Table 3.
Prospective studies of dietary fibre intake and type 2 diabetes included in the meta-analysis
| Author, publication year, country/ region | Study name | Population | Follow-up (years) | Age at baseline (range in years) | No. of cases | Dietary assessment | Exposure | Quantity | RR (95% CI) | Adjustment for confounders |
|---|---|---|---|---|---|---|---|---|---|---|
| Hodge et al, 1993, Australia [39] | NA | 350 M/W | 5 | >20 | 7 | 24 h dietary recall | Total fibre | Per 10 g/day | 0.69 (0.16, 2.96) | Age, sex, BMI, energy |
| Salmeron et al, 1997, USA [8] | Nurses’ Health Study | 85,173 W | 6 | 40–65 | 915 | Validated FFQ, 134 food items | Total fibre | 24.1 vs 11.8 g/day | 0.78 (0.62, 0.98) | Age, BMI, alcohol, smoking, physical activity, family history of diabetes, energy |
| Cereal fibre | 7.5 vs 2.0 g/day | 0.72 (0.58, 0.90) | ||||||||
| Fruit fibre | 7.6 vs 1.4 g/day | 0.87 (0.70, 1.08) | ||||||||
| Vegetable fibre | 9.6 vs 3.4 g/day | 1.17 (0.93, 1.46) | ||||||||
| Salmeron et al, 1997, USA [26] | Health Professionals Follow-up Study | 42,759 M | 6 | 40–75 | 523 | Validated FFQ, 131 food items | Total fibre | 29.7 vs 13.4 g/day | 0.98 (0.73, 1.33) | Age, BMI, alcohol, smoking, physical activity, family history of diabetes, energy |
| Cereal fibre | 10.2 vs 2.5 g/day | 0.70 (0.51, 0.96) | ||||||||
| Fruit fibre | 8.3 vs 1.2 g/day | 1.01 (0.76, 1.36) | ||||||||
| Vegetable fibre | 11.3 vs 3.5 g/day | 1.12 (0.84, 1.49) | ||||||||
| Meyer et al, 2000, USA [10] | Iowa Women’s Health | 35,988 W | 6 | 55–69 | 1,141 | Validated FFQ, 127 food items | Total fibre | 26.5 vs 13.27 g/day | 0.78 (0.64, 0.96) | Age, BMI, WHR, alcohol, smoking, physical activity, education level, energy intake |
| Cereal fibre | 9.43 vs 2.66 g/day | 0.64 (0.53, 0.79) | ||||||||
| Fruit fibre | 8.72 vs 1.71 g/day | 1.17 (0.96, 1.42) | ||||||||
| Vegetable fibre | 11.74 vs 4.71 g/day | 0.97 (0.80, 1.18) | ||||||||
| Soluble fibre | 8.01 vs 4.19 g/day | 0.89 (0.73, 1.08) | ||||||||
| Insoluble fibre | 19.84 vs 9.93 g/day | 0.75 (0.61, 0.91) | ||||||||
| Hu et al, 2001, USA [27] | Nurses’ Health Study | 84,941 W | 16 | 34–59 | 3,300 | Validated FFQ, 61–120 food items | Cereal fibre | Quintile 5 vs 1 | 0.59 (0.52, 0.68) | Age, time, family history of diabetes, menopausal status, HT, smoking status, BMI, moderate-to-vigorous exercise, alcohol, TFA, PUFA/SFA ratio, GL |
| Stevens et al, 2002, USA [28] | Atherosclerosis Risk in Communities Study | 9,529 M/W white participants, 2,722 M/W AA participants | 9 | 45–64 | 971 and 478 | Validated FFQ, 66 food items | Total fibre, whites | Per 1 g/day | 0.999 (0.987, 1.012) | Age, sex, BMI, smoking, physical activity, education level, field centre, energy |
| Cereal fibre, whites | Per 1 g/day | 0.956 (0.925, 0.987) | ||||||||
| Fruit fibre, whites | Per 1 g/day | 1.002 (0.983, 1.021) | ||||||||
| Total fibre, AA | Per 1 g/day | 0.998 (0.980, 1.017) | ||||||||
| Cereal fibre, AA | Per 1 g/day | 0.982 (0.927, 1.039) | ||||||||
| Fruit fibre, AA | Per 1 g/day | 1.009 (0.985, 1.033) | ||||||||
| Cereal fibre, whites | Quintile 5 vs 1 | 0.75 (0.60, 0.92) | ||||||||
| Cereal fibre, AA | Quintile 5 vs 1 | 0.86 (0.65, 1.15) | ||||||||
| Montonen et al, 2003, Finland [6] | Finnish Mobile Clinic Health Examination Survey | 4,318 M/W | 10 | 40–69 | 156 | FFQ, 100 food items | Total fibre | ≥33.2 vs ≤19.2 g/day | 0.51 (0.26, 1.00) | Age, sex, geographical area, smoking BMI, intakes of energy, fruit and berries, vegetables |
| Cereal fibre | ≥24.5 vs ≤12.0 g/day | 0.39 (0.20, 0.77) | ||||||||
| Fruit fibre | ≥3.4 vs ≤0.99 g/day | 0.92 (0.40, 2.13) | ||||||||
| Vegetable fibre | ≥26.5 vs ≤3.7 g/day | 1.19 (0.46, 3.04) | ||||||||
| Soluble fibre | ≥7.4 vs ≤4.5 g/day | 0.57 (0.29, 1.12) | ||||||||
| Insoluble fibre | ≥16.6 vs ≤8.7 g/day | 0.47 (0.25, 0.91) | ||||||||
| Schulze et al, 2004, USA [9] | Nurses’ Health Study II | 91,249 W | 8 | 26–46 | 741 | Validated FFQ, 133 food items | Total fibre | 24.9 vs 12.5 g/day | 1.00 (0.75, 1.34) | Age, BMI, alcohol, smoking, physical activity, family history of diabetes, high blood pressure, high blood cholesterol, HT, OC use, energy intake, Mg, and caffeine, GL, mutual adjustment between fibre types |
| Cereal fibre | 8.8 vs 3.1 g/day | 0.64 (0.48, 0.86) | ||||||||
| Fruit fibre | 6.2 vs 1.1 g/day | 0.79 (0.60, 1.02) | ||||||||
| Vegetable fibre | 10.4 vs 3.4 g/day | 1.12 (0.87, 1.46) | ||||||||
| Hodge et al, 2004, Australia [29] | Melbourne Collaborative Cohort Study | 31,641 M/W | 4 | 40–69 | 365 | FFQ, 121 food items | Fibre | Per 20 g/day | 1.02 (0.81, 1.30) | Age, sex, BMI, WHR, weight change, alcohol, smoking, physical activity, family history of diabetes, education, country of birth, energy intake |
| Cereal fibre | Per 10 g/day | 1.08 (0.88, 1.32) | ||||||||
| Fruit fibre | Per 10 g/day | 0.97 (0.81, 1.16) | ||||||||
| Vegetable fibre | Per 5 g/day | 1.00 (0.86, 1.17) | ||||||||
| Lindström et al, 2006, Finland [7] | The Finnish Diabetes Prevention Study | 172 M, 350 W | 4.1 | 40–64 | 114 | 3 day food record | Fibre | >15.55 vs <10.85 g/4.1868 MJ/day | 0.38 (0.19, 0.77) | Age, sex, baseline weight, baseline 2 h glucose, physical activity, weight change, energy |
| Barclay et al, 2007, Australia [30] | Blue Mountains Eye Study | 1,833 M/W | 10 | ≥49 | 138 | Validated FFQ, 145 food items | Fibre | Per 5 g/day | 0.90 (0.79, 1.02) | Age, sex, smoking, physical activity, family history of diabetes, triacylglycerols, HDL cholesterol |
| Cereal fibre | Per 5 g/day | 0.96 (0.78, 1.20) | ||||||||
| Fruit fibre | Per 5 g/day | 0.94 (0.78, 1.15) | ||||||||
| Vegetable fibre | Per 5 g/day | 0.76 (0.57, 0.99) | ||||||||
| Krishnan et al, 2007, USA [31] | Black Women’s Health Study | 40,078 W | 8 | 21–69 | 1,938 | Validated FFQ, 68 food items | Cereal fibre | 7.6 vs 1.7 g/day | 0.82 (0.70, 0.96) | Age, BMI, smoking, physical activity, family history of diabetes, total fat, protein intake, GI, energy intake |
| Schulze et al, 2007, Germany [5] | EPIC-Potsdam | 9,702 M, 15,365 W | 7 | 35–65 | 844 | Validated FFQ | Soluble fibre | 9.6 vs 5.3 g/day | 0.83 (0.57, 1.22) | Age, sex, BMI, sports activities, education, cycling, occupational activity, smoking, alcohol, total energy intake, waist circumference, PUFA:SFA ratio, MUFA:SFA ratio, carbohydrate, Mg |
| Insoluble fibre | 18.4 vs 10.3 g/day | 0.93 (0.62, 1.40) | ||||||||
| Wannamethee et al, 2009, UK [32] | British Regional Heart Study | 3,428 M | 7 | 60–79 | 162 | Validated 7 day recall FFQ | Total fibre | ≥31.0 vs ≤20 g/day | 0.82 (0.51, 1.32) | Age, waist circumference, alcohol, smoking, physical activity, social class, pre-existing MI or stroke, statin use, energy intake |
| Cereal fibre | Quartile 4 vs 1 | 0.70 (0.44, 1.12) | ||||||||
| Vegetable fibre | Quartile 4 vs 1 | 0.74 (0.46, 1.19) | ||||||||
| Hopping et al, 2010, USA [33] | Multiethnic cohort | 36,256 M, 39,256 W | 14 | 45–75 | 4,555 and 4,032 | Validated FFQ | Total fibre, M | 14.2 vs 7.4 g/4.1868 MJ/day | 0.75 (0.67, 0.84) | Age, BMI, physical activity, education, ethnicity, energy intake |
| Cereal fibre, M | 4.8 vs <1.9 g/4.1868 MJ/day | 0.91 (0.82, 1.00) | ||||||||
| Fruit fibre, M | 3.9 vs <0.8 g/4.1868 MJ/day | 0.93 (0.84, 1.02) | ||||||||
| Vegetable fibre, M | 5.3 vs <2.2 g/4.1868 MJ/day | 0.78 (0.69, 0.88) | ||||||||
| Total fibre, W | 16.2 vs <8.9 g/4.1868 MJ/day | 0.95 (0.85, 1.06) | ||||||||
| Cereal fibre, W | 5.1 vs <2.1 g/4.1868 MJ/day | 0.88 (0.79, 0.97) | ||||||||
| Fruit fibre, W | 5.1 vs <2.1 g/4.1868 MJ/day | 0.95 (0.85, 1.06) | ||||||||
| Vegetable fibre, W | 5.2 vs <1.3 g/4.1868 MJ/day | 0.96 (0.87, 1.08) | ||||||||
| Sakurai et al, 2012, Japan [34] | NA | 1,995 M | 6 | 35–55 | 133 | DHQ, 147 items | Fibre | >6.0 vs <3.7 g/day | 0.99 (0.59, 1.66) | Age, BMI, family history of diabetes, exercise, hypertension, hyperlipidaemia, total energy, GI, GL |
| Wirström et al, 2013, Sweden [35] | NA | 5,477 M/W | 8–10 | 35–56 | 165 | Validated FFQ, NA | Cereal fibre | Per 10 g/day | 0.97 (0.82, 1.14) | Age, family history of diabetes, BMI, leisure-time physical activity, smoking, education, blood pressure |
| >11.6 vs <7.7 g/day | 1.02 (0.68, 1.52) | |||||||||
| Liu et al, 2012, China [37] | NA | 3,461 M/W | >5 | 35–74 | 162 | NA | Total fibre | High vs low | 0.38 (0.17, 0.87) | Age, sex, family history of diabetes, blood pressure, triacylglycerols, fasting glucose |
| Weng et al, 2012, Taiwan [36] | CardioVascular Disease risk FACtor Two-township Study | 1,604 M/W | 4.6 | >30 | 141 | Validated FFQ, 49 food items | Total fibre | 43.3 vs 20.4 g/day | 0.49 (0.28, 0.85)a | Age, sex, age–sex interaction, calories, residential area, family history of diabetes, BMI, central obesity, smoking, drinking, physical activity, hypertension, hypercholesterolaemia, hypertriacylglycerolaemia, low HDL-cholesterol |
| Fruit fibre | 20.6 vs 3.1 g/day | 0.55 (0.32, 0.95)a | ||||||||
| Vegetable fibre | 15.8 vs 3.2 g/day | 0.45 (0.25, 0.82)a | ||||||||
| Qiao et al, 2014, USA [38] | Women’s Health Initiative | 154,493 W | 7.6 | 50–79 | 10,285 | Validated FFQ | Total fibre | ≥13.14 vs <13.14 g/day | 0.98 (0.93, 1.04) | Age, education, cigarette smoking, BMI, WHR, physical activity, family history of diabetes, study arms, HT use |
| Present study, Europe | EPIC-InterAct | 26,088 M/W | 10.8 | 20–79 | 11,559 | Validated FFQs, diet history, 7 day food diaries | Total fibre | >26.4 vs <18.9 g/day | 0.91 (0.81, 1.03) | Age, sex, smoking status, physical activity, education level, alcohol, energy intake, carbohydrates, Mg, vitamin B1, SFA, BMI |
| Cereal fibre | >10.9 vs <5.7 g/day | 0.95 (0.83, 1.08) | ||||||||
| Fruit fibre | >6.3 vs <2.3 g/day | 0.96 (0.83, 1.10) | ||||||||
| Vegetable fibre | >5.3 vs <2.4 g/day | 0.93 (0.84, 1.03) |
4.1868 MJ/day = 1,000 kcal/day
aRisk estimates provided in the article were for a low vs high (reference) comparison. For consistency with the remaining studies, these risk estimates have been converted so the comparison is for the highest vs the lowest intake
AA, African-American, DHQ, diet history questionnaire; FFQ, food frequency questionnaire; GI, glycaemic index; GL, glycaemic load; HT, hormone therapy; M, men; MI, myocardial infarction; MUFA, monounsaturated fatty acids; OC, oral contraceptive; PUFA, polyunsaturated fatty acids; SFA, saturated fatty acids; TFA, trans fatty acids; W; women