dietary fats and mortality among patients with type …...2018/12/14 · 11 elucidated, owing to...

Confidential: For Review OnlyDietary Fats and Mortality among Patients with Type 2

Diabetes

Journal: BMJ

Manuscript ID BMJ-2018-048420

Article Type: Research

BMJ Journal: BMJ

Date Submitted by the Author: 14-Dec-2018

Complete List of Authors: Jiao, Jingjing; School of Public Health, Zhejiang University School of Medicine , Department of Nutrition; Harvard University T H Chan School of Public Health, Department of NutritionZong, Geng; Harvard University T H Chan School of Public Health, Department of Nutrition; Shanghai Institutes for Biological Sciences, Chinese Academy of SciencesLiu, Gang; Harvard University T H Chan School of Public Health, Department of NutritionShin, Hyun Joon ; Brigham and Women’s Hospital and Harvard Medical School, Veterans Affairs Boston Healthcare SystemHu, Frank; Harvard University T H Chan School of Public Health, Department of Nutrition, Department of Epidemiology; Brigham and Women’s Hospital and Harvard Medical School, Channing Division of Network Medicine, Department of MedicineRimm, Eric; Harvard University T H Chan School of Public Health, Department of Nutrition, Department of Epidemiology; Brigham and Women's Hospital and Harvard Medical School, Channing Division of Network Medicine, Department of MedicineKathryn, Kathryn; Brigham and Women's Hospital and Harvard Medical School, Division of Preventive Medicine, Department of Medicine; Brigham and Women's Hospital and Harvard Medical School, Division of Women's Health, Department of MedicineManson, JoAnn; Harvard University T H Chan School of Public Health, Department of Epidemiology; Brigham and Women's Hospital and Harvard Medical School, Channing Division of Network Medicine, Division of Preventive Medicine, Department of MedicineSun, Qi; Harvard University T H Chan School of Public Health, Department of Nutrition; Brigham and Women's Hospital and Harvard Medical School, Channing Division of Network Medicine, Department of Medicine

Keywords: saturated fat, polyunsaturated fat, monounsaturated fat, trans fat, total mortality, cardiovascular disease mortality

https://mc.manuscriptcentral.com/bmj

BMJ

Confidential: For Review Only

Page 1 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


1

1 Dietary Fats and Mortality among Patients with Type 2 Diabetes

2

3 Jingjing Jiao, PhD, Geng Zong, PhD, Gang Liu, PhD, Hyun Joon Shin, ScD, Frank B. Hu,

4 MD, PhD, Eric B. Rimm, ScD, Kathryn M. Rexrode, MD, JoAnn E. Manson, MD, DrPH, Qi

5 Sun, MD, ScD

6

7 AFFILIATIONS: From Department of Nutrition, School of Public Health, Zhejiang University

8 School of Medicine, Hangzhou, Zhejiang, China (J.J.J.); Department of Nutrition, Harvard

9 T.H. Chan School of Public Health, Boston, Massachusetts, USA (J.J.J., G.Z., G.L., F.B.H.,

10 E.B.R., Q.S.); Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences,

11 Shanghai, China (G.Z.); Department of Medicine, Brigham and Women's Hospital,

12 Veterans Affairs Boston Healthcare System, Harvard Medical School, Boston,

13 Massachusetts, USA (H.J.S.);Department of Epidemiology, Harvard T.H. Chan School of

14 Public Health, Boston, Massachusetts, USA (F.B.H., E.B.R., J.E.M.); Channing Division of

15 Network Medicine, Department of Medicine, Brigham and Women's Hospital and Harvard

16 Medical School, Boston, Massachusetts, USA (F.B.H., E.B.R., J.E.M., Q.S.); Division of

17 Preventive Medicine, Department of Medicine, Brigham and Women's Hospital and

18 Harvard Medical School, Boston, Massachusetts, USA (K.M.R.); Division of Women's

19 Health, Department of Medicine, Brigham and Women's Hospital and Harvard Medical

20 School, Boston, Massachusetts, USA (K.M.R., J.E.M.).

21 Correspondence to: Geng Zong, PhD, Department of Nutrition, Harvard T.H. Chan School

22 of Public Health, 677 Huntington Avenue, Boston, Massachusetts 02115, USA. E-mail:

23 [email protected]

24 Short running title: Dietary Fats and Mortality in T2D Patients

25 Word count: 3579 words

26 Number of tables and figures: 3 tables and 1 figure

Page 2 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


2

1 ABSTRACT

2 OBJECTIVE

3 To assess the association of dietary fatty acids with cardiovascular disease (CVD) mortality

4 and total mortality among patients with type 2 diabetes (T2D).

5 DESIGN

6 Prospective, longitudinal cohort study.

7 SETTING

8 Health professionals in the United States.

9 PARTICIPANTS

10 We identified and followed 11,264 participants with T2D in the Nurses’ Health Study

11 (1980-2014) and Health Professionals Follow-Up Study (1986-2014).

12 EXPOSURES

13 Dietary fat intake was assessed using validated food frequency questionnaires and

14 updated every 2-4 years.

15 MAIN OUTCOME MEASURE

16 Total and CVD mortality during follow-up.

17 RESULTS

18 We documented 2,502 deaths, including 646 deaths due to CVD. After multivariate

19 adjustment, in replacement of total carbohydrates, intake of polyunsaturated fatty acids

20 (PUFAs) was associated with a lower CVD mortality: HRs (95%CIs) comparing extreme

21 quartiles were 0.76 (0.58, 0.99; P=0.026) for total PUFAs, 0.69 (0.52, 0.90; P=0.0067) for

22 marine n-3 PUFAs, 1.13 (0.85, 1.51) for α-linolenic acid (ALA), and 0.75 (0.56, 1.01) for

23 linoleic acid (LA). Inverse associations with total mortality were also observed for intakes of

24 total PUFAs, n-3 PUFAs, and LA, while monounsaturated fatty acids (MUFAs) of animal,

25 but not plant, origin were associated with a higher total mortality. Moreover, isocalorically

26 replacing 2% of energy from SFAs with total PUFAs or LA was associated with 13% [HR

Page 3 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


3

1 (95%CI): 0.87 (0.77, 0.99)] or 15% [HR (95%CI): 0.85 (0.73, 0.99)] lower CVD mortality,

2 respectively. A 2% replacement of energy from SFAs with total PUFAs was associated with

3 12% [HR (95% CI): 0.88 (0.83, 0.94)] lower total mortality.

4 CONCLUSIONS

5 In T2D patients, higher intakes of PUFAs, in replacement of carbohydrates or SFAs, are

6 associated with lower CVD and total mortality. These findings highlight the important role of

7 dietary fat quality in the prevention of CVD and total mortality among adults with T2D.

Page 4 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


4

1 WHAT IS ALREADY KNOWN ON THIS TOPIC

2 Current dietary guidelines for patients with diabetes recommend limiting trans fat intake and

3 replacing saturated fats with unsaturated fats for maintaining good health, and these

4 recommendations are largely based on findings among general populations. Little is known

5 about the associations of specific dietary fats with total and cardiovascular disease (CVD)

6 mortality among diabetes patients who have altered metabolism of macronutrients.

7 WHAT THIS STUDY ADDS

8 Among diabetes patients, dietary intakes of total polyunsaturated fatty acids (PUFAs),

9 α-linolenic acid, linoleic acid (LA) and marine n-3 PUFAs were associated with lower total

10 mortality whereas intake of monounsaturated fats from animal source is associated with

11 higher total mortality.

12 Replacing saturated fatty acids (SFAs) with PUFAs, especially LA, was associated with a

13 lower CVD mortality.

14 Our analysis indicate that increasing dietary PUFAs, especially LA and marine n-3 PUFAs,

15 in replacement of SFAs may facilitate long-term survival among patients with diabetes.

Page 5 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


5

1 Introduction

2 The number of individuals with type 2 diabetes (T2D) is estimated to be 422 million

3 globally1 and is projected to reach 642 million by 2040,2 imposing a substantial disease toll

4 and economic burden to patients and healthcare systems. Cardiovascular disease (CVD) is

5 the leading cause of deaths in adults with diabetes.3 Current dietary guidelines for CVD

6 prevention and management among diabetes patients recommend limiting intakes of

7 saturated fats (SFAs), trans fats and cholesterol, while increasing the consumption of foods

8 abundant in omega-3 polyunsaturated fatty acids (PUFAs),4 such as fish, nuts and seeds.

9 However, these recommendations were largely based on findings from general

10 populations, and whether they could be extrapolated to patients with diabetes has yet to be

11 elucidated, owing to the altered metabolism of carbohydrates and fats,5 dyslipidemia, and a

12 prothrombotic profile.6 The CVD risk of T2D patients is two to three times that of the general

13 population, thus any possible effects of dietary fats, especially omega-3 PUFAs, needs to

14 be clarified.7 It has been shown that a Mediterranean diet rich in mono- or polyunsaturated

15 fats improved blood glucose and lipid control among individuals with diabetes,8-11 but the

16 association between dietary fats and CVD or total mortality in such population remains

17 lacking.12-15

18 To fill this knowledge gap, we assessed the associations of major dietary fats with CVD

19 mortality and total mortality among adults with diabetes in two large prospective cohort

20 studies. We hypothesize that quality of fats determines their associations with total and

21 CVD mortality among T2D patients.

22

23 Methods

24 Study population

25 The Nurses’ Health Study (NHS) is an ongoing cohort study consisting of 121,701

26 female nurses aged 30–55 y at enrollment in 1976.16 The Health Professionals Follow-Up

Page 6 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


6

1 Study (HPFS) is a parallel cohort initiated in 1986 that enrolled 51,529 U.S. male health

2 professionals aged 40-75 y.17 Details of these two cohorts have been published

3 elsewhere.17 18 Information on non-dietary lifestyle factors, medical history, as well as

4 incident diseases was collected every two years through validated questionnaires.19 The

5 cumulative response rate during follow-up was over 90% in both cohorts. The Human

6 Subjects Committees at Harvard T.H. Chan School of Public Health and Brigham and

7 Women’s Hospital approved these studies. The return of completed questionnaires was

8 considered informed consent.

9 We identified 1498 prevalent T2D patients at baseline (1980 for NHS and 1986 for

10 HPFS) and 9,766 incident T2D during follow-up. We excluded participants who had CVD or

11 cancer at baseline or before T2D diagnosis, had implausible daily caloric intake (<500

12 or >3,500 kcal/day for women, and <800 or >4,200 kcal/day for men), or had missing data

13 of dietary fats at baseline (defined as 1980 or 1986 for prevalent cases, or the return date of

14 the first dietary questionnaire after T2D diagnosis for incident cases). After these

15 exclusions, 9,053 women and 2,211 men with diabetes were included in the analysis.

16

17 Ascertainment of type 2 diabetes

18 Participants with self-reported diagnosis of diabetes were sent a validated supplementary

19 questionnaire to obtain detailed information on the diabetes diagnosis, including diagnosis

20 date, symptoms, and treatment. Before 1997, cases of T2D were diagnosed according to

21 original National Diabetes Data Group criteria20 if they met at least one of the following

22 criteria: classic symptoms plus a fasting plasma glucose concentration ≥140 mg/dL

23 (7.8mmol/L) or a randomly measured plasma glucose concentration ≥200 mg/dL (11.1

24 mmol/L); ≥2 elevated plasma glucose concentrations on different occasions (≥140 mg/dL

25 with fasting or ≥200 mg/dL with randomly measured, ≥200 mg/dL for ≥2 h after an oral

26 glucose challenge); or treatment with hypoglycemic medication (insulin or an oral

Page 7 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


7

1 hypoglycemic agent). Since 1997, the fasting plasma glucose concentration for diabetes

2 diagnosis was lowered to 7.0 mmol/L, or 126 mg/dL based on the American Diabetes

3 Association diagnostic criteria.21 In validation studies, 98% of questionnaire-confirmed

4 diabetes cases were re-confirmed by medical record review in the NHS22 and 97% of cases

5 were re-confirmed in the HPFS.23 Given such high accuracy, we included all self-reported

6 diabetes cases to increase sample size.

7

8 Dietary assessment

9 Diet was assessed using validated semi-quantitative food frequency questionnaires

10 (sFFQs) administered every 2-4 years.24 25 The overall validity and reliability of the sFFQs

11 have been reported previously.24 25 Nutrient composition was obtained from the Harvard

12 University Food Composition Database, which is continuously updated to account for

13 changes in nutrient contents and food processing. Fat intake was then calculated by

14 multiplying the consumption frequency of each food item with a pre-specified portion size

15 (e.g., one egg or one slice of bread) by the content of fats, taking into account of the brand

16 and type of fat use in preparation. Dietary fats were expressed as percentages of energy

17 and we calculated cumulative averages across valid assessments during the follow-up to

18 represent long-term intake. Dietary variables was not updated upon a diagnosis of incident

19 CVD or cancer during study follow-up because the incidence of these conditions often

20 leads to changes in diet that are not relevant to the associations of interest.

21 In validation studies, the assessment of dietary fats was compared with that estimated

22 by multiple 1-wk diet records and fatty acid biomarkers in adipose tissue. Correlation

23 coefficients between fat intakes assessed by the 1986 FFQ and diet records were 0.70 for

24 SFAs, 0.69 for MUFAs, and 0.64 for PUFAs.26 Correlations between fatty acids estimated

25 from FFQs and measured in adipose tissue were 0.35 for linoleic acid (LA), 0.48 for marine

26 n-3 PUFAs, and 0.51 for trans fatty acids.27

Page 8 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


8

1

2 Covariate assessments

3 Information on lifestyle and other potential risk factors for mortality was assessed at

4 baseline and updated during follow-up through biennial questionnaires, including age,

5 ethnicity, weight, smoking status, alcohol drinking, physical activity, family history of

6 diabetes, family history of myocardial infarction, self-reported hypertension and

7 hypercholesterolemia, use of multivitamin, aspirin, and use of insulin and oral hypoglycemic

8 drugs. Body mass index (BMI) was calculated by dividing weight (kg) by height squared

9 (m2). Physical activity was assessed as hours per week of moderate or vigorous activities

10 (including brisk walking) that require at least the expenditure of 3 metabolic equivalents or

11 more per hour (METs).28

12

13 Ascertainment of deaths

14 Deaths were identified through searching the National Death Index, and reports by next of

15 kin or postal authorities.29 Over 98% deaths were identified.30The cause of death was

16 ascertained by physicians who were blinded of the risk profiles of the participants through

17 reviewing death certificates and medical records. International Classification of Diseases

18 (ICD), 8thand 9thRevision codes were used to classify cardiovascular mortality (ICD-8,

19 390.0-458.9 and ICD-9,390.0-459.9) and cancer mortality (ICD-8, 140.0-207.9, ICD-9,

20 140.0-208.9).31

21

22 Statistical analysis

23 Person-time was calculated from the return date of the first dietary questionnaire after T2D

24 diagnosis to the time of death or the end of follow-up (June 30, 2014 for the NHS and

25 January 30, 2014 for the HPFS), whichever was earlier. Cox proportional hazards models

Page 9 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


9

1 were applied to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for the

2 associations of different dietary fats with CVD, cancer, and total mortality. To ensure

3 enough numbers of deaths for analyses, we pooled participants from the two cohorts (no

4 significant heterogeneity was found).

5 The association of specific dietary fats and mortality was analyzed using nutrient

6 density models that included intakes of total energy, percentage of energy from protein and

7 all other dietary fats, and were further adjusted for age, gender, race, BMI at diagnosis,

8 smoking status, smoking pack-years, physical activity, alcohol consumption, family history

9 of diabetes or myocardial infarction, self-reported hypertension or hypercholesterolemia,

10 diabetes duration, multivitamin use, current aspirin use, and dietary cholesterol intake. The

11 model for cancer mortality was additionally adjusted for family history of cancer. In these

12 models, regression coefficients for the specific dietary fats can be interpreted as the

13 estimated effect of substituting the fat at issue for the same energy from carbohydrates. We

14 also evaluated the effects of replacing SFAs with other types of fats in isocaloric

15 substitution models which simultaneously included total energy intake, percentages of

16 energy from protein, carbohydrates, and all fats except SFAs. Tests for trend were

17 calculated by modelling the median values of each category as a continuous variable.

18 In a secondary analysis, we further analyzed the associations for MUFAs from plant

19 and animal origin separately. Sensitivity analyses were performed by excluding mortality

20 that occurred within 4 years after diabetes diagnosis, excluding prevalent T2D, excluding

21 participants with extreme BMI (<18.5 or >40.0 kg/m2), restricting the analysis to confirmed

22 T2D cases only, or further adjusting for the use of hypoglycemic medication.

23 Statistical analyses were performed with SAS 9.4 (SAS Institute, Cary, NC, USA).

24 Two-sided P values <0.05 were considered as statistically significant.

25

26 Patient involvement

Page 10 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


10

1 Participants were not involved in raising the research question or the outcome measures,

2 nor were they involved in developing plans for recruitment, design, or implementation of the

3 study. No participant was asked to advise on interpretation or writing up of the manuscript.

4 The participants are updated on the study outcomes and developments through the study

5 website (www.nurseshealthstudy.org and https://www.hsph.harvard.edu/hpfs/index.html)

6 and newsletters.

7

8 Results

9 Population characteristics

10 During a mean follow-up of 11 years (124,362 person-years), we documented 2,502

11 deaths, including 646 deaths due to CVD. At diabetes diagnosis, women with higher PUFA

12 intake were on average older, exercised less, and more likely to have hypertension and

13 hypercholesterolemia (table 1). Men with higher PUFA consumption were more likely to be

14 non-smokers, to exercise more, and to use aspirin. They also drank less alcohol and were

15 more likely to use multivitamins, consume more energy, SFAs, MUFAs, and trans fat but

16 less carbohydrates. Participant characteristics according to SFA and MUFA intake were

17 presented in supplementary table1.

18

19 Cardiovascular and total mortality

20 In Table 2, we observed a significant inverse association of total PUFAs with CVD and total

21 mortality after adjusting for age and gender. In multivariate-adjusted model that estimated

22 mortality risk by substituting fats for carbohydrates, the association remained significant:

23 HRs (95% CIs) across quartiles of PUFA intake (from low to high) were 1.00, 0.99 (0.80,

24 1.23), 0.85 (0.67, 1.08), and 0.76 (0.58, 0.99) for CVD mortality (Ptrend = 0.03) and 1.00,

25 0.86 (0.77, 0.95), 0.83 (0.74, 0.94), and 0.68 (0.60, 0.78) for total mortality (Ptrend<0.0001).

26 Intakes of SFAs and trans fats were positively associated with CVD mortality in age and

Page 11 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


11

1 gender-adjusted model (model 1), but became non-significant after adjusting for other

2 covariates (model 2; table 2).

3 Total MUFA intake was not significantly associated with either CVD or total mortality

4 (model 2; table 2). Nonetheless, MUFAs from animal source (mainly red meats and high-fat

5 dairy products) were positively associated with total mortality (HR: 1.23; 95% CI: 1.04, 1.45;

6 Ptrend = 0.0005; supplementary table 2), and non-significantly associated with a higher

7 CVD mortality (HR comparing the highest versus lowest quartile: 1.27; 95% CI: 0.92-1.75;

8 Ptrend = 0.07). No significant associations with mortality were observed for plant-derived

9 MUFA intake.

10 For specific PUFAs, intake of marine n-3 PUFAs was associated with a lower

11 cardiovascular mortality after multivariate adjustment: HRs (95% CIs) comparing extreme

12 quartiles were 0.69 (0.52, 0.90; Ptrend = 0.007; high vs low). The association between LA

13 and CVD mortality did not achieve statistical significance, with an HR (95% CI) of 0.75

14 (0.56, 1.01; Ptrend = 0.06; table 3). Intakes of marine n-3 PUFAs and LA were associated

15 with lower total mortality: HRs (95% CIs) comparing extreme quartiles were 0.71 (0.62,

16 0.82; Ptrend < 0.0001) and 0.81 (0.69, 0.94; Ptrend = 0.008; table 3), respectively. ALA intake

17 was non-significantly associated with lower total mortality (HR: 0.88; 95% CI: 0.76, 1.02;

18 Ptrend = 0.04).

19

20 Cancer mortality

21 No significant associations of SFAs and MUFAs intakes with cancer mortality were

22 observed (supplementary tables 3). As shown in supplementary table 4, participants in

23 the highest quartile of marine n-3 PUFA intake had 28% (HR: 0.72; 95%CI: 0.53, 0.99)

24 lower risk of cancer mortality, comparing with those in the lowest quartile. ALA turned to be

25 inversely associated with cancer mortality, but HR comparing extreme quartiles did not

26 achieve statistical significance (0.75; 95% CI: 0.53, 1.05; Ptrend = 0.04).

Page 12 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


12

1

2 Substitution for SFAs

3 Figure 1 presents mortality risk by isocaloric replacement of SFAs with other fats. CVD

4 mortality was 13% [HR (95% CI): 0.87 (0.77, 0.99)] lower when 2% energy from SFAs was

5 iso-calorically replaced by total PUFAs, and 15% (HR: 0.85; 95% CI: 0.73, 0.99) lower

6 when replaced by LA. Similarly, replacing 2% energy from SFAs with total PUFAs was

7 associated with 12% (HR: 0.88; 95% CI: 0.83, 0.94) lower total mortality. No significant

8 changes in total mortality [HR (95% CI): 0.99 (0.94, 1.05)] or CVD mortality [HR (95% CI):

9 0.99 (0.88, 1.11)] were observed when replacing 2% energy from SFAs with MUFAs.

10 Replacing SFAs with PUFAs was not associated with cancer mortality (supplementary fig

11 1).

12

13 Sensitivity analyses

14 The associations of specific dietary fats with total, CVD, and cancer mortality did not

15 change materially after excluding participants with extreme BMI or further adjusting for the

16 use of hypoglycemic medication (supplementary table 5). Excluding deaths that occurred

17 within 4 years after T2D diagnosis slightly strengthened the results. We observed similar

18 results when restricting the analysis within incidence cases only. Restricting the analysis to

19 confirmed T2D cases only slightly attenuated the associations.

20

21 Discussion

22 In two cohorts of US men and women, we demonstrated an inverse association of dietary

23 PUFAs, including LA and marine n-3 PUFAs, with CVD mortality or total mortality among

24 T2D patients. In addition, animal-derived MUFAs were associated with higher total mortality

25 compared with dietary carbohydrates.

26

Page 13 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


13

1 Comparison with other studies and possible explanations

2 To our knowledge, this is the first prospective analysis that investigated the

3 associations between specific dietary fats and CVD mortality among individuals with T2D.

4 Our findings regarding PUFAs are largely ascribed to LA, which is the most abundant

5 PUFA in the diet.4 Similarly, a recent analysis of NHS and HPFS demonstrated a significant

6 inverse association of dietary LA with total and CVD mortality among participants who were

7 free of major chronic diseases, including diabetes, at baseline.32 However, these findings

8 are seemingly contradictory to evidence from the Sydney Diet Heart Study that showed

9 higher LA intake led to increased mortality.33 It is worth noting that safflower oil and

10 margarines were used as the food source of LA in this intervention, therefore the observed

11 effect may be at least partially ascribed to high trans fat contents of margarines in early

12 days.34 In contrast, a recent meta-analysis of four randomized controlled trials (RCTs)

13 demonstrated that substituting PUFAs for SFAs reduced coronary heart disease (CHD) risk

14 by approximately 30%.35 Our findings are also in line with a 30-week intervention study that

15 showed LA-enriched diet (10.9% of energy) reduced total and LDL-cholesterol levels

16 among individuals with T2D.36 Collectively, these data suggest that intake of n-6 PUFAs

17 may exert beneficial effects on cardiovascular health among individuals with and without

18 T2D.

19 Potential cardiovascular benefits of marine n-3 PUFAs have been reported in

20 numerous cohort studies,37 38 although recent meta-analysis did not support protective

21 effects of marine n-3 supplements on cardiovascular outcomes, including cardiac death.39

22 This inconsistency maybe due to high baseline dietary marine n-3 PUFA intake, few

23 cardiac deaths, revascularization therapy, and use of statin and other lipid-lowering drugs

24 that may mask the effects of low-to-moderate dose of fish oil supplementation on CVD

25 patients.40 Indeed, in the recent VITAL trial, marine n-3 fatty acid supplementation

26 significantly lowered CVD risk among participants who had a lower intake of these fatty

Page 14 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


14

1 acids at baseline.41 Moreover, in the REDUCE-IT trial that used a high dose (4 g/day) of

2 pure EPA supplementation, the risk of developing CVD was substantially reduced.42

3 Evidence specifically pertinent to diabetes patients is relatively sparse. Clinical trials

4 concerning CVD intermediate outcomes among adults with diabetes have consistently

5 reported that marine n-3 supplements could lower triglycerides levels,43 improve arterial

6 blood flow, and ameliorate inflammatory condition that predispose CVD events.44 45

7 Meanwhile, a secondary analysis of JELIS study found that eicosapentaenoic acid (1.8

8 g/day) for 4.6 years led to a 22% reduction in coronary artery disease (CAD) risk among

9 4,565 adults with T2D or impaired glucose metabolism,46 but this effect was not observed in

10 another RCT (0.9 g/day) with a larger sample size (n = 12,536) and a longer follow-up

11 duration (6.2 years).47 In the latest trial of 15,480 patients with diabetes, though daily

12 supplementation with 0.84 g marine n-3 PUFA for 7.4 years did not significantly lower the

13 risk of major vascular events, including nonfatal myocardial infarction or stroke, transient

14 ischemic attack, and vascular death, the vascular deaths were fewer in the marine n-3 fatty

15 acid group compared with the control group [rate ratio (95% CI): 0.82 (0.68–0.98)].7

16 The inverse association of marine n-3 PUFA and ALA with cancer mortality is

17 consistent with a previous cohort study that reported combined intakes of n-3 PUFAs from

18 diet and supplements were associated with lower cancer mortality in a general population.48

19 In addition, meta-analyses have concluded that n-3 PUFA intake was inversely related to

20 cancer risks in breast,49 50 prostate,51 and liver.52 It is worth noting that the morbidity and

21 mortality of some cancers, such as endometrial , breast and colorectal cancers is elevated

22 in individuals with T2D.53 In rodent models, n-3 PUFA supplements also ameliorated the

23 diabetic phenotype and subsequently reduced risk of colorectal cancer via inhibiting colonic

24 expression of HNF-4α and β-catenin/Tcf.54

25 MUFA-enriched diets in comparison with carbohydrates have been reported to improve

26 metabolic risk factors in adults with T2D.55 Consistently, substitution of MUFAs for

Page 15 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


15

1 carbohydrates was associated with lower total mortality among individuals with diabetes in

2 the European Prospective Investigation into Cancer and Nutrition.56 We have recently

3 reported that the associations of dietary MUFAs with CHD risk were dependent on the food

4 sources: those from plant foods was associated with a lower CHD risk, whereas their

5 counterparts from animal sources turned to be associated with a higher risk, possibly

6 because of confounding by other food constitutes.57 In the current study, we did not detect

7 significant associations with total or CVD mortality for total MUFAs or plant-derived MUFAs,

8 yet animal-based MUFAs were significantly associated with total mortality. Similarly, the

9 Kuopio Ischemic Heart Disease Risk Factor study reported that dietary MUFA, which was

10 mainly from animal products was associated with a higher risk of fatal CHD.58 Taken

11 together, these data emphasize the role of food sources in health effects of MUFA intake

12 on cardiovascular health in general populations and diabetes patients.

13

14 Strengths and limitations

15 The current investigation has the strength of using the repeated assessments of diet

16 after diabetes diagnosis, which can help capture potential dietary changes typically

17 observed among individuals with T2D.12 56 Other strengths include the relatively large

18 sample size, long-term follow-up with a high follow-up rate (>90%), and repeated

19 measurements of other lifestyle factors. In addition, when we excluded participants who

20 died within 4 y or those with extreme BMI, the results did not change materially, indicating

21 the robustness of our findings. There are also limitations. First, glycemic control and

22 severity of diabetes were not rigorously assessed in our study. However, when we adjusted

23 for duration of diabetes in the model and further adjusted for the use of insulin and

24 hypoglycemic medications, the results remained unchanged. Prevalent cases may differ

25 from incident cases in terms of their risk profiles, but restricting the analysis within incident

26 cases yields similar results. Third, measurement errors in self-reported diet and lifestyle

Page 16 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


16

1 factors were inevitable, which may likely attenuate the associations of interest owing to the

2 prospective study design. Fourth, we could not rule out unmeasured confounding by

3 medication adherence, psychosocial stress, or other factors that were not assessed in the

4 cohorts. Lastly, causality may not be established because of the observational nature of

5 this study.

6

7 Conclusions and implications

8 In conclusion, among US men and women with T2D, dietary intake of PUFAs,

9 especially LA and marine n-3 PUFAs, is associated with lower CVD and total mortality. Our

10 results suggest that dietary PUFAs, in replacement of SFAs or carbohydrates, may

11 facilitate the long-term survival among adults with T2D.

12

13 Contributors: JJ, GZ, and QS conceived and designed the study. JJ, GZ, GL, and QS

14 analyzed and interpreted data. JJ drafted the manuscript. FBH, EBR, KMR, JEM, and QS

15 revised manuscript critically for important intellectual content. All authors provided final

16 approval of the version to be published. JJ is the guarantor of this work and, as such, had

17 full access to all the data in the study and takes responsibility for the integrity of the data

18 and the accuracy of the data analysis.

19 Funding: This study was sponsored by the National Institutes of Health, CA186107,

20 CA167552, CA87969, DK082486, HL35464, DK058845, HL088521, and

21 HL034594.Thefunders had no role in design and conduct of the study; collection,

22 management, analysis, and interpretation of the data; preparation, review, and approval of

23 the manuscript; or the decision to submit the manuscript for publication.

24 Competing interests: All authors have completed the ICMJE uniform disclosure form at

25 www.icmje.org/coi_disclosure.pdf and declare: support from the National Institutes of

26 Health for the submitted work; GZ is supported by a postdoctoral fellowship funded by

Page 17 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


17

1 Unilever R&D, Vlaardingen, Netherlands; QS has received ad hoc consulting fees from the

2 Emavant Solutions GmbH; FBH has received research support from California Walnut

3 Commission and Metagenics; no other relationships or activities that could appear to have

4 influenced the submitted work.

5 Ethical approval: The study protocol was approved by the institutional review boards of

6 the Brigham and Women's Hospital and the Harvard T H Chan School of Public Health. The

7 completion of the self-administered questionnaire was considered to imply informed

8 consent.

9 Transparency declaration: The manuscript’s guarantor affirms that the manuscript is an

10 honest, accurate, and transparent account of the study being reported; that no important

11 aspects of the study have been omitted; and that any discrepancies from the study as

12 planned (and, if relevant, registered) have been explained.

13 Data sharing: No additional data available.

Page 18 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


18

1 REFERENCES

2 1 Worldwide trends in diabetes since 1980: a pooled analysis of 751 population-based

3 studies with 4.4 million participants. Lancet 2016;387:1513-30.

4 2 Federation ID. Idf diabetes atlas. 2015;7th ed.

5 3 Cardiovascular disease and risk management. Sec. 9. In standards of medical care in

6 diabetes-2017. Diabetes Care 2017;40:S75-S87.

7 4 Lifestyle management: standards of medical care in diabetes-2018. Diabetes Care

8 2018;41:S38

9 5 Guasch-Ferré M, Hruby A, Toledo E, et al. Metabolomics in prediabetes and diabetes:

10 a systematic review and meta-analysis. Diabetes Care 2016;39:833-46.

11 6 Seligman BG, Biolo A, Polanczyk CA, et al. Increased plasma levels of endothelin 1

12 and von Willebrand factor in patients with type 2 diabetes and dyslipidemia. Diabetes

13 Care 2000;23:1395.

14 7 Effects of n-3 Fatty Acid Supplements in Diabetes Mellitus. N Engl J Med

15 2018;379:1540-50.

16 8 Shai I, Schwarzfuchs D, Henkin Y, et al. Weight loss with a low-carbohydrate,

17 mediterranean, or low-fat diet. N Engl J Med 2008;359:229-41.

18 9 Brehm BJ, Lattin BL, Summer SS, et al. One-Year Comparison of a high–

19 monounsaturated fat diet with a high-carbohydrate diet in type 2 diabetes. Diabetes

20 Care 2009;32:215-20.

21 10 Bloomfield HE, Koeller E, Greer N, et al. Effects on Health Outcomes of a

22 mediterranean diet with no restriction on fat intake: a systematic review and

23 meta-analysis. Ann Intern Med 2016;165:491-500.

24 11 Estruch R, Ros E, Salas-Salvadó J, et al. Primary prevention of cardiovascular

25 disease with a mediterranean diet supplemented with extra-virgin olive oil or nuts. N

26 Engl J Med 2018;378:e34.

Page 19 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


19

1 12 dos Santos ALT, Weiss T, Duarte CK, et al. Dietary fat composition and cardiac

2 events in patients with type 2 diabetes. Atherosclerosis 2014;236:31-38.

3 13 Hu FB, Cho E, Rexrode KM, et al. Fish and long-chain ω-3 fatty acid intake and risk of

4 coronary heart disease and total mortality in diabetic women. Circulation

5 2003;107:1852.

6 14 Tanasescu M, Cho E, Manson JE, et al. Dietary fat and cholesterol and the risk of

7 cardiovascular disease among women with type 2 diabetes. Am J Clin Nutr

8 2004;79:999-1005.

9 15 Patel YR, Gadiraju TV, Gaziano JM, et al. Adherence to healthy lifestyle factors and

10 risk of death in men with diabetes mellitus: The Physicians' Health Study. Clin Nutr

11 2018;37:139-43.

12 16 Willett WC, Green A, Stampfer MJ, et al. Relative and absolute excess risks of

13 coronary heart disease among women who smoke cigarettes. N Engl J

14 Med1987;317:1303-09.

15 17 Colditz GA, Rimm EB, Giovannucci E, et al. A prospective study of parental history off

16 myocardial infarction and coronary artery disease in men. Am J

17 Cardiol1991;67:933-38.

18 18 Hu FB, Manson JE, Stampfer MJ, et al. Diet, lifestyle, and the risk of type 2 diabetes

19 mellitus in women. N Engl J Med 2001;345:790-97.

20 19 Mozaffarian D, Hao T, Rimm EB, et al. Changes in diet and lifestyle and long-term

21 weight gain in women and men. N Engl J Med 2011;364:2392-404.

22 20 Classification and diagnosis of diabetes mellitus and other categories of glucose

23 intolerance. Diabetes 1979;28:1039.

24 21 Standards of medical care in diabetes—2010. Diabetes Care 2010;33:S11.

25 22 Manson JE, Stampfer MJ, Colditz GA, et al. Physical activity and incidence of

26 non-insulin-dependent diabetes mellitus in women. Lancet 1991;338:774-78.

Page 20 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


20

1 23 Hu FB, Leitzmann MF, Stampfer MJ, et al. Physical activity and television watching in

2 relation to risk for type 2 diabetes mellitus in men. Arch Intern Med 2001;161:1542-48.

3 24 Rimm EB, Giovannucci EL, Stampfer MJ, et al. Reproducibility and validity of an

4 expanded self-administered semiquantitative food frequency questionnaire among

5 male health professionals. Am J Epidemiol 1992;135:1114-26.

6 25 Willett WC, Sampson L, Stampfer MJ, et al. Reproducibility and validity of a

7 semiquantitative food frequency questionnaire. Am J Epidemiol 1985;122:51-65.

8 26 Willett WC. Reproducibility and validity of food-frequency questionnaires.In:

9 Nutritional Epidemiology. 3rd ed. New York: Oxford University, 2013.

10 27 London SJ, Sacks FM, Caesar J, et al. Fatty acid composition of subcutaneous

11 adipose tissue and diet in postmenopausal US women. Am J Clin Nutr

12 1991;54:340-45.

13 28 Wolf AM, Hunter DJ, Colditz GA, et al. Reproducibility and validity of a

14 self-administered physical activity questionnaire. Int J Epidemiol 1994;23:991-99.

15 29 Hu FB, Willett WC, Li T, et al. Adiposity as compared with physical activity in

16 predicting mortality among women. N Engl J Med 2004;351:2694-703.

17 30 Rich-Edwards JW, Corsano KA, Stampfer MJ. Test of the national death index and

18 equifax nationwide death search. Am J Epidemiol 1994;140:1016-19.

19 31 Tobias DK, Pan A, Jackson CL, et al. Body-mass index and mortality among adults

20 with incident type 2 diabetes. N Engl J Med 2014;370:233-44.

21 32 Wang DD, Li Y, Chiuve SE, et al. Specific dietary fats in relation to total and

22 cause-specific mortality. JAMA Intern Med 2016;176:1134-45.

23 33 Ramsden CE, Zamora D, Leelarthaepin B, et al. Use of dietary linoleic acid for

24 secondary prevention of coronary heart disease and death: evaluation of recovered

25 data from the Sydney Diet Heart Study and updated meta-analysis. BMJ

26 2013;346:e8707.

Page 21 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


21

1 34 Chowdhury R, Warnakula S, Kunutsor S, et al. Association of dietary, circulating, and

2 supplement fatty acids with coronary risk: a systematic review and meta-analysis. Ann

3 Intern Med 2014;160:398-406.

4 35 Sacks FM, Lichtenstein AH, Wu JHY, et al. Dietary fats and cardiovascular disease: a

5 presidential advisory from the American Heart Association. Circulation 2017;136:e1.

6 36 Heine RJ, Mulder C, Popp-Snijders C, et al. Linoleic-acid-enriched diet: long-term

7 effects on serum lipoprotein and apolipoprotein concentrations and insulin sensitivity

8 in noninsulin-dependent diabetic patients. Am J Clin Nutr 1989;49:448-56.

9 37 Lee JH, O'Keefe JH, Lavie CJ, et al. Omega-3 fatty acids: cardiovascular benefits,

10 sources and sustainability. Nat Rev Cardiol 2009;6:753.

11 38 Saravanan P, Davidson NC, Schmidt EB, et al. Cardiovascular effects of marine

12 omega-3 fatty acids. Lancet;376:540-50.

13 39 Aung T, Halsey J, Kromhout D, et al. Associations of omega-3 fatty acid supplement

14 use with cardiovascular disease risks: Meta-analysis of 10 trials involving 77 917

15 individuals. JAMA Cardiol2018;3:225-34.

16 40 Siscovick DS, Barringer TA, Fretts AM, et al. Omega-3 polyunsaturated fatty acid (fish

17 oil) supplementation and the prevention of clinical cardiovascular disease. Circulation

18 2017;135:e867.

19 41 Manson JE, Cook NR, Lee IM, et al. Marine n−3 fatty acids and prevention of

20 cardiovascular disease and cancer. N Engl J Med 2018.

21 42 Bhatt DL, Steg PG, Miller M, et al. Cardiovascular risk reduction with icosapent ethyl

22 for hypertriglyceridemia. N Engl J Med 2018.

23 43 Tajuddin N, Shaikh A, Hassan A. Prescription omega-3 fatty acid products:

24 considerations for patients with diabetes mellitus. Diabetes MetabSyndrObes

25 2016;9:109-18.

Page 22 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


22

1 44 De Luis DA, Conde R, Aller R, et al. Effect of omega-3 fatty acids on cardiovascular

2 risk factors in patients with type 2 diabetes mellitus and hypertriglyceridemia: an open

3 study. Eur Rev Med Pharmacol Sci 2009;13:51-55.

4 45 Hendrich S. (n-3) fatty acids: clinical trials in people with type 2 diabetes. Adv Nutr

5 2010;1:3-7.

6 46 Oikawa S, Yokoyama M, Origasa H, et al. Suppressive effect of EPA on the incidence

7 of coronary events in hypercholesterolemia with impaired glucose metabolism:

8 Sub-analysis of the Japan EPA Lipid Intervention Study (JELIS). Atherosclerosis

9 2009;206:535-39.

10 47 n-3 fatty acids and cardiovascular outcomes in patients with dysglycemia. N Engl J

11 Med 2012;367:309-18.

12 48 Bell GA, Kantor ED, Lampe JW, et al. Intake of long-chain ω-3 fatty acids from diet

13 and supplements in relation to mortality. Am J Epidemiol 2014;179:710-20.

14 49 Zheng J-S, Hu X-J, Zhao Y-M, et al. Intake of fish and marine n-3 polyunsaturated

15 fatty acids and risk of breast cancer: meta-analysis of data from 21 independent

16 prospective cohort studies. BMJ 2013;346.

17 50 Liu J, Ma DWL. The role of n-3 polyunsaturated fatty acids in the prevention and

18 treatment of breast cancer. Nutrients 2014;6:5184-223.

19 51 Fu Y-Q, Zheng J-S, Yang B, et al. Effect of individual omega-3 fatty acids on the risk

20 of prostate cancer: a systematic review and dose-response meta-analysis of

21 prospective cohort studies. J Epidemiol 2015;25:261-74.

22 52 Gao M, Sun K, Guo M, et al. Fish consumption and n-3 polyunsaturated fatty acids,

23 and risk of hepatocellular carcinoma: systematic review and meta-analysis. Cancer

24 Causes Control 2015;26:367-76.

25 53 Szablewski L. Diabetes mellitus: influences on cancer risk. Diabetes Metab Res Rev

26 2014;30:543-53.

Page 23 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


23

1 54 Algamas-Dimantov A, Yehuda-Shnaidman E, Hertz R, et al. Prevention of

2 diabetes-promoted colorectal cancer by (n-3) polyunsaturated fatty acids and (n-3)

3 PUFA mimetic. Oncotarget 2014;5:9851-63.

4 55 Qian F, Korat AA, Malik V, et al. Metabolic effects of monounsaturated fatty acid–

5 enriched diets compared with carbohydrate or polyunsaturated fatty acid–enriched

6 diets in patients with type 2 diabetes: a systematic review and meta-analysis of

7 randomized controlled trials. Diabetes Care 2016;39:1448-57.

8 56 Campmans-Kuijpers MJ, Sluijs I, Nöthlings U, et al. The association of substituting

9 carbohydrates with total fat and different types of fatty acids with mortality and weight

10 change among diabetes patients. Clin Nutr 2016;35:1096-102.

11 57 Zong G, Li Y, Sampson L, et al. Monounsaturated fats from plant and animal sources

12 in relation to risk of coronary heart disease among US men and women. Am J Clin

13 Nutr 2018;107:445-53.

14 58 Virtanen JK, Mursu J, Tuomainen T-P, et al. Dietary fatty acids and risk of coronary

15 heart disease in men: the Kuopio Ischemic Heart Disease Risk Factor Study.

16 Arterioscler Thromb Vasc Biol 2014;34:2679.

Page 24 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


2

Table 1. Characteristics of participants at diabetes diagnosis*

Quartiles of Polyunsaturated Fat Intake

Characteristic 1 2 3 4

NHS (n=9,053) 　　

Age (years) 63.4 (10.5) 63.1 (10.8) 62.9 (10.9) 64.4 (11.2)†White (%) 96 96.1 96.7 95.9Body Mass Index (kg/m2) 29.3 (5.9) 29.9 (6.3) 30.1 (6.3) 29.8 (6.4)

Alcohol consumption (g/day) 4.0 (9.6) 2.9 (7.0) 3.0 (7.2) 2.7 (6.4)†Physical activity (hours/week) 1.7 (2.8) 1.6 (2.7) 1.4 (2.4) 1.5 (2.6)†Current smoker (%) 12.7 12.4 12.2 12.4Smoking pack-year 26.4 (22.0) 24.6 (20.7) 25.6 (21.5) 24.8 (20.6)Multivitamin use (%) 49.3 49.5 54.4 51.6†Family history of diabetes (%) 49.4 49.3 49.3 50.9Family history of MI (%) 27.9 28 29.5 28.6Family history of cancer (%) 14.5 14.4 12.8 12.5Aspirin use (%) 52.8 54.6 55.4 55.2Hypoglycemic medication use (%) 41.1 44.8 47.8 45.1Duration of diabetes (years) 4.5 (4.2) 4.8 (4.5) 4.8 (4.3) 4.7 (4.2)Hypercholesterolemia (%) 61.8 61.1 65.8 65.9†Hypertension (%) 74.4 74.8 75.4 75.9†Total energy intake (kcal/day) 1629.4 (521.9) 1655.6 (500.1) 1676.1 (504.8) 1692.0 (542.6)†Carbohydrates (% energy) 52.1 (10.3) 48.6 (9.3) 46.8 (8.7) 44.1 (8.4)†Protein (% energy) 19.6 (3.8) 19.5 (3.6) 19.2 (3.4) 18.8 (3.6)†Total fats (% energy) 28.2 (7.4) 32.4 (6.9) 34.5 (6.5) 37.9 (6.4)†Total PUFAs (% energy) 4.2 (0.7) 5.4 (0.5) 6.4 (0.6) 8.5 (1.8)†n-3 PUFAs (% energy) 0.6 (0.2) 0.7 (0.2) 0.8 (0.2) 1.0 (0.5)† ALA (% energy) 0.4 (0.1) 0.5 (0.1) 0.6 (0.2) 0.8 (0.4)† Marine n-3 PUFAs (% energy) 0.1 (0.1) 0.1 (0.2) 0.1 (0.2) 0.2 (0.2)†n-6 PUFAs (% energy) 3.6 (0.9) 4.8 (0.8) 5.6 (0.9) 7.4 (1.8)† LA (% energy) 3.6 (0.9) 4.7 (0.8) 5.5 (0.9) 7.3 (1.8)† AA (% energy) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0)†SFAs (% energy) 10.7 (4.0) 11.7 (3.8) 11.8 (3.5) 11.9 (3.3)†MUFAs (% energy) 10.7 (3.4) 12.5 (3.4) 13.4 (3.5) 14.7 (3.6)†Trans fats (% energy) 1.3 (0.5) 1.5 (0.6) 1.6 (0.7) 1.7 (0.9)†　　　

HPFS (n=2,211) 　　

Age (years) 68.6 (9.4) 67.5 (8.4) 67.6 (8.7) 67.6 (8.1)White (%) 91.2 90.3 89.3 89.7Body Mass Index (kg/m2) 28.6 (4.9) 29.0 (5.0) 28.3 (4.6) 28.3 (4.6)

Alcohol consumption (g/day) 12.7 (19.7) 8.9 (12.9) 9.8 (14.2) 7.8 (11.3)†Physical activity (hours/week) 2.9 (5.0) 3.1 (7.9) 3.2 (5.1) 3.5 (6.8)†Current smoker (%) 5.9 5.2 5.3 3.8†Smoking pack-year 18.4 (22.3) 16.2 (20.9) 16.8 (20.2) 16.4 (20.3)Multivitamin use (%) 51.2 54.8 62.6 61†

Page 25 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


3

Family history of diabetes (%) 36.4 40 37.7 41.1Family history of MI (%) 36.5 33.1 30.3 31Family history of cancer (%) 31.2 41.4 36.2 36Aspirin use (%) 53 55.7 56.7 58.9†Hypoglycemic medication use (%) 21.5 20.8 26.2 25Duration of diabetes (years) 3.1 (1.0) 3.3 (1.3) 3.3 (1.5) 3.3 (1.4)†Hypercholesterolemia (%) 65.9 61.6 69.9 66.7Hypertension (%) 70.9 72.1 72.3 72.1Total energy intake (kcal/day) 1944.5 (566.7) 1927.8 (570.6) 1950.3 (648.2) 2014.2 (666.8)†

Carbohydrates (% energy) 51.8 (8.9) 49.2 (7.7) 47.3 (7.6) 43.4 (8.2)†Protein (% energy) 18.6 (3.5) 19.1 (3.1) 18.9 (3.7) 18.6 (3.4)Total fats (% energy) 27.2 (6.0) 30.5 (5.3) 32.4 (5.3) 37.6 (6.3)†Total PUFAs (% energy) 4.3 (0.6) 5.5 (0.4) 6.4 (0.5) 8.6 (2.0)†n-3 PUFAs (% energy) 0.6 (0.2) 0.7 (0.2) 0.8 (0.3) 1.0 (0.5)† ALA (% energy) 0.4 (0.1) 0.5 (0.1) 0.6 (0.1) 0.8 (0.4)† Marine n-3 PUFAs (% energy) 0.1 (0.1) 0.2 (0.2) 0.2 (0.2) 0.2 (0.2)†n-6 PUFAs (% energy) 4.0 (0.6) 5.0 (0.6) 5.8 (0.7) 7.8 (1.8)† LA (% energy) 3.9 (0.6) 4.9 (0.6) 5.7 (0.7) 7.7 (1.8)† AA (% energy) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0) 0.1 (0.1)†SFAs (% energy) 9.8 (3.1) 10.2 (2.8) 10.3 (2.6) 10.9 (2.6)†MUFAs (% energy) 10.5 (2.6) 11.9 (2.5) 12.7 (2.7) 14.8 (3.4)†Trans fats (% energy) 1.3 (0.6) 1.5 (0.7) 1.5 (0.7) 1.6 (0.8)†

AA=arachidonic acid; ALA=α-linolenic acid; HPFS=Health Professionals Follow-Up Study; LA=linoleic

acid; MI=myocardial infarction; MUFA=monounsaturated fatty acids; NHS=Nurses’ Health Study;

PUFA=polyunsaturated fatty acids; SFA=saturated fatty acids.

*Values are age-adjusted means (standard deviation) for continues variables or percentages for

categorical variable.

†P for trend <0.05 analyzed by ANOVA for continuous variables or chi-square test for categorical

variables.

Page 26 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


4

Table 2 Associations of Dietary Fats with CVD Mortality and Total Mortality Among

Adults with Type 2 Diabetes Mellitus (comparison is isocaloric substitution for total

carbohydrates)

　 Quartiles of Fatty Acids Intake 　

　 Q1 Q2 Q3 Q4 P trendSaturated fatty acids Range ≤9.28 9.29-11.09 11.10-13.18 ≥13.19 Median, % energy 8.03 10.03 11.69 14.34 CVD Mortality No. of cases/person-years 138/31,501 166/31,535 154/31,558 188/31,400

Model 1: HR (95% CI)* 1.00 1.30 (1.03, 1.64) 1.26 (1.00, 1.59) 1.71 (1.37, 2.14) <.0001 Model 2: HR (95% CI)† 1.00 1.14 (0.88, 1.48) 0.98 (0.73, 1.33) 1.13 (0.80, 1.59) 0.62 Total Mortality No. of cases/person-years 594/31,087 628/31,103 613/31,184 667/30,988

Model 1: HR (95% CI)* 1.00 1.14 (1.02, 1.28) 1.21 (1.08, 1.36) 1.39 (1.24, 1.56) <.0001 Model 2: HR (95% CI)† 1.00 1.05 (0.92, 1.19) 1.03 (0.89, 1.19) 1.00 (0.85, 1.19) 0.88

Polyunsaturated fat intake Range ≤5.06 5.07-5.97 5.98-7.06 ≥7.07 Median, % energy 4.48 5.50 6.39 7.95 CVD Mortality No. of cases/person-years 198/31,335 176/31,490 147/31,569 125/31,600

Model 1: HR (95% CI)* 1.00 0.98 (0.79, 1.20) 0.84 (0.67, 1.04) 0.74 (0.59, 0.93) 0.0040 Model 2: HR (95% CI)† 1.00 0.99 (0.80, 1.23) 0.85 (0.67, 1.08) 0.76 (0.58, 0.99) 0.026 Total Mortality No. of cases/person-years 855/30,754 627/31,082 575/31,189 445/31,336

Model 1: HR (95% CI)* 1.00 0.81 (0.73, 0.90) 0.78 (0.70, 0.86) 0.61 (0.55, 0.69) <.0001 Model 2: HR (95% CI)† 1.00 0.86 (0.77, 0.95) 0.83 (0.74, 0.94) 0.68 (0.60, 0.78) <.0001

Monounsaturated fatty acids Range ≤10.77 10.78-12.61 12.62-14.64 ≥14.65 Median, % energy 9.51 11.66 13.37 16.01 CVD Mortality No. of cases/person-years 160/31,453 163/31,524 151/31,559 172/31,458


Model 1: HR (95% CI)* 1.00 1.01 (0.90, 1.12) 0.97 (0.87, 1.09) 1.08 (0.97, 1.21) 0.23 Model 2: HR (95% CI)† 1.00 0.93 (0.82, 1.05) 0.83 (0.72, 0.96) 0.90 (0.76, 1.06) 0.21

Page 27 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


5

Trans fatty acids Range ≤1.16 1.17-1.53 1.54-1.95 ≥1.96 Median, % energy 0.94 1.37 1.71 2.24 CVD Mortality No. of cases/person-years 140/31,499 149/31,542 165/31,546 192/31,407


Model 1: HR (95% CI)* 1.00 1.06 (0.94, 1.19) 1.21 (1.08, 1.36) 1.42 (1.27, 1.58) <.0001 Model 2: HR (95% CI)† 1.00 0.93 (0.82, 1.06) 1.01 (0.88, 1.15) 1.08 (0.94, 1.25) 0.11

CI=confidence interval; CVD=cardiovascular disease; HR=hazard ratio; PUFA=polyunsaturated fatty

acids.

*Adjusted for age (in months), gender and survey period.

†Further adjusted for ethnicity (White, others), body mass index at diagnosis (<23.0, 23.0-24.9,

25.0-29.9, 30.0-34.9, or ≥35.0, kg/m2), physical activity (0-0.4, 0.5-1.9, 2.0-3.4, 3.5-5.4, ≥5.5

hours/week), smoking status (never, past, current 1-14 cigarettes/d, or current ≥15 cigarettes/d),

smoking pack-years (0, <20, ≥20 pack-year), alcohol consumption (0, 0.1-4.9, 5.0-14.9, 15.0-29.9, or

≥30.0 g/d), multivitamin use (yes vs. no), current aspirin use (yes vs. no), family history of myocardial

infarction (yes vs. no), family history of diabetes (yes vs. no), history of hypercholesterolemia (yes vs.

no), history of hypertension (yes vs. no), diabetes duration (<5, 5-10, >10 years), total energy intake

(quartiles), dietary cholesterol (quartiles), and percentage of energy from dietary protein, and remaining

fatty acids where appropriate (PUFAs, MUFAs, trans fats, LA, AA, ALA, and marine n-3 PUFA, all

continuous variables). Comparison is isocaloric substitution for total carbohydrates in this model.

Page 28 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


6

Table 3 CVD Mortality and Total Mortality Among Adults with Type 2 Diabetes

Mellitus by Iso-calorically Substituting Specific PUFAs for Total Carbohydrates

　 Quartile of Fatty Acids Intake 　

　 Q1 Q2 Q3 Q4 P trendMarine n-3 PUFA* Interquartile range ≤0.05 0.06-0.09 0.10-0.17 ≥0.17 Median, % energy 0.03 0.07 0.13 0.25 CVD Mortality No. of cases/person-years 198/31,217 169/31,439 166/31,618 113/31,721

Model 1: HR (95% CI)† 1.00 0.83 (0.67, 1.02) 0.78 (0.63, 0.97) 0.54 (0.42, 0.68) <.0001 Model 2: HR (95% CI)‡ 1.00 0.92 (0.74, 1.14) 0.91 (0.72, 1.14) 0.69 (0.52, 0.90) 0.0067 Total Mortality No. of cases/person-years 791/30,674 703/30,974 606/31,239 402/31,475

Model 1: HR (95% CI)† 1.00 0.89 (0.80, 0.98) 0.77 (0.69, 0.85) 0.52 (0.46, 0.59) <.0001 Model 2: HR (95% CI)‡ 1.00 0.99 (0.89, 1.10) 0.91 (0.81, 1.03) 0.71 (0.62, 0.82) <.0001

α-Linolenic acid Interquartile range ≤0.45 0.46-0.54 0.55-0.65 ≥0.66 Median, % energy 0.40 0.49 0.58 0.76 CVD Mortality No. of cases/person-years 195/31,291 168/31,509 142/31,578 141/31,616

Model 1: HR (95% CI)† 1.00 0.96 (0.78, 1.18) 0.83 (0.66, 1.03) 0.84 (0.67, 1.04) 0.074 Model 2: HR (95% CI)‡ 1.00 1.06 (0.85, 1.32) 0.98 (0.76, 1.26) 1.13 (0.85, 1.51) 0.44 Total Mortality No. of cases/person-years 827/30,715 680/31,066 510/31,253 485/31,328

Model 1: HR (95% CI)† 1.00 0.89 (0.80, 0.99) 0.69 (0.62, 0.77) 0.65 (0.58, 0.73) <.0001 Model 2: HR (95% CI)‡ 1.00 0.98 (0.87, 1.09) 0.80 (0.70, 0.91) 0.88 (0.76, 1.02) 0.037

Linoleic acid Interquartile range ≤4.21 4.22-5.13 5.14-6.16 ≥6.17 Median, % energy 3.65 4.72 5.55 7.03 CVD Mortality No. of cases/person-years 209/31,299 161/31,509 149/31,545 127/31,641


Model 1: HR (95% CI)† 1.00 0.80 (0.72, 0.88) 0.76 (0.68, 0.84) 0.62 (0.55, 0.69) <.0001 Model 2: HR (95% CI)‡ 1.00 0.89 (0.80, 1.00) 0.90 (0.80, 1.02) 0.81 (0.69, 0.94) 0.0080

Arachidonic acid Interquartile range ≤0.06 0.07-0.08 0.09-0.10 ≥0.11

Page 29 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


7

Median, % energy 0.05 0.07 0.09 0.12 CVD Mortality No. of cases/person-years 168/31,429 142/31,531 165/31,533 171/31,502


Model 1: HR (95% CI)† 1.00 0.93 (0.83, 1.04) 1.01 (0.91, 1.13) 0.98 (0.88, 1.10) 0.90 Model 2: HR (95% CI)‡ 1.00 0.91 (0.80, 1.02) 1.05 (0.91, 1.21) 1.01 (0.85, 1.19) 0.64

CI=confidence interval; CVD=cardiovascular disease; HR=hazard ratio; PUFA=polyunsaturated fatty

acids.

*The sum of eicosapentaenoic acid and docosahexaenoic acid.

†Adjusted for age (in months), gender and survey period.

‡Further adjusted for white race (yes vs. no), body mass index at diagnosis (<23.0, 23.0-24.9, 25.0-29.9,

30.0-34.9, or ≥35.0), physical activity (0-0.4, 0.5-1.9, 2.0-3.4, 3.5-5.4, ≥5.5 h /week), smoking status

(never, past, current 1-14 cigarettes/d, or current ≥15 cigarettes/d), smoking pack-years (0, <20, ≥20

pack-year), alcohol consumption (0, 0.1-4.9, 5.0-14.9, 15.0-29.9, or ≥30.0 g/d), multivitamin use (yes vs.

no), current aspirin use (yes vs. no), family history of myocardial infarction (yes vs. no), family history of

diabetes (yes vs. no), history of hypercholesterolemia (yes vs. no), history of hypertension (yes vs. no),

diabetes duration (<5, 5-10, >10 years), total energy intake (quartiles), dietary cholesterol (quartiles),

and percentage of energy from dietary protein, and remaining fatty acids where appropriate (PUFAs,

MUFAs, trans fats, LA, AA, ALA, and marine n-3 PUFA, all continuous variables).Comparison is

isocaloric substitution for total carbohydrates in this model.

Page 30 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


1

Fig 1-Multivariate-adjustedd hazard ratios (HRs) of CVD mortality and total mortality by isocaloric substitution of dietary fats for saturated fatty acids. Hazard ratios of CVD mortality (A) and total mortality (B) by iso-calorically replacing energy from SFAs with specific types of fats. AA=arachidonic acid; ALA=α-linolenic acid;

CI=confidence interval; CVD=cardiovascular disease; HR=hazard ratio; LA=linoleic acid;

MUFA=monounsaturated fatty acids; PUFA=polyunsaturated fatty acids; SFA=saturated

fatty acids. The model was adjusted for age (in months), gender, survey period, white race

(yes vs. no), body mass index at diagnosis (<23.0, 23.0-24.9, 25.0-29.9, 30.0-34.9, or

≥35.0), physical activity (0-0.4, 0.5-1.9, 2.0-3.4, 3.5-5.4, ≥5.5 h /week), smoking status

(never, past, current 1-14 cigarettes/d, or current ≥15 cigarettes/d), smoking pack-years (0,

<20, ≥20 pack-year), alcohol consumption (0, 0.1-4.9, 5.0-14.9, 15.0-29.9, or ≥30.0 g/d),

multivitamin use (yes vs. no), current aspirin use (yes vs. no), family history of myocardial

infarction (yes vs. no), family history of diabetes (yes vs. no), history of

hypercholesterolemia (yes vs. no), history of hypertension (yes vs. no), diabetes duration

(<5, 5-10, >10 years), total energy intake (quartiles), dietary cholesterol (quartiles), and

percentage of energy from dietary protein, carbohydrates and remaining fatty acids where

appropriate (PUFAs, MUFAs, trans fats, LA, AA, ALA, and marine n-3 PUFA, all continuous

variables).

Page 31 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


1

Appendix: Supplementary figure and tables

Content

Supplementary table 1. Characteristics of patients with type 2 diabetes mellitus at diagnosis

according to saturated fat or monounsaturated fat intake ..........................................................2

Supplementary table 2. Associations between monounsaturated fat from plant or animal

source with mortality from CVD, cancer and all causes among patients with type 2 diabetes

mellitus (isocaloric substitution for total carbohydrates)...........................................................5

Supplementary table 3. Associations between dietary fats and cancer mortality among

patients with type 2 diabetes mellitus (isocaloric substitution for total carbohydrates) ............7

Supplementary table 4. Associations between specific PUFA with mortality from cancer

among patients with type 2 diabetes mellitus (isocaloric substitution for total carbohydrates) 9

Supplementary table 5. Sensitivity analyses ............................................................................11

Supplementary figure 1. Multivariable hazard ratios (HRs) of cancer mortality isocaloric

substitution of dietary fats for saturated fatty acids .................................................................17

Page 32 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


2

Supplementary table 1. Characteristics of patients with type 2 diabetes mellitus at diagnosis according to saturated fat or

monounsaturated fat intake*

　 Quartile of Saturated Fat Intake 　 Quartile of Monounsaturated Fat Intake

Characteristic 1 2 3 4 　 1 2 3 4

NHS (N=9,053)

Age (years) 64.4 (10.2) 62.9 (11.0) 62.9 (11.2) 63.5 (11.3) 64.8 (10.5) 63.3 (10.8) 62.4 (11.2) 63.0 (11.1)White (%) 93.9 96.7 97.5 97.6 95.1 95.9 97.4 96.4Body Mass Index (kg/m2) 28.7 (5.9) 30.0 (6.2) 30.3 (6.3) 30.7 (6.5) 29.2 (6.0) 29.8 (6.2) 30.3 (6.5) 30.1 (6.3)

Alcohol consumption (g/day) 3.7 (9.3) 3.3 (7.5) 2.7 (6.3) 2.6 (6.6) 3.5 (8.9) 2.9 (6.9) 3.0 (7.0) 3.3 (7.6)Physical activity (hours/week) 1.9 (2.9) 1.5 (2.6) 1.4 (2.4) 1.3 (2.4) 1.7 (2.8) 1.5 (2.4) 1.4 (2.6) 1.5 (2.6)Current smoker (%) 11.9 11.7 12.7 13.7 12.2 11.7 12.7 13.3Smoking pack-year 24.1 (20.6) 25.2 (20.3) 24.9 (21.1) 28.3 (23.0) 25.1 (20.9) 24.7 (21.1) 25.7 (21.6) 26.1 (21.3)Multivitamin use (%) 54.6 51.8 49.4 48.2 52 50.8 52.4 49.8Family history of diabetes (%) 48.6 52.1 49.5 48.1 49.8 51 49.6 47.6Family history of MI (%) 28.9 29.1 27.2 28.8 28.2 28.3 28 29.4Family history of cancer (%) 13.6 13.9 13.4 13.1 13.5 14.2 13.8 12.5Aspirin use (%) 55.2 56 54.3 52.5 55.3 53.1 54.8 54.5Hypoglycemic medication use (%) 38.4 46.2 47.4 50.8 40.6 44.3 46.9 48.8Duration of diabetes (years) 4.6 (4.3) 4.6 (4.0) 4.7 (3.9) 5.0 (4.6) 4.5 (4.2) 4.7 (4.3) 4.6 (4.0) 4.9 (4.4)Hypercholesterolemia (%) 64.8 63.5 63.5 62.6 62.7 63.1 63.4 66.4Hypertension (%) 73.3 75.7 76.6 75.9 73.9 75.7 75.3 76.1Total energy intake (kcal/day) 1602.0 (496.2) 1683.2 (515.0) 1705.5 (530.7) 1705.9 (537.9) 1590.8 (501.8) 1664.7 (509.6) 1710.8 (513.2) 1739.9 (547.0)

Carbohydrate (% energy) 54.2 (8.3) 48.2 (7.0) 44.9 (7.4) 39.8 (8.9) 54.8 (7.8) 48.5 (6.4) 44.7 (6.8) 38.9 (8.5)Protein (% energy) 19.4 (3.9) 19.4 (3.5) 19.1 (3.4) 19.1 (3.5) 19.4 (4.0) 19.2 (3.5) 19.1 (3.3) 19.1 (3.5)Total fat (% energy) 27.0 (5.4) 32.9 (4.7) 36.4 (5.3) 41.2 (6.4) 26.3 (4.6) 32.8 (3.7) 36.5 (4.3) 42.2 (6.1)

Page 33 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


3

Total PUFA (% energy) 5.8 (2.0) 6.3 (1.9) 6.6 (2.0) 6.4 (2.0) 5.2 (1.5) 6.2 (1.6) 6.7 (1.9) 7.4 (2.3)n-3 PUFA (% energy) 0.8 (0.4) 0.8 (0.3) 0.8 (0.3) 0.7 (0.3) 0.7 (0.3) 0.7 (0.3) 0.8 (0.4) 0.8 (0.4) ALA (% energy) 0.6 (0.4) 0.6 (0.3) 0.6 (0.3) 0.6 (0.2) 0.5 (0.3) 0.6 (0.3) 0.7 (0.3) 0.7 (0.3) Marine n-3 PUFA (% energy) 0.2 (0.2) 0.1 (0.2) 0.1 (0.1) 0.1 (0.1) 0.1 (0.2) 0.1 (0.2) 0.1 (0.1) 0.1 (0.2)n-6 PUFA (% energy) 5.1 (1.8) 5.6 (1.8) 5.8 (1.9) 5.5 (2.0) 4.5 (1.4) 5.4 (1.6) 5.9 (1.8) 6.6 (2.2) LA (% energy) 5.0 (1.8) 5.5 (1.8) 5.7 (1.9) 5.4 (2.0) 4.5 (1.4) 5.3 (1.6) 5.8 (1.8) 6.5 (2.2) AA (% energy) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0)SFA (% energy) 8.4 (2.0) 11.0 (1.9) 12.8 (2.3) 15.9 (3.2) 9.2 (2.5) 11.5 (2.7) 12.6 (3.1) 14.1 (4.1)MUFA (% energy) 10.3 (2.7) 12.7 (2.8) 14.1 (3.2) 15.8 (3.9) 9.4 (1.8) 12.3 (1.5) 14.2 (1.9) 17.7 (3.2)Trans-fat (% energy) 1.2 (0.6) 1.5 (0.6) 1.7 (0.7) 1.9 (0.7) 1.2 (0.5) 1.6 (0.6) 1.7 (0.7) 1.8 (0.9)　　　　　

HPFS (N=2,211)Age (years) 67.9 (9.0) 67.8 (8.5) 67.7 (8.2) 67.6 (8.2) 68.5 (9.3) 67.9 (8.5) 67.5 (8.6) 67.3 (7.9)White (%) 87 91.4 92.3 90.2 88.1 90 90.8 91.2Body Mass Index (kg/m2) 27.8 (4.5) 28.6 (4.7) 28.8 (5.0) 29.3 (4.8) 28.3 (5.0) 28.3 (4.4) 28.7 (4.8) 28.4 (4.7)

Alcohol consumption (g/day) 10.9 (16.0) 10.2 (14.3) 8.0 (12.0) 6.9 (11.8) 10.4 (16.9) 9.5 (13.0) 9.2 (13.7) 8.5 (12.0)Physical activity (hours/week) 3.6 (5.5) 3.5 (6.6) 2.9 (7.9) 2.7 (5.4) 3.8 (8.3) 3.0 (5.2) 3.3 (5.6) 3.2 (6.1)Current smoker (%) 3.9 4.6 4.9 7 4.5 4.3 4.4 5.9Smoking pack-year 15.0 (19.1) 16.9 (20.5) 17.2 (20.7) 20.7 (24.0) 15.5 (20.9) 16.9 (20.1) 17.3 (21.2) 19.2 (21.9)Multivitamin use (%) 61 55.9 55.3 58.6 58.3 58.3 53.5 61.8Family history of diabetes (%) 41 38.6 36.4 38.3 39.8 40.5 38.1 37.3Family history of MI (%) 35.8 34.7 26.2 30 36.1 33.8 31.3 28.9Family history of cancer (%) 35.9 36.4 36.4 35.2 36.3 35.8 37.9 35.8Aspirin use (%) 58.2 57.7 56.3 53.9 56.3 60.3 52.3 58.7Hypoglycemic medication use (%) 22.2 22.2 23.1 26.7 21.3 22.8 23.5 25.3Duration of diabetes (years) 3.2 (1.4) 3.1 (1.0) 3.2 (1.2) 3.4 (1.6) 3.2 (1.3) 3.1 (1.2) 3.1 (1.0) 3.4 (1.5)Hypercholesterolemia (%) 71.3 65.7 59.6 63 68.6 68.8 63.9 65Hypertension (%) 73 73.2 70.6 71.1 72 72.7 71.9 71.1

Page 34 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


4

Total energy intake (kcal/day) 1842.4 (553.7) 1994.0 (602.2) 1995.7 (653.6) 2076.7 (683.4) 1818.9 (551.6) 1905.0 (599.0) 2077.9 (620.3) 2052.9 (658.9)

Carbohydrate (% energy) 53.0 (7.8) 48.0 (6.7) 44.5 (6.4) 39.3 (7.7) 55.0 (7.5) 49.5 (5.4) 46.1 (5.6) 39.9 (7.5)Protein (% energy) 18.9 (3.7) 18.4 (3.2) 18.8 (3.2) 19.0 (3.3) 19.0 (3.9) 18.7 (3.3) 18.6 (3.1) 18.8 (3.2)Total fat (% energy) 26.7 (5.1) 32.1 (4.3) 35.8 (4.5) 40.7 (5.4) 24.6 (3.7) 30.5 (2.4) 34.2 (2.8) 40.4 (5.2)Total PUFA (% energy) 6.1 (2.1) 6.6 (2.0) 6.8 (2.0) 6.9 (2.0) 5.3 (1.4) 6.2 (1.5) 6.7 (1.8) 7.9 (2.4)n-3 PUFA (% energy) 0.8 (0.4) 0.8 (0.4) 0.8 (0.3) 0.8 (0.3) 0.7 (0.3) 0.8 (0.3) 0.8 (0.4) 0.9 (0.5) ALA (% energy) 0.6 (0.3) 0.6 (0.3) 0.6 (0.3) 0.6 (0.2) 0.5 (0.2) 0.6 (0.2) 0.6 (0.3) 0.7 (0.4) Marine n-3 PUFA (% energy) 0.2 (0.2) 0.2 (0.2) 0.2 (0.2) 0.1 (0.2) 0.2 (0.2) 0.2 (0.2) 0.2 (0.2) 0.2 (0.2)n-6 PUFA (% energy) 5.5 (1.8) 6.0 (1.8) 6.3 (1.8) 6.3 (1.9) 4.7 (1.2) 5.6 (1.3) 6.1 (1.6) 7.2 (2.1) LA (% energy) 5.4 (1.8) 5.9 (1.8) 6.2 (1.8) 6.2 (1.8) 4.7 (1.2) 5.5 (1.3) 6.0 (1.6) 7.1 (2.1) AA (% energy) 0.1 (0.1) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0) 0.1 (0.0) 0.1 (0.1)SFA (% energy) 7.5 (1.3) 9.9 (0.6) 11.6 (0.7) 14.6 (1.9) 7.9 (1.9) 9.9 (1.7) 11.1 (2.3) 12.3 (2.7)MUFA (% energy) 10.5 (2.7) 12.6 (2.7) 14.1 (2.9) 15.6 (3.0) 9.0 (1.4) 11.5 (0.5) 13.3 (0.6) 16.8 (2.6)Trans-fat (% energy) 1.2 (0.6) 1.5 (0.7) 1.7 (0.7) 1.9 (0.7) 　 1.1 (0.5) 1.5 (0.6) 1.7 (0.7) 1.7 (0.8)

AA=arachidonic acid; ALA=α-linolenic acid; HPFS=Health Professionals Follow-Up Study; LA=linoleic acid; MI=myocardial infarction;

MUFA=monounsaturated fatty acids; NHS=Nurses’ Health Study; PUFA=polyunsaturated fatty acids; SFA=saturated fatty acids.

*Values are age-adjusted means (SD) or percentages.

Page 35 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


5

Supplementary table 2. Associations between monounsaturated fat from plant or animal

source with mortality from CVD, cancer and all causes among patients with type 2

diabetes mellitus (isocaloric substitution for total carbohydrates)


　 Q1 Q2 Q3 Q4 P trend

Plant source

Interquartile range ≤4.24 4.25-5.69 5.70-7.62 ≥7.63

Median, % energy 3.35 4.99 6.54 9.35

CVD mortality

No. of cases 162/27,935 155/28,025 145/28,045 101/28,124

Multivariate HR (95% CI)* 1.00 1.01 (0.79, 1.29) 1.03 (0.79, 1.35) 0.84 (0.61, 1.15) 0.26

Cancer mortality

No. of cases 106/27,989 103/28,077 104/28,086 93/28,132

Multivariate HR (95% CI) 1.00 1.09 (0.81, 1.47) 1.13 (0.82, 1.55) 1.12 (0.79, 1.60) 0.57

Total mortality

No. of cases 684/27,470 595/27,647 524/27,709 422/27,843


Animal source

Interquartile range ≤4.35 4.36-5.75 5.76-7.48 ≥7.49

Median, % energy 3.47 5.05 6.42 8.82

CVD mortality

No. of cases 122/28,040 130/28,101 118/28,095 193/27,893


Cancer mortality

No. of cases 98/28,062 104/28,128 86/28,125 118/27,969


Total mortality

No. of cases 504/27,700 472/27,788 521/27,751 728/27,431

Page 36 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


6


CI=confidence interval; CVD=cardiovascular disease; HR=hazard ratio.

*The multivariate model was adjusted for age (in months), gender, survey period, white race

(yes vs. no), body mass index at diagnosis (<23.0, 23.0-24.9, 25.0-29.9, 30.0-34.9, or ≥35.0),

physical activity (0-0.4, 0.5-1.9, 2.0-3.4, 3.5-5.4, ≥5.5 h /week), smoking status (never, past,

current 1-14 cigarettes/d, or current ≥15 cigarettes/d), smoking pack-years (0, <20, ≥20

pack-year), alcohol consumption (0, 0.1-4.9, 5.0-14.9, 15.0-29.9, or ≥30.0 g/d),



hypercholesterolemia (yes vs. no), history of hypertension (yes vs. no), diabetes duration (<5,

5-10, >10 years), total energy intake (quartiles), dietary cholesterol (quartiles), and

percentage of energy from dietary protein, and remaining fatty acids where appropriate

(SFAs, PUFAs, animal MUFAs, plant MUFAs, trans fats, all continuous variables). The

model for cancer mortality also included family history of cancer (yes vs. no).

Page 37 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


7

Supplementary table 3. Associations between dietary fats and cancer mortality among

patients with type 2 diabetes mellitus (isocaloric substitution for total carbohydrates)


　 Q1 Q2 Q3 Q4 P trendSaturated fat Interquartile range ≤9.28 9.29-11.09 11.10-13.18 ≥13.19 Median, % energy 8.03 10.03 11.69 14.34 Cancer mortality No. of cases/person-years 119/31,518 116/31,582 102/31,626 114/31,470 Model 1: HR (95% CI)* 1.00 1.01 (0.78, 1.31) 1.00 (0.76, 1.30) 1.16 (0.89, 1.50) 0.29 Model 2: HR (95% CI)† 1.00 1.04 (0.77, 1.40) 0.96 (0.67, 1.35) 0.99 (0.67, 1.47) 0.90

Polyunsaturated fat Interquartile range ≤5.06 5.07-5.97 5.98-7.06 ≥7.07 Median, % energy 4.48 5.50 6.39 7.95 Cancer mortality No. of cases/person-years 141/31,399 100/31,567 115/31,597 95/31,633 Model 1: HR (95% CI)* 1.00 0.76 (0.59, 0.99) 0.91 (0.71, 1.17) 0.75 (0.58, 0.98) 0.085 Model 2: HR (95% CI)† 1.00 0.81 (0.62, 1.06) 0.95 (0.72, 1.25) 0.74 (0.55, 1.01) 0.11

Monounsaturated fat Interquartile range ≤10.77 10.78-12.61 12.62-14.64 ≥14.65 Median, % energy 9.51 11.66 13.37 16.01 Cancer mortality No. of cases/person-years 125/31,485 99/31,591 109/31,612 118/31,508 Model 1: HR (95% CI)* 1.00 0.81 (0.62, 1.06) 0.93 (0.71, 1.21) 1.06 (0.82, 1.37) 0.48 Model 2: HR (95% CI)† 1.00 0.88 (0.64, 1.19) 0.98 (0.70, 1.39) 1.09 (0.74, 1.60) 0.49

Trans-fat Interquartile range ≤1.16 1.17-1.53 1.54-1.95 ≥1.96 Median, % energy 0.94 1.37 1.71 2.24 Cancer mortality No. of cases/person-years 136/31,490 94/31,615 104/31,604 117/31,486 Model 1: HR (95% CI)* 1.00 0.74 (0.57, 0.97) 0.80 (0.62, 1.04) 0.93 (0.72, 1.19) 0.68 Model 2: HR (95% CI)† 1.00 0.69 (0.52, 0.91) 0.72 (0.54, 0.97) 0.74 (0.54, 1.01) 0.11

CI=confidence interval; HR=hazard ratio; PUFA=polyunsaturated fatty acids.


†Further adjusted for white race (yes vs. no), body mass index at diagnosis (<23.0, 23.0-24.9,

25.0-29.9, 30.0-34.9, or ≥35.0), physical activity (0-0.4, 0.5-1.9, 2.0-3.4, 3.5-5.4, ≥5.5 h

/week), smoking status (never, past, current 1-14 cigarettes/d, or current ≥15 cigarettes/d),

Page 38 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


8

smoking pack-years (0, <20, ≥20 pack-year), alcohol consumption (0, 0.1-4.9, 5.0-14.9,

15.0-29.9, or ≥30.0 g/d), multivitamin use (yes vs. no), current aspirin use (yes vs. no), family

history of myocardial infarction (yes vs. no), family history of diabetes (yes vs. no), family

history of cancer (yes vs. no), history of hypercholesterolemia (yes vs. no), history of

hypertension (yes vs. no), diabetes duration (<5, 5-10, >10 years), total energy intake

(quartiles), dietary cholesterol (quartiles), and percentage of energy from dietary protein, and

other fatty acids when appropriate (SFAs, PUFAs, MUFAs, and trans fats, all were

considered as continuous variables).

Page 39 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


9

Supplementary table 4. Associations between specific PUFA with mortality from cancer

among patients with type 2 diabetes mellitus (isocaloric substitution for total

carbohydrates)


　 Q1 Q2 Q3 Q4 P trendMarine n-3 PUFA Interquartile range ≤0.05 0.06-0.09 0.10-0.17 ≥0.17 Median, % energy 0.03 0.07 0.13 0.25 Cancer mortality No. of cases/person-years 119/31,291 122/31,483 123/31,672 87/31,750 Model 1: HR (95% CI)* 1.00 0.95 (0.74, 1.23) 0.94 (0.73, 1.22) 0.63 (0.48, 0.84) 0.0009 Model 2: HR (95% CI)† 1.00 1.02 (0.78, 1.33) 1.02 (0.78, 1.35) 0.72 (0.53, 0.99) 0.026

α-Linolenic acid Interquartile range ≤0.45 0.46-0.54 0.55-0.65 ≥0.66 Median, % energy 0.40 0.49 0.58 0.76 Cancer mortality No. of cases/person-years 149/31,342 130/31,557 79/31,629 93/31,668 Model 1: HR (95% CI)* 1.00 0.95 (0.74, 1.20) 0.60 (0.46, 0.79) 0.72 (0.56, 0.94) 0.0032 Model 2: HR (95% CI)† 1.00 1.00 (0.78, 1.29) 0.63 (0.46, 0.85) 0.75 (0.53, 1.05) 0.043

Linoleic acid Interquartile range ≤4.21 4.22-5.13 5.14-6.16 ≥6.17 Median, % energy 3.65 4.72 5.55 7.03 Cancer mortality No. of cases/person-years 135/31,374 103/31,565 112/31,586 101/31,670 Model 1: HR (95% CI)* 1.00 0.80 (0.61, 1.04) 0.88 (0.68, 1.13) 0.79 (0.60, 1.02) 0.12 Model 2: HR (95% CI)† 1.00 0.87 (0.66, 1.15) 1.04 (0.77, 1.40) 1.00 (0.71, 1.42) 0.83

Arachidonic acid Interquartile range ≤0.06 0.07-0.08 0.09-0.10 ≥0.11 Median, % energy 0.05 0.07 0.09 0.12 Cancer mortality No. of cases/person-years 128/31,481 105/31,567 125/31,577 93/31,571 Model 1: HR (95% CI)* 1.00 0.88 (0.68, 1.15) 1.05 (0.82, 1.35) 0.78 (0.59, 1.02) 0.14 Model 2: HR (95% CI)† 1.00 0.88 (0.66, 1.17) 1.07 (0.78, 1.47) 0.78 (0.53, 1.15) 0.28

CI=confidence interval; HR=hazard ratio; PUFA=polyunsaturated fatty acids.


Page 40 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


10

†Further adjusted for white race (yes vs. no), body mass index at diagnosis (<23.0, 23.0-24.9,

25.0-29.9, 30.0-34.9, or ≥35.0), physical activity (0-0.4, 0.5-1.9, 2.0-3.4, 3.5-5.4, ≥5.5 h

/week), smoking status (never, past, current 1-14 cigarettes/d, or current ≥15 cigarettes/d),

smoking pack-years (0, <20, ≥20 pack-year), alcohol consumption (0, 0.1-4.9, 5.0-14.9,

15.0-29.9, or ≥30.0 g/d), multivitamin use (yes vs. no), current aspirin use (yes vs. no), family

history of myocardial infarction (yes vs. no), family history of diabetes (yes vs. no), family

history of cancer (yes vs. no), history of hypercholesterolemia (yes vs. no), history of

hypertension (yes vs. no), diabetes duration (<5, 5-10, >10 years), total energy intake

(quartiles), dietary cholesterol (quartiles), and percentage of energy from dietary protein, and

remaining fatty acids where appropriate (SFAs, MUFAs, trans fats, LA, AA, ALA, and

marine n-3 PUFA, all continuous variables).

Page 41 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


11

Supplementary table 5. Sensitivity analyses*


　 Q1 Q2 Q3 Q4 P trendExcluding deaths within first 4 years since baselineSaturated fatty acids CVD mortality 1.00 1.09 (0.83, 1.42) 0.91 (0.67, 1.24) 1.02 (0.72, 1.44) 0.95 Cancer mortality 1.00 1.13 (0.82, 1.55) 1.01 (0.70, 1.45) 1.05 (0.70, 1.60) 0.95 Total mortality 1.00 1.05 (0.92, 1.19) 1.02 (0.88, 1.19) 0.98 (0.82, 1.16) 0.63

Polyunsaturated fatty acids CVD mortality 1.00 0.94 (0.75, 1.17) 0.81 (0.64, 1.03) 0.70 (0.53, 0.92) 0.0060 Cancer mortality 1.00 0.83 (0.63, 1.10) 0.88 (0.66, 1.17) 0.74 (0.54, 1.02) 0.10 Total mortality 1.00 0.85 (0.76, 0.94) 0.82 (0.73, 0.92) 0.67 (0.58, 0.77) <.0001

Marine n-3 PUFA CVD mortality 1.00 0.91 (0.73, 1.13) 0.93 (0.74, 1.17) 0.68 (0.51, 0.90) 0.0076 Cancer mortality 1.00 1.01 (0.77, 1.34) 1.03 (0.77, 1.37) 0.74 (0.53, 1.03) 0.049 Total mortality 1.00 0.98 (0.88, 1.09) 0.92 (0.82, 1.03) 0.71 (0.62, 0.82) <.0001

α-Linolenic acid CVD mortality 1.00 1.08 (0.86, 1.35) 1.01 (0.78, 1.30) 1.20 (0.89, 1.60) 0.26 Cancer mortality 1.00 0.98 (0.75, 1.28) 0.61 (0.45, 0.85) 0.74 (0.52, 1.06) 0.048 Total mortality 1.00 0.98 (0.87, 1.09) 0.82 (0.72, 0.93) 0.90 (0.77, 1.04) 0.083

Linoleic acid CVD mortality 1.00 0.84 (0.67, 1.06) 0.78 (0.61, 1.01) 0.69 (0.51, 0.94) 0.016 Cancer mortality 1.00 0.84 (0.63, 1.12) 0.98 (0.72, 1.34) 0.96 (0.67, 1.38) 0.96 Total mortality 1.00 0.87 (0.78, 0.98) 0.87 (0.76, 0.99) 0.76 (0.65, 0.89) 0.0009

Arachidonic acid CVD mortality 1.00 0.82 (0.63, 1.06) 1.05 (0.79, 1.40) 0.95 (0.68, 1.35) 0.92 Cancer mortality 1.00 0.90 (0.66, 1.22) 1.16 (0.84, 1.62) 0.77 (0.51, 1.16) 0.27 Total mortality 1.00 0.91 (0.80, 1.03) 1.07 (0.93, 1.23) 1.00 (0.84, 1.18) 0.74

Monounsaturated fatty acids CVD mortality 1.00 1.01 (0.78, 1.32) 0.88 (0.65, 1.21) 1.08 (0.76, 1.53) 0.67 Cancer mortality 1.00 0.92 (0.66, 1.28) 1.08 (0.75, 1.55) 1.15 (0.77, 1.73) 0.36 Total mortality 1.00 0.96 (0.84, 1.09) 0.86 (0.74, 1.00) 0.93 (0.78, 1.10) 0.39

Trans- fatty acids CVD mortality 1.00 0.96 (0.75, 1.25) 1.09 (0.83, 1.41) 1.17 (0.88, 1.56) 0.18 Cancer mortality 1.00 0.64 (0.47, 0.86) 0.71 (0.52, 0.96) 0.72 (0.52, 1.00) 0.14

Page 42 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


12

Total mortality 1.00 0.95 (0.83, 1.08) 1.05 (0.92, 1.20) 1.13 (0.98, 1.30) 0.030

Further adjusted for hypoglycemic medication†Saturated fatty acids CVD mortality 1.00 1.13 (0.87, 1.46) 0.98 (0.72, 1.32) 1.11 (0.79, 1.56) 0.67 Cancer mortality 1.00 1.04 (0.77, 1.40) 0.95 (0.67, 1.35) 0.99 (0.67, 1.46) 0.88 Total mortality 1.00 1.03 (0.91, 1.17) 1.01 (0.87, 1.17) 0.98 (0.83, 1.17) 0.74







Trans- fatty acids CVD mortality 1.00 0.96 (0.75, 1.23) 1.02 (0.78, 1.32) 1.15 (0.87, 1.52) 0.23 Cancer mortality 1.00 0.68 (0.51, 0.91) 0.72 (0.54, 0.97) 0.74 (0.54, 1.01) 0.12 Total mortality 1.00 0.92 (0.82, 1.05) 1.00 (0.88, 1.14) 1.10 (0.95, 1.26) 0.066

Page 43 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


13

Excluding extreme BMISaturated fatty acids CVD mortality 1.00 1.15 (0.88, 1.52) 1.00 (0.72, 1.37) 1.12 (0.78, 1.60) 0.68 Cancer mortality 1.00 1.09 (0.80, 1.47) 0.98 (0.69, 1.40) 1.04 (0.70, 1.56) 0.93 Total mortality 1.00 1.04 (0.91, 1.19) 1.04 (0.89, 1.21) 1.00 (0.84, 1.19) 0.92







Trans- fatty acids CVD mortality 1.00 0.98 (0.75, 1.27) 1.06 (0.80, 1.39) 1.19 (0.89, 1.60) 0.17 Cancer mortality 1.00 0.73 (0.55, 0.98) 0.72 (0.53, 0.98) 0.76 (0.55, 1.04) 0.14 Total mortality 1.00 0.96 (0.84, 1.09) 1.01 (0.88, 1.16) 1.09 (0.94, 1.26) 0.14

Excluding prevalent T2D Saturated fatty acid CVD mortality 1.00 1.17 (0.87, 1.58) 1.05 (0.75, 1.48) 1.02 (0.69, 1.50) 0.82

Page 44 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


14

Cancer mortality 1.00 0.93 (0.67, 1.29) 1.00 (0.69, 1.45) 1.10 (0.72, 1.67) 0.52 Total mortality 1.00 0.98 (0.84, 1.13) 1.02 (0.87, 1.20) 0.95 (0.79, 1.14) 0.61

Polyunsaturated fat intake CVD mortality 1.00 0.99 (0.78, 1.27) 0.84 (0.64, 1.11) 0.82 (0.60, 1.11) 0.14 Cancer mortality 1.00 0.75 (0.56, 1.01) 1.02 (0.76, 1.37) 0.80 (0.58, 1.12) 0.40 Total mortality 1.00 0.84 (0.74, 0.94) 0.81 (0.71, 0.93) 0.69 (0.59, 0.80) <.0001

Marine ω-3 PUFA CVD mortality 1.00 0.99 (0.77, 1.26) 0.97 (0.75, 1.26) 0.73 (0.53, 1.00) 0.039 Cancer mortality 1.00 1.04 (0.79, 1.39) 1.05 (0.78, 1.40) 0.74 (0.53, 1.04) 0.050 Total mortality 1.00 1.00 (0.89, 1.13) 0.91 (0.80, 1.04) 0.73 (0.63, 0.85) <.0001




Monounsaturated fatty acid CVD mortality 1.00 0.98 (0.73, 1.31) 0.74 (0.53, 1.05) 0.99 (0.67, 1.45) 0.90 Cancer mortality 1.00 0.89 (0.63, 1.24) 0.94 (0.65, 1.36) 0.89 (0.59, 1.36) 0.71 Total mortality 1.00 0.96 (0.84, 1.11) 0.86 (0.73, 1.01) 0.87 (0.72, 1.04) 0.092

Trans fat CVD mortality 1.00 0.94 (0.71, 1.25) 1.01 (0.75, 1.35) 1.07 (0.78, 1.46) 0.55 Cancer mortality 1.00 0.68 (0.50, 0.92) 0.70 (0.51, 0.95) 0.70 (0.50, 0.97) 0.066 Total mortality 1.00 0.93 (0.81, 1.07) 1.02 (0.88, 1.18) 1.04 (0.89, 1.22) 0.35

Restricting to confirmed T2D casesSaturated fatty acid CVD 1.00 1.19 (0.89, 1.59) 1.02 (0.73, 1.43) 1.12 (0.76, 1.64) 0.74 Cancer 1.00 1.08 (0.77, 1.52) 1.10 (0.75, 1.62) 1.21 (0.78, 1.88) 0.38 Total 1.00 1.02 (0.88, 1.18) 1.05 (0.89, 1.24) 1.01 (0.84, 1.23) 0.91

Page 45 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


15

Polyunsaturated fat intake CVD 1.00 0.92 (0.72, 1.17) 0.71 (0.54, 0.93) 0.74 (0.56, 1.00) 0.023 Cancer 1.00 0.78 (0.58, 1.07) 1.02 (0.75, 1.38) 0.79 (0.56, 1.11) 0.34 Total 1.00 0.83 (0.74, 0.94) 0.80 (0.70, 0.92) 0.67 (0.58, 0.78) <.0001

Marine ω-3 PUFA CVD 1.00 1.02 (0.80, 1.30) 0.94 (0.73, 1.21) 0.80 (0.59, 1.08) 0.097 Cancer 1.00 1.15 (0.85, 1.55) 1.09 (0.80, 1.49) 0.81 (0.56, 1.15) 0.098 Total 1.00 1.05 (0.93, 1.18) 0.94 (0.83, 1.07) 0.77 (0.66, 0.90) 0.0001

α-Linolenic acid CVD 1.00 0.97 (0.76, 1.24) 0.87 (0.65, 1.15) 1.08 (0.78, 1.48) 0.69 Cancer 1.00 1.02 (0.77, 1.36) 0.61 (0.43, 0.87) 0.75 (0.51, 1.11) 0.079 Total 1.00 0.98 (0.87, 1.11) 0.80 (0.69, 0.92) 0.85 (0.72, 1.01) 0.033

Linoleic acid CVD 1.00 0.86 (0.67, 1.10) 0.77 (0.58, 1.02) 0.79 (0.56, 1.10) 0.13 Cancer 1.00 0.93 (0.69, 1.27) 1.02 (0.73, 1.43) 1.01 (0.68, 1.50) 0.88 Total 1.00 0.89 (0.78, 1.01) 0.86 (0.74, 0.99) 0.81 (0.68, 0.96) 0.013

Arachidonic acid CVD 1.00 0.93 (0.70, 1.23) 1.16 (0.85, 1.59) 1.15 (0.79, 1.67) 0.34 Cancer 1.00 0.95 (0.69, 1.31) 1.14 (0.80, 1.62) 0.85 (0.55, 1.32) 0.53 Total 1.00 0.92 (0.80, 1.05) 1.03 (0.88, 1.20) 1.02 (0.84, 1.23) 0.61

Monounsaturated fatty acid CVD 1.00 0.96 (0.72, 1.28) 0.77 (0.55, 1.09) 0.96 (0.65, 1.40) 0.82 Cancer 1.00 0.90 (0.64, 1.28) 0.92 (0.63, 1.36) 0.97 (0.63, 1.50) 0.98 Total 1.00 0.97 (0.84, 1.12) 0.83 (0.71, 0.98) 0.89 (0.74, 1.08) 0.18

Trans fat CVD 1.00 0.93 (0.70, 1.24) 1.10 (0.82, 1.46) 1.20 (0.88, 1.63) 0.14 Cancer 1.00 0.61 (0.45, 0.84) 0.67 (0.48, 0.92) 0.67 (0.48, 0.95) 0.063 Total 1.00 0.93 (0.80, 1.07) 1.03 (0.89, 1.19) 1.12 (0.96, 1.31) 0.050

CI=confidence interval; CVD=cardiovascular disease; HR=hazard ratio;

PUFA=polyunsaturated fatty acids.

*The multivariate model was adjusted for age (in months), gender, survey period, white race

(yes vs. no), body mass index at diagnosis (<23.0, 23.0-24.9, 25.0-29.9, 30.0-34.9, or ≥35.0),

physical activity (0-0.4, 0.5-1.9, 2.0-3.4, 3.5-5.4, ≥5.5 h /week), smoking status (never, past,

current 1-14 cigarettes/d, or current ≥15 cigarettes/d), smoking pack-years (0, <20, ≥20

pack-year), alcohol consumption (0, 0.1-4.9, 5.0-14.9, 15.0-29.9, or ≥30.0 g/d),

Page 46 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


16



hypercholesterolemia (yes vs. no), history of hypertension (yes vs. no), diabetes duration (<5,

5-10, >10 years), total energy intake (quartiles), dietary cholesterol (quartiles), and

percentage of energy from dietary protein, and remaining fatty acids where appropriate

(PUFAs, MUFAs, trans fats, LA, AA, ALA, and marine n-3 PUFA, all continuous variables).

The model for cancer mortality also included family history of cancer (yes vs. no).

†Hypoglycemic medication use (none, oral medication only, insulin or oral medication, or

both).

Page 47 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


17

Supplementary figure 1. Multivariable hazard ratios (HRs) of cancer mortality isocaloric

substitution of dietary fats for saturated fatty acids

AA=arachidonic acid; ALA=α-linolenic acid; CI=confidence interval; HR=hazard ratio; LA=linoleic acid;

PUFA=polyunsaturated fatty acids; MUFA=monounsaturated fatty acids; SFA=saturated fatty acids.

Multivariable hazard ratios of mortality associated with replacing the percentage of energy from saturated

fat by the same energy from specific types of fat were used. The model was adjusted for age (in months),

gender, survey period, white race (yes vs. no), body mass index at diagnosis (<23.0, 23.0-24.9, 25.0-29.9,

30.0-34.9, or ≥35.0), physical activity (0-0.4, 0.5-1.9, 2.0-3.4, 3.5-5.4, ≥5.5 h /week), smoking status

(never, past, current 1-14 cigarettes/d, or current ≥15 cigarettes/d), smoking pack-years (0, <20, ≥20

pack-year), alcohol consumption (0, 0.1-4.9, 5.0-14.9, 15.0-29.9, or ≥30.0 g/d), multivitamin use (yes vs.

no), current aspirin use (yes vs. no), family history of myocardial infarction (yes vs. no), family history of

diabetes (yes vs. no), family history of cancer (yes vs. no), history of hypercholesterolemia (yes vs. no),

history of hypertension (yes vs. no), diabetes duration (<5, 5-10, >10 years), total energy intake

(quartiles), dietary cholesterol (quartiles), and percentage of energy from dietary protein, carbohydrates,

and remaining fatty acids where appropriate (PUFAs, MUFAs, trans fats, LA, AA, ALA, and marine n-3

PUFA, all continuous variables).

Page 48 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

dietary fats and mortality among patients with type …...2018/12/14 · 11 elucidated, owing to...

Documents