Abstract
Objective
We examined the association between Mediterranean dietary pattern as measured by Alternate Mediterranean Diet Score (aMed) and risk of incident rheumatoid arthritis (RA) in US women.
Methods
We prospectively followed 83,245 participants from the Nurses’ Health Study (NHS, 1980–2008) and 91,393 participants from NHS II (1991–2009) who were initially free of baseline connective tissue diseases. Dietary information was obtained via validated food frequency questionnaires (FFQ) at baseline and approximately every 4 years during follow-up. The aMed was calculated according to the consumption status of 9 food components using cumulative average value. Time-varying Cox proportional hazards models were used to calculate hazard ratios (HR) for RA, seropositive RA and seronegative RA after adjustment for potential confounding factors. Results from 2 cohorts were pooled by an inverse variance–weighted, fixed-effects model.
Results
During 3,511,050 person-years of follow-up, 913 incident cases of RA were documented in the two cohorts. After adjustment for several lifestyle and dietary variables, in both cohorts, greater adherence to Mediterranean dietary pattern was not significantly associated with altered risk of RA. The pooled HR for women in the highest quartile of aMed score was 0.98 (95% CI: 0.80–1.20) compared with those in the bottom quartile. Similar non-significant results were observed for both seropositive and seronegative RA. We did not find significant associations between each individual food component (except for alcohol) of aMed score and risk of incident RA.
Conclusion
We did not find a significant association between Mediterranean dietary pattern and risk of RA in women.
The traditional Cretan Mediterranean diet characterized by a high consumption of fruit, vegetables, whole grains, legumes, fish, olive oil, less red meat and moderate alcohol is generally regarded as a healthy diet pattern (1). A meta-analysis of prospective cohort studies showed that greater adherence to a Mediterranean diet was associated with significantly lower mortality rate from cardiovascular disease, lower incidence of cancer, Parkinson’s disease and Alzheimer’s disease (2). It has been suggested that the beneficial effects of the Mediterranean diet pattern are mediated through improvements of inflammatory markers, lipid profile and blood pressure (3–5).
Several randomized controlled clinical trials have shown a beneficial effect of Mediterranean diet intervention on physical function and vitality among patients with existing RA (6,7). This beneficial effect was found to be associated with improved fatty acid profile but not related with levels of plasma antioxidants (8,9). Several case-control studies suggested that higher consumption of fish, olive oil and cooked vegetables all of which were key components of the Mediterranean diet was associated with reduced risk or lower severity of RA (10–12). However, recall bias may be a particular critical issue in the dietary assessment because individuals with RA may be more likely to misreport their actual food consumption in the past. The case-control design is also unable to measure the long-term effects of certain dietary factors. Reverse causation may be a possible bias in case-control studies because individuals with early symptoms might change their usual diet. In most cases, many important time-varying confounders are rarely sufficiently controlled.
To the best of our knowledge, no previous prospective cohort studies have evaluated the association between the overall Mediterranean dietary pattern and risk of developing RA. We therefore investigate the relationships between the Mediterranean diet dietary pattern represented by Alternate Mediterranean Diet Score (aMed) (13)and RA risk in 2 well-established large cohorts of middle-aged and old women, controlling for a series of lifestyle and dietary factors.
MATERIALS AND METHODS
Study design and participants
The Nurses’ Health Study (NHS) was a cohort study including 121,700 female registered nurses of age 30 to 55 years initiated at 1976. The Nurse’s Health Study II (NHS II) was a parallel cohort established in 1989 and consisted of 116,671 female registered nurses of age 25–42 years. The participants in both cohorts responded to a baseline questionnaire about their lifestyles and medical histories, and they were followed biennially through validated questionnaires that obtained updated information. Follow-up was complete for more than 90% for every 2-year period in the two cohorts (14,15). Dietary information was collected using a validated food-frequency questionnaire (FFQ) since 1980 in NHS and 1991 in NHS II and the information was updated approximately every 4 years during the follow up period. We used 1980 as baseline for NHS and 1991 for NHS II when the dietary information was first collected.
For this analysis, we included women who completed the 1980 FFQ in NHS and 1991 FFQ in NHS II with <70 missing items and total energy intake between 500 and 3500 kcal/d. We censored all women who reported psoriasis, psoriatic arthritis and connective tissue diseases, in which the diagnosis was not subsequently confirmed as RA at self-reported date. Participants with missing aMed score were excluded. After exclusions, this left a total of 83,245 NHS participants and 91,393 NHS II participants for the analysis. Women lost to follow-up were censored at their last response to questionnaires because incident cases could not subsequently be identified. All aspects of this study were approved by the Partners HealthCare Institutional Review Board.
RA Case Identification
The ascertainment of RA cases in the NHS and NHS II was a 2-step process. The connective tissue disease (CTD) screening questionnaire (CSQ) was mailed to participants who self-reported a new physician diagnosis of RA (16). Two board-certified rheumatologists trained in chart abstraction conducted independent medical record reviews according to American College of Rheumatology classification 1987 version for RA (17). The serologic phenotype of RA was determined by positive rheumatoid factors (RF) (available since baseline) or cyclic citrullinated protein (CCP) antibodies in the medical record (available since the early 1990’s). Seropositive RA comprises of the majority of RA cases, and about 50–80% of patients with RA have positive RF, ACPA,(anti-citrullinated peptide antibody) or both (18). Compared with seronegative RA, seropositive RA has been more strongly associated with environmental risk factors such as cigarette smoking, which may induce the immunologic reaction against citrullinated peptides (19) and is also associated with a poorer prognosis (20). Detailed RA assessment procedures have been described elsewhere (21,22).
Dietary Assessment
Self-reported FFQ were used to assess average food intake over the preceding year. Participants were asked the frequency of certain food consumption according to commonly used unit or portion size. Nine possible frequency of consumption responses, ranging from ‘never or less than once per month’ to ‘ ≥ 6 times per day’ were provided for each food. Total energy and nutrient intakes were calculated by summing energy or nutrient intakes from all foods. A previous validation study among participants of NHS found reasonably high correlation coefficients between FFQ and multiple dietary records for most of the dietary factors including the 9 components of the aMed score (23).
The aMed score was a modified version of Mediterranean diet scale from a previously published study (24). There were 9 components in the aMed score including vegetables (excluding potatoes), fruits, nuts, whole grains, legume, fish, ratio of monounsaturated to saturated fat, red and processed meats and alcohol (13). Participants who consumed alcohol between 5 and 15g/d were assigned 1 point, and otherwise they received 0 point. This represents approximately one 12-oz can of regular beer, 5 oz of wine or 1.5 oz of liquor. For other food components of the aMed score, women with intake above the median intake were assigned 1 point and they otherwise received 0 point. Therefore, participants would have maximum 9 points and minimum 0 point in aMed score and a higher score indicated stronger adherence to the Mediterranean diet. We used cumulative average estimates of aMed score to reflect long-term dietary habits and reduce measurement errors (25). Because patients with preclinical RA typically have early symptoms such as joint pain which may lead them to change their usual diet, we performed a lag analysis in which aMed score was used to predict incident RA that occurred at least 4 years later to reduce the possibility of such reverse causation. For example, to predict RA incidence during the 1994–1998 time period, we used cumulative aMed score calculated between 1980 and 1990 (excluding 1994 measure, the most recent exposure). We stopped updating dietary variables at the first report of cancer because changes in diet after development of cancer may confound the relation between diet and RA. In sensitivity analyses, however, we also examined the association for cumulative average intake that was continually updated even after the development of cancer.
Assessment of Covariates
All covariates information was self-reported on the mailed questionnaires administered every 2 years since 1976 in the NHS, and since 1989 in the NHS II. Information on age, body weight, height, smoking status, menopausal status, use of postmenopausal hormone therapy, multivitamin use, history of diabetes, physical activity, census tract family median income, parity, breast feeding status and age of menarche was collected in both cohorts. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters and it was categorized in 5 levels: <20, 20–22.9, 23–24.9, 25–29.9 and ≥ 30kg/m2. Smoking status was categorized as never, past, current 1–14 cigarettes/day, current ≥ 15 cigarettes/day. As a proxy of socioeconomic status, we included the 2000 U.S. Census tract median income for the nurses’ residence. The age at menarche was categorized as < 12 years old, 12 years old and >12 years old. The information about menopause status and hormone use was grouped into 4 levels: pre-menopausal, post-menopausal with never use, current use and past use. The parity and breast feeding status were also integrated into a single variable that had 4 levels: nulliparous, parous/no breastfeeding, parous/1–12 months breastfeeding and parous/ >12 months breastfeeding. Both history of diabetes and multivitamin use was dichotomized as yes or no. In the NHS, recreational physical activity was measured biennially beginning in 1986 with a validated questionnaire asking about the average time spent on 10 common activities. In the NHS II, similar measures were performed in 1989, 1991, 1997, 2001 and 2005. The information was summed and calculated as weekly energy expenditure in metabolic equivalent hours weighting each activity by its intensity level (26). Coffee and sugar-sweetened soda consumption were assessed every 4 years using validated FFQ and were cumulatively averaged to represent the long-term intake. The frequency of soda consumption had 4 levels: <1 month, 1–4/month, 2–6/week and>1/day, and for coffee consumption, they were <1/day, 1–2/day, 3–4/day and ≥ 4/day.
Statistical Analysis
We used time-varying Cox proportional hazards models to assess the association between the aMed score and risk of RA, including separate models for individual food components of this derived Mediterranean diet score. Total RA, seropositive and seronegative RA were separate outcomes in each analysis. Baseline characteristics were presented as mean±SD for continuous variables and percentage was used for categorical variables. Cumulative average aMed score was included in the model as time-varying exposure and updated until 2 questionnaires prior to RA diagnosis (excluding the most recent questionnaire), and similarly, age, smoking status, physical activity, postmenopausal hormone use, dietary variables, and BMI were included in the multivariate model as time-varying covariates. aMed score was first included in the model using quartiles and then the median value of each quartile was used as a continuous variable to calculate the p-value for trend. We adjusted for the following potential confounders which were updated every 2years: age, census tract median family income (quartiles), cigarette smoking pack-years (never, past, current 1–14 cigarettes/day, current ≥ 15 cigarettes/day), age at menarche (<12, 12, >12 years), parity and breast feeding (nulliparous, parous/no breastfeeding, parous/1–12 months breastfeeding, parous/ >12 months breastfeeding), hormone use (pre-menopausal, post-menopausal with never use, current use and past use), physical activity (0–3, 3–9, 9–18, 18–27, ≥ 27 METs /week), body mass index (<20, 20–22.9, 23–24.9, 25–29.9, ≥ 30kg/m2), diabetes history, multi-vitamin use, sugar-sweetened soda consumption, coffee consumption, and total energy (Kcal, quintiles). Two sensitivity analyses were performed to assess the robustness of the main results. First, we kept updating dietary information after self-reported cancer. Second, we used the baseline aMed score without updating to examine the association with RA risk.
Missing data was carried forward one cycle. The inverse-variance weighted fixed-effects model was used to combine the results from 2 cohorts. All statistical tests were 2-sided and performed by using SAS 9.2 for UNIX (SAS Institute Inc, Cary, NC).
RESULTS
During 3,516,236 person-years of follow-up (28 years for NHS and 20 years for NHS II), we documented 916 cases of RA (631 cases in NHS, 282 in NHS II). Table 1 shows the age-standardized baseline characteristics of the study population by quartiles of aMed score. For both cohorts, women who were in higher quartiles of aMed score tended to be older, have higher census tract median family income, higher levels of physical activity, higher total energy intake, and lower BMI. Participants with higher aMed scores were also more likely to start menarche before age 12, have multivitamin supplement, breast-feeding more than 12 month and have current hormone use and less likely be current smoker. Parous status and postmenopausal status tended to be similar across aMed score quartiles in both cohorts. As expected, the average red and processed meat consumption was lower in the higher aMed score quartiles, while fish, vegetable, fruit, legume and whole grain consumption was systematically higher in higher aMed score quartiles in both cohorts. The nuts consumption was very low in both cohorts and the variation between aMed score quartiles was small. The average values of monounsaturated to saturated fat ratio were also similar across aMed score quartiles.
Table 1.
Variables | Nurses’ Health Study (N=83,245) | Nurses’ Health Study II (N=91,393) | ||||||
---|---|---|---|---|---|---|---|---|
| ||||||||
Q1 | Q2 | Q3 | Q4 | Q1 | Q2 | Q3 | Q4 | |
No. of participants | 19,140 | 17,279 | 32,501 | 14,325 | 19,071 | 33,881 | 16,375 | 22,066 |
aMed score range | 0–2 | 3 | 4–5 | 6–9 | 1–2 | 3–4 | 5 | 6–9 |
Age (years) | 47.4±7.0 | 48.2±7.1 | 49.0±7.2 | 50.1±7.2 | 38.2±4.8 | 38.6±4.7 | 38.8±4.6 | 39.1±4.5 |
BMI (kg/m2) | 24.7±4.7 | 24.7±4.6 | 24.6±4.5 | 24.4±4.4 | 25.7±6.0 | 25.3±5.7 | 25.2±5.5 | 24.7±5.2 |
Census tract median family income $ (*1000) | 62.4±24.7 | 63.9±25.2 | 65.3±26.1 | 67.9±27.6 | 59.7±20.5 | 62.2±22.7 | 63.4±23.3 | 64.8±24.5 |
Physical activity(MET-h/wk) | 11.1±17.3 | 12.7±19.6 | 14.8±20.8 | 18.0±22.9 | 15.9±22.7 | 19.2±24.7 | 22.8±28.6 | 26.9±32.4 |
AHEI score | 45.7±10.2 | 48.5±10.3 | 51.7±10.4 | 56.6±10.3 | 34.6±7.8 | 41.4±8.4 | 46.9±8.4 | 53.1±9.3 |
Alcohol consumption (g/d) | 6.4±12.3 | 6.6±11.3 | 6.3±9.9 | 6.5±8.2 | 2.4±5.3 | 2.9±5.3 | 3.2±5.3 | 3.8±5.2 |
Current smoking, % | 34.4 | 29.7 | 24.9 | 19.2 | 14.6 | 11.7 | 9.8 | 8.7 |
Multivitamin use, % | 31.0 | 34.1 | 37.9 | 43.3 | 33.6 | 38.0 | 41.2 | 45.9 |
Menarche before age 12, % | 21.7 | 22.5 | 23.7 | 24.0 | 23.4 | 24.0 | 25.1 | 25.2 |
Parous, % | 92.4 | 92.7 | 92.4 | 92.8 | 69.9 | 69.8 | 70.0 | 68.2 |
Breastfeeding >= 12 months, % | 14.1 | 16.2 | 18.4 | 22.9 | 20.5 | 24.7 | 28.5 | 31.4 |
Postmenopausal, % | 41.2 | 41.5 | 41.7 | 41.2 | 5.2 | 4.6 | 4.7 | 4.5 |
Current hormone use, % | 15.5 | 15.5 | 16.3 | 17.3 | 37.4 | 36.8 | 36.7 | 36.1 |
Total energy intake (Kcal/d) | 1435±477 | 1498±493 | 1609±503 | 1717±482 | 1508±481 | 1705±507 | 1909±521 | 2087±525 |
Red/processed meat (servings/d) | 1.8±1.0 | 1.7±1.0 | 1.6±1.0 | 1.4±0.9 | 1.2±0.7 | 1.2±0.7 | 1.2±0.7 | 1.1±0.7 |
Fish (servings/d) | 0.2±0.2 | 0.3±0.3 | 0.4±0.5 | 0.7±0.7 | 0.2±0.2 | 0.3±0.2 | 0.3±0.3 | 0.4±0.3 |
Vegetable (servings/d) | 1.1±0.5 | 1.4±0.8 | 1.9±1.0 | 2.4±1.1 | 1.8±1.0 | 2.8±1.6 | 3.8±1.9 | 4.8±2.2 |
Fruit (servings/d) | 0.8±0.8 | 1.2±1.0 | 1.6±1.3 | 2.2±1.3 | 0.6±0.5 | 1.0±0.8 | 1.4±1.0 | 1.8±1.1 |
Legume (servings/d) | 0.3±0.2 | 0.4±0.3 | 0.5±0.4 | 0.7±0.4 | 0.2±0.2 | 0.3±0.3 | 0.4±0.3 | 0.6±0.4 |
Whole grain (servings/d) | 0.3±0.6 | 0.5±0.8 | 0.9±0.9 | 1.3±1.0 | 0.7±0.7 | 1.2±1.0 | 1.6±1.1 | 2.1±1.3 |
Nuts (servings/d) | 0.1±0.2 | 0.1±0.3 | 0.2±0.4 | 0.3±0.5 | 0.0±0.1 | 0.0±0.1 | 0.1±0.1 | 0.1±0.2 |
Monounsaturated to saturated fat ratio | 1.0±0.1 | 1.0±0.1 | 1.0±0.1 | 1.1±0.2 | 1.0±0.1 | 1.1±0.1 | 1.1±0.2 | 1.2±0.2 |
Values (except age) are standardized to the age distribution of the study population. Mean ± SD unless otherwise indicated. aMed = Alternate Mediterranean Diet Score; BMI= body mass index (calculated as weight in kilograms divided by height in meters squared); MET=metabolic equivalent task; AHEI=Alternate Healthy Eating Index.
After adjusting for potential confounders, we did not find a significant association between aMed score and RA among NHS and NHS II participants. In the pooled results, the HR was 0.98 (95%CI: 0.80–1.20, p for trend 0.85) for women in the highest quartiles of aMed score compared with those in the bottom quartiles. Similar insignificant results were observed for seropositive RA (HR: 1.10, 95%CI: 0.85–1.42, p for trend 0.51) and seronegative RA (HR: 0.80, 95%CI: 0.57–1.13, p for trend 0.60) (Table 2). We further examined the relationship between each component of the aMed score and risk of RA (Table 3). We found a modest association between greater legume intake and increased risk of developing RA (p for trend 0.04). In a separate study, we found that long-term moderate alcohol drinking was associated with a reduced risk of RA in NHS and NHSII (27). However, no other food components of aMed score were observed to be associated with RA in the multivariable adjusted models. In both sensitivity analyses that updated dietary information after diagnosis of cancer and use baseline aMed score, the estimates changed modestly but the associations remained unchanged (data not shown).
Table 2.
aMed Score
|
|||||
---|---|---|---|---|---|
Q1 | Q2 | Q3 | Q4 | p for trend | |
All RA | |||||
NHS | |||||
Cases/ Person-years | 164/510,785 | 139/508,567 | 173/531,429 | 155/524,681 | |
Age adjusted | 1.00 | 0.85(0.68,1.07) | 1.01(0.81,1.25) | 0.89(0.71,1.11) | 0.56 |
Multivariable adjusted | 1.00 | 0.87(0.69,1.09) | 1.04(0.83,1.30) | 0.94(0.73,1.20) | 0.91 |
NHS II | |||||
Cases/ Person-years | 75/315,806 | 72/460,984 | 54/300,158 | 81/358,641 | |
Age adjusted | 1.00 | 0.65(0.47,0.91) | 0.71(0.50,1.00) | 0.88(0.64,1.21) | 0.58 |
Multivariable adjusted | 1.00 | 0.71(0.51,0.98) | 0.80(0.55,1.16) | 1.09(0.75,1.57) | 0.60 |
NHS and NHS II pooled | |||||
Age adjusted | 1.00 | 0.78(0.65, 0.94) | 0.91(0.76, 1.10) | 0.89(0.74, 1.06) | 0.43 |
Multivariable adjusted | 1.00 | 0.81(0.67, 0.98) | 0.97(0.80, 1.18) | 0.98(0.80, 1.20) | 0.85 |
| |||||
Seropositive RA | |||||
NHS | |||||
Cases/ Person-years | 103/481,583 | 90/479,660 | 98/497,621 | 100/429,830 | |
Age adjusted | 1.00 | 0.89(0.67,1.19) | 0.92(0.70,1.22) | 0.94(0.71,1.24) | 0.74 |
Multivariable adjusted | 1.00 | 0.93(0.70,1.24) | 0.98(0.73,1.31) | 1.03(0.76,1.40) | 0.79 |
NHS II | |||||
Cases/ Person-years | 50/310,102 | 48/452,201 | 30/295,056 | 58/352,392 | |
Age adjusted | 1.00 | 0.66(0.44,0.98) | 0.58(0.37,0.92) | 0.96(0.66,1.41) | 0.75 |
Multivariable adjusted | 1.00 | 0.73(0.48,1.09) | 0.70(0.43,1.12) | 1.26(0.81,1.98) | 0.43 |
NHS and NHS II pooled | |||||
Age adjusted | 1.00 | 0.81(0.64, 1.02) | 0.81(0.64, 1.03) | 0.95(0.76, 1.19) | 0.65 |
Multivariable adjusted | 1.00 | 0.85(0.68, 1.08) | 0.89(0.69, 1.14) | 1.10(0.85, 1.42) | 0.51 |
| |||||
Seronegative RA | |||||
NHS | |||||
Cases/ Person-years | 61/509,353 | 49/507,020 | 75/529,827 | 55/522,969 | |
Age adjusted | 1.00 | 0.79(0.54,1.15) | 1.15(0.82,1.62) | 0.82(0.56,1.18) | 0.64 |
Multivariable adjusted | 1.00 | 0.78(0.53,1.15) | 1.14(0.79,1.63) | 0.80(0.53,1.20) | 0.61 |
NHS II | |||||
Cases /Person-years | 25/315,231 | 24/460,221 | 24/299,590 | 23/357,938 | |
Age adjusted | 1.00 | 0.64(0.36,1.12) | 0.95(0.54,1.67) | 0.73(0.41,1.29) | 0.60 |
Multivariable adjusted | 1.00 | 0.66(0.37,1.18) | 0.99(0.54,1.80) | 0.81(0.43,1.55) | 0.87 |
NHS and NHS II pooled | |||||
Age adjusted | 1.00 | 0.74(0.54, 1.01) | 1.09(0.82, 1.47) | 0.79(0.58, 1.08) | 0.50 |
Multivariable adjusted | 1.00 | 0.74(0.54, 1.02) | 1.10(0.80, 1.49) | 0.80(0.57, 1.13) | 0.60 |
aMed = Alternate Mediterranean Diet Score; NHS = Nurses’ Health Study; NHS II =Nurses’ Health Study II; RA= rheumatoid arthritis.
Adjusted for age, census tract median family income (quartiles), cigarette smoking status (never, past, current 1–14 cigarettes/day, current ≥15 cigarettes/day), age at menarche (<12, 12, >12 years), parity and breast feeding (nulliparous, parous/no breastfeeding, parous/1–12 months breastfeeding, parous/>12 months breastfeeding), hormone use (pre-menopausal, post-menopausal with never use, current use and past use), physical activity (0–3, 3–9, 9–18, 18–27, ≥27 METs /week), body mass index (<20, 20–22.9, 23–24.9, 25–29.9, ≥30kg/m2), multi-vitamin use, diabetes history, sugar-sweetened soda (<1 month, 1–4/month, 2–6/week, >1/day), coffee consumption (<1/day, 1–2/day, 3–4/day, >=4/day) and total energy (Kcal, quintiles).
Table 3.
aMed Score
|
|||||
---|---|---|---|---|---|
Q1 | Q2 | Q3 | Q4 | p for trend | |
Red/processed meat | |||||
NHS | |||||
Cases/ Person-years | 146/521,016 | 163/516,642 | 165/516,887 | 157/520,917 | |
Age adjusted | 1.00 | 1.12(0.90,1.41) | 1.14(0.91,1.43) | 1.08(0.86,1.35) | 0.60 |
Multivariable adjusted | 1.00 | 1.09(0.86,1.37) | 1.12(0.88,1.42) | 1.10(0.85,1.43) | 0.51 |
NHS II | |||||
Cases/ Person-years | 56/357,340 | 66/358,752 | 77/359,436 | 83/360,061 | |
Age adjusted | 1.00 | 1.16(0.81,1.66) | 1.35(0.96,1.91) | 1.46(1.04,2.05) | 0.02 |
Multivariable adjusted | 1.00 | 1.05(0.73,1.51) | 1.15(0.80,1.65) | 1.12(0.76,1.65) | 0.55 |
NHS and NHS II pooled | |||||
Age adjusted | 1.00 | 1.13(0.94, 1.37) | 1.20(0.99, 1.45) | 1.19(0.98, 1.43) | 0.10 |
Multivariable adjusted | 1.00 | 1.08(0.89, 1.31) | 1.13(0.92, 1.37) | 1.11(0.89, 1.37) | 0.39 |
| |||||
Fish | |||||
NHS | |||||
Cases/ Person-years | 117/466,509 | 168/548,258 | 179/565,017 | 167/495,678 | |
Age adjusted | 1.00 | 1.21(0.95,1.54) | 1.26(1.00,1.60) | 1.31(1.03,1.66) | 0.07 |
Multivariable adjusted | 1.00 | 1.18(0.93,1.51) | 1.26(0.99,1.60) | 1.29(1.00,1.64) | 0.10 |
NHS II | |||||
Cases/ Person-years | 80/361,761 | 60/352,959 | 60/363,665 | 82/357,204 | |
Age adjusted | 1.00 | 0.74(0.53,1.04) | 0.70(0.50,0.98) | 0.96(0.70,1.30) | 0.91 |
Multivariable adjusted | 1.00 | 0.74(0.53,1.03) | 0.69(0.49,0.97) | 0.95(0.69,1.32) | 0.88 |
NHS and NHS II pooled | |||||
Age adjusted | 1.00 | 1.03(0.85, 1.25) | 1.04(0.86, 1.26) | 1.16(0.96, 1.40) | 0.09 |
Multivariable adjusted | 1.00 | 1.01(0.83, 1.23) | 1.03(0.85, 1.26) | 1.15(0.95, 1.40) | 0.12 |
| |||||
Vegetables | |||||
NHS | |||||
Cases/ Person-years | 146/518,607 | 149/518,556 | 160/519,156 | 176/519,143 | |
Age adjusted | 1.00 | 1.00(0.80,1.26) | 1.07(0.85,1.34) | 1.16(0.93,1.44) | 0.15 |
Multivariable adjusted | 1.00 | 0.99(0.78,1.25) | 1.06(0.84,1.34) | 1.18(0.92,1.50) | 0.13 |
NHS II | |||||
Cases /Person-years | 72/357,029 | 66/358,288 | 60/359,973 | 84/360,299 | |
Age adjusted | 1.00 | 0.86(0.62,1.21) | 0.77(0.54,1.08) | 1.03(0.75,1.41) | 0.73 |
Multivariable adjusted | 1.00 | 0.86(0.61,1.21) | 0.76(0.53,1.10) | 1.04(0.73,1.49) | 0.67 |
NHS and NHS II pooled | |||||
Age adjusted | 1.00 | 0.96(0.79, 1.15) | 0.97(0.80, 1.17) | 1.11(0.93, 1.33) | 0.18 |
Multivariable adjusted | 1.00 | 0.94(0.78, 1.14) | 0.96(0.79, 1.17) | 1.13(0.92, 1.38) | 0.15 |
| |||||
Fruit | |||||
NHS | |||||
Cases/ Person-years | 153/520,416 | 164/515,583 | 181/519,863 | 133/519,600 | |
Age adjusted | 1.00 | 1.04(0.84,1.30) | 1.16(0.93,1.44) | 0.84(0.66,1.06) | 0.16 |
Multivariable adjusted | 1.00 | 1.07(0.86,1.35) | 1.21(0.97,1.52) | 0.91(0.70,1.17) | 0.48 |
NHS II | |||||
Cases /Person-years | 75/352,518 | 56/362,022 | 76/361,593 | 75/359,456 | |
Age adjusted | 1.00 | 0.70(0.50,0.99) | 0.95(0.69,1.31) | 0.91(0.66,1.26) | 0.90 |
Multivariable adjusted | 1.00 | 0.74(0.52,1.05) | 1.04(0.74,1.46) | 1.05(0.73,1.50) | 0.37 |
NHS and NHS II pooled | |||||
Age adjusted | 1.00 | 0.93(0.77, 1.12) | 1.09(0.91, 1.30) | 0.86(0.71, 1.04) | 0.25 |
Multivariable adjusted | 1.00 | 0.96(0.80, 1.16) | 1.16(0.96, 1.40) | 0.95(0.77, 1.17) | 0.86 |
| |||||
Legume | |||||
NHS | |||||
Cases/ Person-years | 145/516,719 | 160/535,523 | 176/505,554 | 150/517,666 | |
Age adjusted | 1.00 | 1.07(0.86,1.34) | 1.25(1.00,1.56) | 1.04(0.83,1.31) | 0.72 |
Multivariable adjusted | 1.00 | 1.09(0.87,1.37) | 1.29(1.03,1.62) | 1.13(0.89,1.44) | 0.31 |
NHS II | |||||
Cases /Person-years | 66/364,070 | 58/373,428 | 68/338,852 | 90/359,239 | |
Age adjusted | 1.00 | 0.82(0.57,1.16) | 1.03(0.73,1.44) | 1.28(0.93,1.76) | 0.03 |
Multivariable adjusted | 1.00 | 0.82(0.57,1.17) | 1.02(0.72,1.44) | 1.27(0.91,1.79) | 0.04 |
NHS and NHS II pooled | |||||
Age adjusted | 1.00 | 0,99(0.82, 1.20) | 1.18(0.98, 1.42) | 1.12(0.93, 1.35) | 0.11 |
Multivariable adjusted | 1.00 | 1.00(0.83, 1.22) | 1.20(0.99, 1.45) | 1.18(0.97, 1.43) | 0.04 |
| |||||
Whole grain | |||||
NHS | |||||
Cases/ Person-years | 158/533,912 | 159/530,808 | 148/472,824 | 166/539,917 | |
Age adjusted | 1.00 | 1.00(0.80,1.25) | 1.03(0.82,1.29) | 1.01(0.81,1.27) | 0.87 |
Multivariable adjusted | 1.00 | 1.00(0.80,1.25) | 1.06(0.84,1.34) | 1.08(0.85,1.36) | 0.49 |
NHS II | |||||
Cases /Person-years | 76/351,023 | 60/364,511 | 70/360,040 | 76/360,014 | |
Age adjusted | 1.00 | 0.75(0.53,1.05) | 0.86(0.62,1.20) | 0.94(0.68,1.29) | 0.90 |
Multivariable adjusted | 1.00 | 0.76(0.54,1.07) | 0.91(0.64,1.28) | 1.04(0.73,1.48) | 0.46 |
NHS and NHS II pooled | |||||
Age adjusted | 1.00 | 0.92(0.76, 1.10) | 0.97(0.81, 1.17) | 0.99(0.82, 1.18) | 0.84 |
Multivariable adjusted | 1.00 | 0.92(0.76, 1.11) | 1.01(0.83, 1.22) | 1.06(0.87, 1.30) | 0.32 |
| |||||
Nuts | |||||
NHS | |||||
Cases/ Person-years | 151/529,925 | 128/384,444 | 197/639,686 | 155/521,407 | |
Age adjusted | 1.00 | 1.14(0.89,1.44) | 1.07(0.86,1.32) | 1.02(0.81,1.27) | 0.75 |
Multivariable adjusted | 1.00 | 1.11(0.87,1.42) | 1.08(0.87,1.34) | 1.03(0.82,1.30) | 0.90 |
NHS II | |||||
Cases /Person-years | 74/379,761 | 56/347,896 | 75/373,176 | 77/334,756 | |
Age adjusted | 1.00 | 0.79(0.53,1.18) | 0.98(0.70,1.36) | 1.07(0.76,1.49) | 0.24 |
Multivariable adjusted | 1.00 | 0.78(0.52,1.16) | 0.97(0.70,1.36) | 1.11(0.78,1.56) | 0.15 |
NHS and NHS II pooled | |||||
Age adjusted | 1.00 | 1.03(0.84, 1.27) | 1.04(0.87, 1.25) | 1.03(0.86, 1.24) | 0.88 |
Multivariable adjusted | 1.00 | 1.01(0.82, 1.24) | 1.05(0.87, 1.25) | 1.06(0.87, 1.28) | 0.68 |
| |||||
Monounsaturated to saturated fat ratio | |||||
NHS | |||||
Cases/ Person-years | 163/518,408 | 164/518,646 | 153/518,216 | 151/520,192 | |
Age adjusted | 1.00 | 1.00(0.81,1.25) | 0.94(0.76,1.18) | 0.92(0.74,1.15) | 0.39 |
Multivariable adjusted | 1.00 | 1.00(0.80,1.24) | 0.94(0.76,1.18) | 0.93(0.74,1.16) | 0.43 |
NHS II | |||||
Cases /Person-years | 73/358,162 | 60/359,484 | 72/358,987 | 77/358,956 | |
Age adjusted | 1.00 | 0.79(0.56,1.12) | 0.92(0.66,1.28) | 0.95(0.69,1.32) | 0.99 |
Multivariable adjusted | 1.00 | 0.77(0.55,1.09) | 0.90(0.65,1.24) | 1.00(0.72,1.39) | 0.75 |
NHS and NHS II pooled | |||||
Age adjusted | 1.00 | 0.94(0.78, 1.13) | 0.94(0.78, 1.13) | 0.93(0.77, 1.12) | 0.50 |
Multivariable adjusted | 1.00 | 0.93(0.77, 1.11) | 0.93(0.77, 1.12) | 0.95(0.79, 1.14) | 0.66 |
aMed = Alternate Mediterranean Diet Score; NHS = Nurses’ Health Study; NHS II =Nurses’ Health Study II; RA= rheumatoid arthritis.
Adjusted for age, census tract median family income (quartiles), cigarette smoking status (never, past, current 1–14 cigarettes/day, current ≥15 cigarettes/day), age at menarche (<12, 12, >12 years), parity and breast feeding (nulliparous, parous/no breastfeeding, parous/1–12 months breastfeeding, parous/ >12 months breastfeeding), hormone use (pre-menopausal, post-menopausal with never use, current use and past use), physical activity (0–3, 3–9, 9–18, 18–27, ≥27 METs /week), body mass index (<20, 20–22.9, 23–24.9, 25–29.9, ≥30kg/m2), multi-vitamin use, diabetes history, sugar-sweetened soda (<1 month, 1–4/month, 2–6/week, >1/day), coffee consumption (<1/day, 1–2/day, 3–4/day, >=4/day) and total energy (Kcal, quintiles).
DISCUSSION
In the present large cohort of women with up to 28 years of follow-up, greater adherence to a Mediterranean dietary pattern, as measured by a higher aMed score, was not significantly associated with reduced risk of incident RA. Except for alcohol consumption, and legume intake, we did not observe significant associations between individual food component of the aMed score and risk of developing RA. The results for alcohol consumption were published elsewhere (27).
Although several studies have investigated the association between components of Mediterranean diet such as vegetables, fruits and fish and risk of RA, this is the first study, to our knowledge, specifically designed to examine whether the overall Mediterranean dietary pattern could play a role in RA development using two large prospective cohorts. The traditional Mediterranean diet is shown to be effective in lowering CVD risk and it is also hypothesized to have a possible protective effect for development of RA because of its anti-inflammatory properties (28). This anti-inflammatory effect is particularly relevant to the treatment and prevention RA which is a disease characterized by persistent synovitis, systematic inflammation, and auto-antibodies (18,29).
Red meat consumption may be biologically plausible to be associated with the development of RA, while fish, fruits and vegetables may be associated with decreased risk of RA, however, the epidemiological results are inconsistent. The link between red meat consumption and increased RA risk may be attributed to its rich source of iron, which has been shown to accumulate in rheumatoid synovial membrane and exacerbate synovial inflammation (30,31). A prospective nested case-control study found that higher intake of meat and total protein was associated with an increased risk of RA (32), but the results from NHS found that neither red meat nor total protein was associated with increased RA risk (33). Fruits and vegetables are two major sources of antioxidants and their anti-inflammatory effects may be relevant in RA prevention. Several case-control studies suggest that higher consumption of fruits and vegetables are associated with reduced risk of RA (11,12), however results from the Iowa Women’s Health Study, a large prospective cohort including 29,368 participants, did not observe a significant association for fruit intake, but a modest inverse association for cruciferous vegetables (34). Dietary fish and fish oil have been suggested to be protective for development of RA because of their long-chain omega-3 fatty acids which are precursors of anti-inflammatory eicosanoids (35). A large case-control study with 1,889 incident RA cases found that those with higher oily fish consumption had 20% reduced risk of RA compared with those with never or seldom intake (10), while another prospective cohort study showed that the protective effects were only restricted to fat fish, and the medium fat fish was associated with significantly increased risk of RA (36). In our recently study, we found a modest association between long-term moderate alcohol drinking and reduced risk of RA in NHS and NHSII (27). Compared to non-drinkers, women who consumed alcohol 5–9.9 gram/day had 22% lower risk of developing RA (HR: 0.78, 95% CI, 0.61–1.00) and the association was stronger for seropositive RA (HR: 0.69, 95% CI, 0.50–0.95). The underlying mechanisms remain to be clarified, but several studies suggested that alcohol consumption might play a role in immunologic regulation, which is able to diminishing the response to immunogens in animals as well as in humans, and to significantly suppressing the synthesis of proinflammatory cytokines and chemokines, such as tumor necrosis factor (TNF-α), interleukin 6(IL-6), and interleukin 8 (IL-8), both in vivo and in vitro in alveolar macrophages and human blood monocytes (37,38).
In the current study, we also examined whether legume, nuts and whole grain consumption could play a role in RA development as very few studies have been conducted to examine the associations between these foods and RA risk. Prospective studies have found dietary fiber intake to be inversely associated with inflammation biomarkers such as IL-6, TNF-α-2, plasma fibrinogen and hs-CRP (39). Thus, nuts and whole grain consumption might be beneficial for RA prevention. In addition, whole grains are also rich in antioxidants, including vitamin E, phytic acid, and selenium (40) which may attribute to the anti-inflammatory process. However, we did not observe a significant inverse association between these food items and RA risk in the current analysis. These insignificant findings were possibly because dietary fibers might have small-to-modest effects on RA risk that could not be detected with our sample size. We found a modest association of legume intake with increased risk of developing RA, but it remains unclear whether this is due to chance. It is suggested that dietary staples such as cereal grains and legumes contain lectins that have the ability to interact with components of the immune system which may facilitate the autoimmune process (41). More studies are warranted to confirm our findings
There were several potential interpretations for the null findings which have been hypothesized to be protective against RA risk. First, a previous study suggested that only cooked vegetables but not raw were associated with reduced risk of RA (11). However, we were not able to differentiate whether it was cooked or raw in the current study. Second, we did not collect information on certain types of fish (e.g., fat or medium fish). Thus the insignificant findings for vegetables and fish consumption might reflect a mixed of food with opposite disease prevention properties. Olive oil, which contains rich oleic acid that can be metabolized to eicosattrienoic acid with anti-inflammatory properties, is the primary source of monounsaturated fat in the traditional Mediterranean diet, however, in our cohorts the major sources of monounsaturated fat come from beef and other meats (13). Therefore, because the red meat also has potential pro-inflammatory effects, the insignificant results from monounsaturated/saturated fat ratio might also reflect a mixed food effects. We did not find fruit consumption was associated with reduced risk of RA in both cohorts. One possible explanation could be that not all kinds of fruits were effective in RA prevention. A recent study using our cohorts has found that only greater consumption of specific whole fruits such as blue berries, grapes and raisins were associated with reduced risk of type 2 diabetes while some others were even associated with increased type 2 diabetes risk (42). Another important reason that may contribute to our null findings is that the variation of aMed score is too small to detect any modest associations. This issue also applies to the individual food components as we found the mean values for nuts and monounsaturated to saturated fat ratio were very similar across aMed score quartiles.
Our study had several strengths. First, prospective assessment of diet and lifestyle information and high rate of follow-up reduces the possibility of selection and recall bias that may be the major methodological limitations in case-control studies. Second, a validated FFQ was used to collect dietary information every 4 years during the follow up and we cumulatively averaged the consumptions to reflect participants’ long-term intake and reduce measurement error. Third, we performed a lag analysis that excluded the most recent dietary information to prevent the possible reverse causation that people with early symptoms of RA were likely to change their usual dietary. Finally, although residual confounding is unavoidable in observational study, given our detailed and updated adjustment for potential confounders, it was unlikely that the results would be severely flawed. Despite these advantages, interpretation of the current data should be cautious because the study population consisted of all women with higher health awareness. Actually, the incidence of RA in our cohorts was lower than prior estimates from the Olmsted County, Minnesota’s hospital case-based reports (43). Hence our ability to assess the association with men and capture subjects with early symptoms, or who were not health professionals was limited, and the generalizability of our findings was limited to women with middle-age range. However, the rigorous assessment of cases and dietary information guaranteed the internal validity of the current analysis.
In conclusion, no significant association was found between overall Mediterranean dietary pattern and RA risk in these two large prospective cohorts. Because this is the first prospective report regarding the potential effects of Mediterranean diet on RA risk, our results need to be replicated in other populations.
Significance & Innovations.
Mediterranean diet is associated with reduced risk of some inflammatory diseases because of its anti-inflammatory properties. Clinical trials suggested beneficial effects of this dietary pattern on patients with existing rheumatoid arthritis.
This is the first study to evaluate the association between overall Mediterranean dietary pattern and risk of incident RA using prospective cohorts.
The greater adherence to Mediterranean diet might not have substantial benefits in reducing RA risk that might reflect mixed effects of the components of this dietary pattern. Except alcohol consumption, none of the components of the Mediterranean diet was found to be significantly associated with RA risk.
Acknowledgments
Supported by the National Institutes of Health under Award Number AR061362, AR049880, AR052403, AR047782, AR059073, CA87969, CA50385, and CA176726. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
We thank the participants in NHS and NHS II cohorts for their dedication and continued participation in these longitudinal studies, as well as NHS staffs in the Channing Division of Network Medicine, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School for their assistance with this project.
Footnotes
Conflict of Interests Disclosure: None reported.
AUTHOR CONTRIBUTIONS:
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Mr. Hu and Dr. Lu had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design. Hu and Lu.
Acquisition of data. Lu, Karlson and Costenbader.
Analysis and interpretation of data. Hu, Lu, Karlson, Costenbader, Gao and Hu.
References
- 1.Willett WC. The Mediterranean diet: science and practice. Public Health Nutr. 2007;9:105–110. doi: 10.1079/phn2005931. [DOI] [PubMed] [Google Scholar]
- 2.Sofi F, Cesari F, Abbate R, Gensini GF, Casini A. Adherence to Mediterranean diet and health status: meta-analysis. BMJ Br Med J. 2008:337. doi: 10.1136/bmj.a1344. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Tzima N, Pitsavos C, Panagiotakos DB, Skoumas J, Zampelas A, Chrysohoou C, et al. Mediterranean diet and insulin sensitivity, lipid profile and blood pressure levels, in overweight and obese people; the Attica study. Lipids Health Dis. 2007;6:22. doi: 10.1186/1476-511X-6-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Esposito K, Marfella R, Ciotola M, Di Palo C, Giugliano F, Giugliano G, et al. Effect of a mediterranean-style diet on endothelial dysfunction and markers of vascular inflammation in the metabolic syndrome: a randomized trial. JAMA. 2004;292:1440–1446. doi: 10.1001/jama.292.12.1440. [DOI] [PubMed] [Google Scholar]
- 5.Fitó M, Guxens M, Corella D, Sáez G, Estruch R, de la Torre R, et al. Effect of a traditional Mediterranean diet on lipoprotein oxidation: a randomized controlled trial. Arch Intern Med. 2007;167:1195–1203. doi: 10.1001/archinte.167.11.1195. [DOI] [PubMed] [Google Scholar]
- 6.Sköldstam L, Hagfors L, Johansson G. An experimental study of a Mediterranean diet intervention for patients with rheumatoid arthritis. Ann Rheum Dis. 2003;62:208–214. doi: 10.1136/ard.62.3.208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.McKellar G, Morrison E, McEntegart A, Hampson R, Tierney A, Mackle G, et al. A pilot study of a Mediterranean-type diet intervention in female patients with rheumatoid arthritis living in areas of social deprivation in Glasgow. Ann Rheum Dis. 2007;66:1239–1243. doi: 10.1136/ard.2006.065151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Hagfors L, Leanderson P, Sköldstam L, Andersson J, Johansson G. Antioxidant intake, plasma antioxidants and oxidative stress in a randomized, controlled, parallel, Mediterranean dietary intervention study on patients with rheumatoid arthritis. Nutr J. 2003;2:5. doi: 10.1186/1475-2891-2-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hagfors L, Nilsson I, Sköldstam L, Johansson G. Fat intake and composition of fatty acids in serum phospholipids in a randomized, controlled, Mediterranean dietary intervention study on patients with rheumatoid arthritis. Nutr Metab (Lond) 2005;2:26. doi: 10.1186/1743-7075-2-26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Rosell M, Wesley A-M, Rydin K, Klareskog L, Alfredsson L. Dietary fish and fish oil and the risk of rheumatoid arthritis. Epidemiology. 2009;20:896–901. doi: 10.1097/EDE.0b013e3181b5f0ce. [DOI] [PubMed] [Google Scholar]
- 11.Linos A, Kaklamani VG, Kaklamani E, Koumantaki Y, Giziaki E, Papazoglou S, et al. Dietary factors in relation to rheumatoid arthritis: a role for olive oil and cooked vegetables? Am J Clin Nutr. 1999;70:1077–1082. doi: 10.1093/ajcn/70.6.1077. [DOI] [PubMed] [Google Scholar]
- 12.Shapiro JA, Koepsell TD, Voigt LF, Dugowson CE, Kestin M, Nelson JL. Diet and rheumatoid arthritis in women: a possible protective effect of fish consumption. Epidemiology. 1996;7:256–263. doi: 10.1097/00001648-199605000-00007. [DOI] [PubMed] [Google Scholar]
- 13.Fung TT, Rexrode KM, Mantzoros CS, Manson JE, Willett WC, Hu FB. Mediterranean diet and incidence of and mortality from coronary heart disease and stroke in women. Circulation. 2009;119:1093–1100. doi: 10.1161/CIRCULATIONAHA.108.816736. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Schulze MB, Manson JE, Ludwig DS, Colditz GA, Stampfer MJ, Willett WC, et al. Sugar-sweetened beverages, weight gain, and incidence of type 2 diabetes in young and middle-aged women. JAMA. 2004;292:927–934. doi: 10.1001/jama.292.8.927. [DOI] [PubMed] [Google Scholar]
- 15.Fung TT, Malik V, Rexrode KM, Manson JE, Willett WC, Hu FB. Sweetened beverage consumption and risk of coronary heart disease in women. Am J Clin Nutr. 2009;89:1037–1042. doi: 10.3945/ajcn.2008.27140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Karlson EW, Sanchez-Guerrero J, Wright EA, Lew RA, Daltroy LH, Katz JN, et al. A connective tissue disease screening questionnaire for population studies. Ann Epidemiol. 1995;5:297–302. doi: 10.1016/1047-2797(94)00096-c. [DOI] [PubMed] [Google Scholar]
- 17.Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum. 1988;31:315–324. doi: 10.1002/art.1780310302. [DOI] [PubMed] [Google Scholar]
- 18.Scott DL, Wolfe F, Huizinga TWJ. Rheumatoid arthritis. Lancet. 2010;376:1094–108. doi: 10.1016/S0140-6736(10)60826-4. [DOI] [PubMed] [Google Scholar]
- 19.Szodoray P, Szabó Z, Kapitány A, Gyetvai Á, Lakos G, Szántó S, et al. Anti-citrullinated protein/peptide autoantibodies in association with genetic and environmental factors as indicators of disease outcome in rheumatoid arthritis. Autoimmun Rev. 2010;9:140–143. doi: 10.1016/j.autrev.2009.04.006. [DOI] [PubMed] [Google Scholar]
- 20.van der Linden MPM, van der Woude D, Ioan-Facsinay A, Levarht EWN, Stoeken-Rijsbergen G, Huizinga TWJ, et al. Value of anti–modified citrullinated vimentin and third-generation anti–cyclic citrullinated peptide compared with second-generation anti–cyclic citrullinated peptide and rheumatoid factor in predicting disease outcome in undifferentiated arthritis and rh. Arthritis Rheum. 2009;60:2232–2241. doi: 10.1002/art.24716. [DOI] [PubMed] [Google Scholar]
- 21.Karlson EW, Mandl LA, Hankinson SE, Grodstein F. Do breast-feeding and other reproductive factors influence future risk of rheumatoid arthritis? Results from the Nurses’ Health Study. Arthritis Rheum. 2004;50:3458–67. doi: 10.1002/art.20621. [DOI] [PubMed] [Google Scholar]
- 22.Karlson EW, Chibnik LB, Tworoger SS, Lee I-M, Buring JE, Shadick Na, et al. Biomarkers of inflammation and development of rheumatoid arthritis in women from two prospective cohort studies. Arthritis Rheum. 2009;60:641–52. doi: 10.1002/art.24350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Salvini S, Hunter DJ, Sampson L, Stampfer MJ, Colditz GA, Rosner B, et al. Food-based validation of a dietary questionnaire: the effects of week-to-week variation in food consumption. Int J Epidemiol. 1989;18:858–867. doi: 10.1093/ije/18.4.858. [DOI] [PubMed] [Google Scholar]
- 24.Trichopoulou A, Costacou T, Bamia C, Trichopoulos D. Adherence to a Mediterranean diet and survival in a Greek population. N Engl J Med. 2003;348:2599–2608. doi: 10.1056/NEJMoa025039. [DOI] [PubMed] [Google Scholar]
- 25.Hu FB, Stampfer MJ, Rimm E, Ascherio A, Rosner BA, Spiegelman D, et al. Dietary fat anc coronary heart disease: a comparison of approaches for adjusting for total energy intake and modeling repeated dietary measurements. Am J Epidemology. 1999;149:531–540. doi: 10.1093/oxfordjournals.aje.a009849. [DOI] [PubMed] [Google Scholar]
- 26.Ainsworth BE, Haskell WL, Leon AS, Jacobs DR, Montoye HJ, Sallis JF, et al. Compendium of physical activities: classification of energy costs of human physical activities. Med Sci Sports Exerc. 1993;25:71–80. doi: 10.1249/00005768-199301000-00011. [DOI] [PubMed] [Google Scholar]
- 27.Lu B, Solomon DH, Costenbader KH, Karlson EW. Alcohol Consumption and Risk of Incident Rheumatoid Arthritis in Women: a Prospective Study. Arthritis Rheumatol. 2014 doi: 10.1002/art.38634. Epub ahead. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Oliviero F, Punzi L, Spinella P. Mediterranean Food Pattern in Rheumatoid Arthritis. Curr Rheumatol Rev. 2009;5:233–240. [Google Scholar]
- 29.McInnes IB, Schett G. The Pathogenesis of Rheumatoid Arthritis. N Engl J Med. 2011;365:2205–2219. doi: 10.1056/NEJMra1004965. [DOI] [PubMed] [Google Scholar]
- 30.Winyard PG, Blake DR, Chirico S, Gutteridge JM, Lunec J. Mechanism of exacerbation of rheumatoid synovitis by total-dose iron-dextran infusion: in-vivo demonstration of iron-promoted oxidant stress. Lancet. 1987;1:69–72. doi: 10.1016/s0140-6736(87)91909-x. [DOI] [PubMed] [Google Scholar]
- 31.Blake DR, Hall ND, Bacon PA, Dieppe PA, Halliwell B, Gutteridge JM. The importance of iron in rheumatoid disease. Lancet. 1981;2:1142–1144. doi: 10.1016/s0140-6736(81)90590-0. [DOI] [PubMed] [Google Scholar]
- 32.Pattison DJ, Symmons DPM, Lunt M, Welch A, Luben R, Bingham SA, et al. Dietary risk factors for the development of inflammatory polyarthritis: evidence for a role of high level of red meat consumption. Arthritis Rheum. 2004;50:3804–3812. doi: 10.1002/art.20731. [DOI] [PubMed] [Google Scholar]
- 33.Benito-Garcia E, Feskanich D, Hu FB, Mandl LA, Karlson EW. Protein, iron, and meat consumption and risk for rheumatoid arthritis: a prospective cohort study. Arthritis Res Ther. 2007;9:R16. doi: 10.1186/ar2123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Cerhan JR, Saag KG, Merlino LA, Mikuls TR, Criswell LA. Antioxidant micronutrients and risk of rheumatoid arthritis in a cohort of older women. Am J Epidemiol. 2003;157:345–354. doi: 10.1093/aje/kwf205. [DOI] [PubMed] [Google Scholar]
- 35.Rennie KL, Hughes J, Lang R, Jebb SA. Nutritional management of rheumatoid arthritis: a review of the evidence. J Hum Nutr Diet. 2003;16:97–109. doi: 10.1046/j.1365-277x.2003.00423.x. [DOI] [PubMed] [Google Scholar]
- 36.Pedersen M, Stripp C, Klarlund M, Olsen SF, Tjønneland AM, Frisch M. Diet and risk of rheumatoid arthritis in a prospective cohort. J Rheumatol. 2005;32:1249–1252. [PubMed] [Google Scholar]
- 37.Grimaldi CM, Hill L, Xu X, Peeva E, Diamond B. Hormonal modulation of B cell development and repertoire selection. Molecular Immunology. 2005;42:811–820. doi: 10.1016/j.molimm.2004.05.014. [DOI] [PubMed] [Google Scholar]
- 38.Waldschmidt TJ, Cook RT, Kovacs EJ. Alcohol and inflammation and immune responses: summary of the 2006 Alcohol and Immunology Research Interest Group (AIRIG) meeting. Alcohol. 2008;42:137–142. doi: 10.1016/j.alcohol.2007.11.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Satija A, Hu FB. Cardiovascular Benefits of Dietary Fiber. Curr Atheroscler Rep. 2012;14:505–514. doi: 10.1007/s11883-012-0275-7. [DOI] [PubMed] [Google Scholar]
- 40.Slavin JL, Martini MC, Jacobs DR, Marquart L. Plausible mechanisms for the protectiveness of whole grains. Am J Clin Nutr. 1999;70:459S–463S. doi: 10.1093/ajcn/70.3.459s. [DOI] [PubMed] [Google Scholar]
- 41.Cordain L, Toohey L, Smith MJ, Hickey MS. Modulation of immune function by dietary lectins in rheumatoid arthritis. Br J Nutr. 2000;83:207–217. doi: 10.1017/s0007114500000271. [DOI] [PubMed] [Google Scholar]
- 42.Muraki I, Imamura F, Manson JE, Hu FB, Willett WC, van Dam RM, et al. Fruit consumption and risk of type 2 diabetes: results from three prospective longitudinal cohort studies. BMJ. 2013;347:f5001. doi: 10.1136/bmj.f5001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Myasoedova E, Crowson CS, Kremers HM, Therneau TM, Gabriel SE. Is the incidence of rheumatoid arthritis rising? : results from Olmsted County, Minnesota, 1955–2007. Arthritis Rheum. 2010;62:1576–1582. doi: 10.1002/art.27425. [DOI] [PMC free article] [PubMed] [Google Scholar]