omega-3 fatty acid/fish intake and brachial artery flow-mediated dilation

Download Omega-3 Fatty Acid/Fish Intake and Brachial Artery Flow-Mediated Dilation

Post on 23-Feb-2016




0 download

Embed Size (px)


in the Multi-Ethnic Study of Atherosclerosis. Omega-3 Fatty Acid/Fish Intake and Brachial Artery Flow-Mediated Dilation. Jennifer S. Anderson, MD, PhD, MEd Cardiology Fellow/Research Fellow. Background. - PowerPoint PPT Presentation


Omega-3 Fatty Acid/Fish Intake and Brachial Artery Flow-Mediated Dilation

Omega-3 Fatty Acid/Fish Intake and Brachial Artery Flow-Mediated Dilation

Jennifer S. Anderson, MD, PhD, MEdCardiology Fellow/Research Fellow

in the Multi-Ethnic Study of Atherosclerosis

1BackgroundMultiple studies have shown an inverse relationship between fish/omega-3 fatty acid consumption and cardiovascular events/deathOne potential mechanism includes omega-3 fatty acids impact on endothelial function Flow-mediated dilation (FMD) is a noninvasive surrogate for endothelial functionFew studies have examined the relationship between dietary fish and FMD(mention in the last point: furthermore, few studies have examined this relationship among men and women of mixed race/ethnicity.)2Study HypothesesConsumption of non-fried fish foods rich in long-chain omega-3 fatty acids (eicosapentanoic acid (EPA), docosahexanoic acid (DHA)) will be positively associated with larger FMD across ethnic groups.

Omega-3 rich (EPA, DHA) dietFlow-mediated dilation (FMD)

2.Concentrations of EPA and DHA in plasma phospholipids will demonstrate positive associations with FMD.[PRESENT TENSE STATEMENT OF HYPOTHESES]-And, as others did show associations between other brachial artery measures and CV outcomes, we also explored the relationship b/t fish and brachial measures other than FMD, including baseline artery diameter and maximal diameter3

Cross-sectional analysis of MESA participantsn=6,814 men and women aged 45-85 years, free of CVDMulti-center participation, 6 U.S. sitesBaseline data collection including:120-item food frequency questionnaireFMD measures (n = 3,137)plasma phospholipid/fatty acid measures (n = 1,642)Study Design

4Exposure Measures: Dietary Fish

Non-fried fishtuna, salmon, sardines, sashimi or sushi; other broiled, steamed, baked, or raw fish including trout, sole, halibut, poke, grouper Shellfish (non-fried)shrimp, lobster, crab, oysters, mussels

Fish in mixed dishes(non-fried)fish in enchiladas, burritos, quesadilla, pasta, stew, gumbo, paella, or salad

Fried fishfried fish or fish sandwich, fried shrimp, calamari

-for measurement of our exposure variables, a FFQ was used that included a total of 10 items that relate to fish, as listed here [2, 1, 6, 1][-highlight that top row = n-3 rich group]

5Exposure Measures: Plasma EPA and DHA

6Outcome: Brachial Artery Measures

Baseline brachial artery diameterMaximum brachial artery diameter

-our outcome measures included brachial artery measures -in essence, this is performed using ultrasound technology and a standard blood pressure cuff -the cuff is positioned around the right arm, 2 inches below the antecubital fossa, and the brachial artery is imaged -baseline images are obtained, the cuff is inflated to 50 mmHg above the participants systolic blood pressure for 5 minutes, and images are captured continuously for two minutes to document the vasodilator response, ie maximum diameter. -% FMD is thus computed from these measures by subtracted baseline from maximum brachial artery diameter, dividing by baseline diameter, and multiplying by 100-For our analysis, our main interest was the relationship b/t fish/plasma n-3 fatty acids and FMD, but we also chose to explore the relationship between fish/n-3 plasma fatty acids and other brachial artery measures including baseline diameter, as others have recently demonstrated associations between baseline artery diameter and incident cardiovascular events.

7Multivariate linear regressionStepwise approachFinal model included:

Stratified by gender, race/ethnicity

StatisticsAgeGenderRace/ethnicitySmokingEducationIncome levelPhysical activitySystolic & diastolic BPLipid levelsBMI-used stepwise to identify list of potentially impt confounders from a large list of potential covariates-our final model included-As other data have suggested differences in FMD by both gender as well as race/ethnicity, we also chose to perform analyses with stratification by these covariates

[age, gender, ethnicity, smoking, alcohol consumption, education, income level, physical activity, SBP, DBP, diabetes, medication use, lipid levels, BMI, n-6 PUFA intake, alpha-linolenic acid intake, vitamin C, E, and folate supplement use, n-3 fatty acid supplement use, and consumption of polyphenolic-rich foods initially examined]


Non-fried Fish & Brachial Measures: Overall Results, Baseline Artery Diameter

(n = 642) (n = 943) (n = 838) (n = 714)*p < 0.001

-bottom line here was that there was no sig association with FMD and non-fried fish, which was quite clear in the unadj model but also apparent in the adj model-NOTHING FOR FMD-(say why only adjusted from here on out at the end of this slide)because we feel that understanding the relationships b/t these variables . Only appropriate after adjustment for covariates EVERYTHING ELSE IS ADJUSTED FROM HERE ON OUT

9Non-fried Fish & Brachial Measures: Baseline Diameter, Stratification by Gender*MenWomen*p = 0.02(n = 338) (n = 455) (n = 411) (n = 344)(n = 304) (n = 488) (n = 427) (n = 370)p = 0.10*adjusted for age, race, BMI, smoking, diabetes, SBP, DBP, total:HDL, exercise, income, education, and HRT (women)When stratification by gender was performed, again a significant inverse relationship was noted between NFF and baseline diameter, but this time it was in men only, and in fact women showed a trend toward the opposite effect as illustrated here tho results were NS for women.-you may also note that participant numbers were becoming progressively smaller within each of these subgroups (point out in X axis), which may limit our ability to detect significant associations within subgroups10Non-fried Fish & Brachial Measures: FMD, Stratification by Gender*MenWomenTest for gender-fish interaction significant (p = 0.008 and 0.05 for baseline diameter & FMD, respectively)(n = 304) (n = 488) (n = 427) (n = 370)(n = 338) (n = 455) (n = 411) (n = 344)*p = 0.007p = 0.23*adjusted for age, race, BMI, smoking, diabetes, SBP, DBP, total:HDL, exercise, income, education, and HRT (women)-it also became apparent that FMD was inversely associated with non-fried fish intake in women, which was sig a p-level of 0.007, while men showed a trend toward an opposite effect (i.e. improved FMD with more NFF intake), though the latter was nonsignificant.-otherwise, there no other notable associations detected within other brachial measures and/or other fish types by gender11Non-fried Fish & Brachial Measures: FMD, Stratification by Gender* & Race/Ethnicity*results for women shown, fully adjusted model*p = 0.02-when stratified by both gender and race/ethnicity, which we were obligated to do when the significant interaction was noted between gender and NFF the participant numbers were so small that it was diff to say anything with confidence as indicated by the wide CIs. -we may note some trends, though, as illustrated here where women within each race/ethnic group showed the same pattern of high fish intake & low fmd (seen in all groups except for blacks); similar findings seen with men & FMD (most male ethnic/race groups showed inverse relationship with NFF)12Plasma EPA+DHA & Brachial Measures: FMD, Stratification by Gender*p = 0.33 (n = 221) (n = 173) (n = 203) (n = 178)*adjusted for age, race, BMI, smoking, diabetes, SBP, DBP, total:HDL, exercise, income, education, and HRT (women)

13Plasma EPA+DHA & Brachial Measures: FMD, Stratification by Gender*p = 0.33 (n = 221) (n = 173) (n = 203) (n = 178)*adjusted for age, race, BMI, smoking, diabetes, SBP, DBP, total:HDL, exercise, income, education, and HRT (women)

[Consider d/cing comparison to diet; SAY similar to what I just showed you with diet etc]14Summary Gender differences were observed in the relationship between non-fried fish and both baseline artery diameter and FMD within the MESA cohortResults examining plasma EPA+DHA levels corroborated results observed in dietary non-fried fish intakeSome race/ethnic differences were suggested

15Next StepsAre gender differences in brachial measures explained by variation in plasma %EPA vs %DHA?Gender differences in the relationship between fish consumption & clinical events within MESA? Gender differences in fish consumption & other markers of subclinical CVD (i.e. IMT, LV structure/function)?Results replicable in other large cohorts with mixed race/ethnic groups?

next steps from here should includeTo further examine the gender diffs observed b/t fish intake & brachial measures, we might ask:(a) are there diffs in plasma levels of EPA and DHA, independent of dietary intake? I examined this, it and does appear that women have higher DHA compared to men. Such findings are also supported by others, with population studies as well as experimental data suggesting hormonal effects on n-3 metabolism. (b) However, examining the association between plasma EPA and DHA levels with brachial measures did not yield any significant findings in this study though it is likely that this study was underpowered to observe any significant effects. Thus one might ask (2), whether gender diffs in fish consumption & clinical events are observed2. Few studies to date have examined the question of gender diffs in fish consumption & clinical events, the most notable of which was Nurses Health Study where women showed an inverse relationship between these variables, similar to men. I suspect we will again be underpowered to examine this question in MESA, thus we may consider (3),3. Other markers of subclinical CVD, including CIMT, aortic stiffness, and/or LV mass. I am c


View more >