. of participants
Initial diet score
(Largest
(Relatively
decrease) no change)
8900
8901
59.1 (10.0) 50.0(9.5)
(Largest
increase)
8900
44.1(9.5)
(Largest
decrease)
4620
60.6(10.6)
(Relatively
no change)
4620
52.1(10.5)
(Largest
increase)
4620
46.2(10.1)
Changes in diet score
Age, years
Initial BMI, Kg/m2
Weight change, Kg
Changes in physical activity
-11.1 (5.0)
65.5 (7.1)
25.5 (4.6)
10.5 (18.1)
0.1 (2.3)
3.0(1.4)
63.6(7.0)
25.4(4.7)
8.3(16.0)
0.4(2.3)
17.2(5.2)
62.6(6.7)
25.0(4.6)
6.1(15.9)
0.5(2.4)
-10.7(5.0)
65.2(9.4)
25.4(3.0)
7.4(14.3)
1.0(4.7)
3.3(1.4)
64.1(9.0)
25.2(3.0)
5.5(12.8)
1.4(3.8)
17.3(5.1)
63.8(9.0)
25.1(3.0)
3.5(12.4)
1.7(4.4)
Initial alcohol intake, g/d
5.8 (8.4)
6.0(10.4)
7.4(13.0)
11.4(12.9)
11.4(15.0)
12.3(17.5)
Changes in alcohol intake,
g/d
Total energy intake, Kcal/d
-0.3 (8.2)
-1.2(7.4)
-2.4(10.2)
-0.5(10.8)
-2.5(13.5)
1742 (521)
1777(516)
1773(513)
1972(604)
2029(613)
1981(571)
Change in total energy intake, -11.6 (527)
kcal/d
-46.0(496)
-72.6(502)
59.1(597)
-10.3(559)
-24.6(574)
2.0(13.1)
RESULTS
12-year changes (1986-1998) in diet quality scores and risk of total mortality
Q1
Q2
Q3
Q4
Q5
P-trend
0.94
(0.88, 1.00)
0.89
(0.83, 0.95)
<.0001
0.92
(0.82, 1.03)
0.83
(0.65, 1.05)
<.0001
0.88
(0.83, 0.94)
<.0001
Alternative Healthy Eating Index (range 0-110)
MV-adjusted model2
1.14
(1.07, 1.22)
1.06
(0.99, 1.13) 1 [Ref]
Alternate Mediterranean Diet (range 0-9)
MV-adjusted model2
1.08
(0.97, 1.19)
0.97
(0.91, 1.03) 1 [Ref]
Dietary Approach to Stop Hypertension (range 8-40)
MV-adjusted model2
1.10
(1.03, 1.16)
1.02
(0.91, 1.15) 1 [Ref]
0.92
(0.86, 0.99)
Improvement in diet quality (14-33%)
11-17% total mortality
Decrease in diet quality (9-22%)
8-14% total mortality
Abbreviations: MV, multivariable; Multivariable-adjusted model adjusted for age (in month),
Model2=age, initial diet quality score , race, family history of MI, diabetes, and cancer, aspirin use, multivitamin use, initial body mass index and simultaneous changes in other
lifestyle factors: smoking status and initial and changes (all in quintiles) in physical activity and total energy intake and menopausal status and hormone use in women. history of
hypertension, hypercholesterolemia and type 2 diabetes, and weight change (quintiles) during the 4-year period. For DASH additionally adjusted for change and initial alcohol intake
(in quintiles). Results for NHS and HPFS from the multivariate model were combined with the use of the random-effects model. P>0.05 for heterogeneity between women and men
in all categories of diet change categories.
RESULTS
12-year changes (1986-1998) in diet quality scores and total mortality, CVD mortality
and cancer mortality per 20-percentile of increase in each score (calculated from the
median value of each quintile )
Overall
mortality
CVD
mortality
Cancer
mortality
0.80
(0.76, 0.85)
0.84
(0.74, 0.94)
0.93
(0.84, 1.03)
0.91
(0.88, 0.94)
0.94
(0.88, 0.99)
0.96
(0.90, 1.02)
0.87
(0.83, 0.91)
0.92
(0.85, 1.02)
0.90
(0.83, 0.98)
AHEI (range 0-110)
MV model
A 20% increase was associated:
AMED (range 0-9)
MV model
DASH (range 8-40)
MV model
9-20% total mortality
6-16% CVD mortality
4-10% Cancer mortality
Maintaining a higher adherence:
- 23-24% lower total mortality
- 19-27% lower risk of CVD deaths
- 13-20% lower risk of cancer death
RESULTS
Hazard Ration (95%CI) of total
mortality
shorter- and longer-term changes
in diet quality
0.5
AHEI (range 0-110)
AMED (range 0-9)
DASH (range 8-40)
8y
12y
16y
8y
12y
16y
8y
12y
16y
Changes Changes Changes Changes Changes Changes Changes Changes Changes
The association was strengthened when
longer changes were evaluated
1. INTRODUCCIÓN
1. INTRODUCCIÓN
Limitations
Strengths
 Measurement error and
misclassification from selfreported dietary information.
 Prospective population-based
design
 Individual components
 A large sample size
 High rates and long follow-up
 White nurses and health
professionals could limit the
generalizability of the results.
 Overall quality diet, multiple diet
quality indices
 Residual confounding
 Changes in diet
STRENGTHS/LIMITATIONS
 Improving adherence to ANY of the three diet quality
scores over 12 y is associated with significantly lower risk
of mortality, CVD and cancer mortality.
 Longer term changes in diet strengthened the association
 Maintenance of your diet quality over time as an adult can
have a meaningful effect on risk of total mortality, CVD and
cancer death.
 These results underscore the importance of the strategies
to promote and sustain a healthy diet in improving longevity
among middle-aged and older adults.
CONCLUSION
 As an example, a person increasing ~22 points (20%)
out of 110 for the AHEI score over a 12-year lower
the risk of total mortality by 20%
 Emphasizing even small dietary changes should be
an important part of nutrition and public health
policies.
What can you do to promote healthy eating and translate the public
health message? As a community, population level, as a politician? As
health policies?
 Department of Nutrition, Harvard T.H. Chan School of Public Health,
Boston, MA
 Coauthors:
Shilpa N Bhupathiraju, PhD
Josiemer Mattei, PhD, MPH
Teresa T Fung, ScD
Yanping Li, PhD
An Pan, PhD
Walter C Willett, MD, DrPH
Eric B Rimm, ScD
Frank B Hu, MD, PhD
 Fundacion Alfonso Martin Escudero, Spain
ACKNOWLEDGEMENTS
THANK YOU
Mercedes Sotos Prieto, PhD
Shilpa N Bhupathiraju, Josiemer Mattei, Teresa T Fung, Yanping Li, An Pan,
Walter C Willett, Eric B Rimm, Frank B Hu
Department of Nutrition, Harvard T.H. Chan School of Public Health, Boston, MA
1.4
12 years later Low
12 years later Medium
1.4
12 years later High
12 years later Medium
12 years later High
1.2
1
*
*
*
*
*
0.8
Hazard Ratio CVD mortality
Hazard Ratio total mortality
1.2
12 years later Low
*
1
*
0.8
0.6
*
*
0.6
0.4
0.4
0.2
0.2
0
0
Medium
High
Low
AHEI
Baseline AHEI score, Pooled
Medium
High
AHEI
Baseline AHEI score, Pooled
1.4
12 years later Low
12 years later Medium
12 years later High
1.2
Hazard Ratio cancer mortality
Low
1
*
*
0.8
0.6
0.4
0.2
0
Low
Medium
High
AHEI
Baseline AHEI score, Pooled
RESULTS