Dietary Fiber Intake and Total Mortality
The results of the present meta-analysis indicated that dietary fiber intake is inversely associated with total mortality. Based on the meta-analysis of the highest category of dietary fiber intake (mean ≈ 26.9 g/day) versus the lowest (mean ≈ 15.0 g/day), people with a high fiber intake had a 23% lower risk of total mortality than those who had a relatively low fiber intake. The inverse association did not vary by sex or geographical region. The results of the dose-response meta-analysis suggested that each additional 10 g of fiber intake daily may lower the risk of total mortality by 11%. By source of fiber, cereal and to a lesser extent vegetable fiber were significantly associated with lower total mortality, while fruit fiber showed no association.
Accumulating evidence from observational studies has shown significant inverse associations between dietary fiber intake and risks of several chronic diseases, including stroke, type 2 diabetes, colorectal adenoma, gastric cancer, and breast cancer, all of which could affect total mortality. Several potential mechanisms may explain the beneficial role of dietary fiber intake in the risks of chronic diseases. Dietary fiber intake has been associated with increased satiety resulting from the prolonged time needed for nutrient absorption and delayed gastric emptying, as well as reduced postprandial glucose responses. High dietary fiber intake has also been associated with decreased serum cholesterol levels by reducing the absorption of cholesterol from the small intestine and increasing the excretion of bile acids in feces. In addition, dietary fiber intake is associated with lower levels of inflammatory markers such as C-reactive protein, interleukin-6, and tumor necrosis factor α. Short-chain fatty acids, which are produced from fermentation of dietary fiber by microbiota, function as a key regulator of antiinflammatory actions. These characteristics of dietary fiber intake may decrease the incidence of chronic diseases and thus help reduce the risk of death from these diseases.
Several studies that examined the association between fiber intake from different foods and disease risk reported that cereal fiber was strongly associated with lower risks of colorectal cancer, gastric cancer, type 2 diabetes, and stroke, while vegetable fiber or fruit fiber was weakly associated with disease risk. The results were similar to our findings in that cereal fiber showed the strongest inverse association with total mortality. However, the results from these subgroup analyses should be interpreted with caution, because relatively few studies were included in our analysis of fiber intake from different food sources.
Among the 7 studies included in this meta-analysis, 5 studies also reported relative risks of cardiovascular disease–specific mortality. Three studies showed significant inverse trends for dietary fiber intake, and 2 studies showed a nonsignificant, weak inverse association between dietary fiber intake and death from cardiovascular disease. Three studies provided risk estimates for death from cancer along with total mortality. One study from Europe showed a significant inverse association between dietary fiber intake and death from all cancers in men and women, and another study from the United States found a significant inverse association in men but not in women. In the analyses stratified by cancer type, the inverse association was found only for smoking-related cancers, including cancers of the oral cavity, esophagus, stomach, colorectum, liver, pancreas, lung, and kidney. On the contrary, in the Nurses' Health Study, Baer et al. reported a significant inverse association with non-smoking-related cancers but not with smoking-related cancers.
We found no evidence of heterogeneity among the studies for the analysis of high dietary fiber intake versus low intake. In the dose-response analysis, however, there was some evidence of heterogeneity among the studies. The sensitivity analysis showed that the significant heterogeneity observed in the dose-response analysis disappeared after exclusion of 1 study (I = 1.7%, P = 0.384); this was the largest study and it had a relatively short follow-up period (<10 years), showing modest inverse associations in both men and women.
Our present meta-analysis had some strengths. To the best of our knowledge, this was the first comprehensive meta-analysis to explore the relationship between dietary fiber intake and total mortality. All of the studies included in this meta-analysis were prospective cohort studies. A prospective study design can minimize the possibility that the results were affected by recall or selection bias, which could be of concern in case-control studies. In addition, we included a large number of cases (n = 62,314) and subjects (n = 908,135) in the meta-analysis, which provided good statistical power for evaluating the association between dietary fiber intake and total mortality risk. Although a limited number of studies examined the association between dietary fiber from different food sources and total mortality, our findings may be useful for generating a hypothesis for future research. The quality assessment indicated that all of the studies included in the meta-analysis were of either high quality or relatively good quality, and the majority of studies adjusted for important confounders, including BMI, smoking, alcohol drinking, and physical activity.
Despite these strengths, several limitations also need to be acknowledged. First, some nondifferential misclassification of dietary fiber intake may have occurred in each study and thus in the meta-analysis, which may have attenuated any true association between dietary fiber intake and total mortality to some extent. All of the studies included in the meta-analysis assessed dietary fiber intake at a single time, except for 1 study that measured dietary fiber intake repeatedly. Some possible exposure misclassification may be especially high for studies with long-follow-up periods that assessed dietary fiber intake at baseline only. However, the subgroup analyses by follow-up period in our meta-analysis showed no significant difference in relative risks between the studies with short follow-up periods and the studies with long follow-up periods. Second, because our quantitative assessment was based on observational studies, we cannot rule out the possibility that unknown and/or residual confounding may still have affected the results in each study and thus the pooled estimates in the meta-analyses. Finally, in a meta-analysis of published studies, publication bias could be of concern. Although we found no significant evidence for publication bias, small studies that find no association are less likely to be published.
In conclusion, results from this meta-analysis of prospective cohort studies provide quantitative evidence regarding the inverse association between dietary fiber intake and total mortality. For practical reasons (including time and expense), it is hard to conduct experimental studies to investigate the association between dietary fiber intake and total mortality. Although the results of meta-analyses were based on the observational studies only, our findings may have public health implications, given that the mean dietary fiber intake in the United States and in many European countries (15 g/day) is less than half of recommended levels (28–36 g/day). Further well-designed large prospective cohort studies with repeated measurements of dietary fiber intake from different food sources, long-follow-up periods, and adjustment for all potential confounders are needed to verify the association of dietary fiber intake with total mortality.
Discussion
The results of the present meta-analysis indicated that dietary fiber intake is inversely associated with total mortality. Based on the meta-analysis of the highest category of dietary fiber intake (mean ≈ 26.9 g/day) versus the lowest (mean ≈ 15.0 g/day), people with a high fiber intake had a 23% lower risk of total mortality than those who had a relatively low fiber intake. The inverse association did not vary by sex or geographical region. The results of the dose-response meta-analysis suggested that each additional 10 g of fiber intake daily may lower the risk of total mortality by 11%. By source of fiber, cereal and to a lesser extent vegetable fiber were significantly associated with lower total mortality, while fruit fiber showed no association.
Accumulating evidence from observational studies has shown significant inverse associations between dietary fiber intake and risks of several chronic diseases, including stroke, type 2 diabetes, colorectal adenoma, gastric cancer, and breast cancer, all of which could affect total mortality. Several potential mechanisms may explain the beneficial role of dietary fiber intake in the risks of chronic diseases. Dietary fiber intake has been associated with increased satiety resulting from the prolonged time needed for nutrient absorption and delayed gastric emptying, as well as reduced postprandial glucose responses. High dietary fiber intake has also been associated with decreased serum cholesterol levels by reducing the absorption of cholesterol from the small intestine and increasing the excretion of bile acids in feces. In addition, dietary fiber intake is associated with lower levels of inflammatory markers such as C-reactive protein, interleukin-6, and tumor necrosis factor α. Short-chain fatty acids, which are produced from fermentation of dietary fiber by microbiota, function as a key regulator of antiinflammatory actions. These characteristics of dietary fiber intake may decrease the incidence of chronic diseases and thus help reduce the risk of death from these diseases.
Several studies that examined the association between fiber intake from different foods and disease risk reported that cereal fiber was strongly associated with lower risks of colorectal cancer, gastric cancer, type 2 diabetes, and stroke, while vegetable fiber or fruit fiber was weakly associated with disease risk. The results were similar to our findings in that cereal fiber showed the strongest inverse association with total mortality. However, the results from these subgroup analyses should be interpreted with caution, because relatively few studies were included in our analysis of fiber intake from different food sources.
Among the 7 studies included in this meta-analysis, 5 studies also reported relative risks of cardiovascular disease–specific mortality. Three studies showed significant inverse trends for dietary fiber intake, and 2 studies showed a nonsignificant, weak inverse association between dietary fiber intake and death from cardiovascular disease. Three studies provided risk estimates for death from cancer along with total mortality. One study from Europe showed a significant inverse association between dietary fiber intake and death from all cancers in men and women, and another study from the United States found a significant inverse association in men but not in women. In the analyses stratified by cancer type, the inverse association was found only for smoking-related cancers, including cancers of the oral cavity, esophagus, stomach, colorectum, liver, pancreas, lung, and kidney. On the contrary, in the Nurses' Health Study, Baer et al. reported a significant inverse association with non-smoking-related cancers but not with smoking-related cancers.
We found no evidence of heterogeneity among the studies for the analysis of high dietary fiber intake versus low intake. In the dose-response analysis, however, there was some evidence of heterogeneity among the studies. The sensitivity analysis showed that the significant heterogeneity observed in the dose-response analysis disappeared after exclusion of 1 study (I = 1.7%, P = 0.384); this was the largest study and it had a relatively short follow-up period (<10 years), showing modest inverse associations in both men and women.
Our present meta-analysis had some strengths. To the best of our knowledge, this was the first comprehensive meta-analysis to explore the relationship between dietary fiber intake and total mortality. All of the studies included in this meta-analysis were prospective cohort studies. A prospective study design can minimize the possibility that the results were affected by recall or selection bias, which could be of concern in case-control studies. In addition, we included a large number of cases (n = 62,314) and subjects (n = 908,135) in the meta-analysis, which provided good statistical power for evaluating the association between dietary fiber intake and total mortality risk. Although a limited number of studies examined the association between dietary fiber from different food sources and total mortality, our findings may be useful for generating a hypothesis for future research. The quality assessment indicated that all of the studies included in the meta-analysis were of either high quality or relatively good quality, and the majority of studies adjusted for important confounders, including BMI, smoking, alcohol drinking, and physical activity.
Despite these strengths, several limitations also need to be acknowledged. First, some nondifferential misclassification of dietary fiber intake may have occurred in each study and thus in the meta-analysis, which may have attenuated any true association between dietary fiber intake and total mortality to some extent. All of the studies included in the meta-analysis assessed dietary fiber intake at a single time, except for 1 study that measured dietary fiber intake repeatedly. Some possible exposure misclassification may be especially high for studies with long-follow-up periods that assessed dietary fiber intake at baseline only. However, the subgroup analyses by follow-up period in our meta-analysis showed no significant difference in relative risks between the studies with short follow-up periods and the studies with long follow-up periods. Second, because our quantitative assessment was based on observational studies, we cannot rule out the possibility that unknown and/or residual confounding may still have affected the results in each study and thus the pooled estimates in the meta-analyses. Finally, in a meta-analysis of published studies, publication bias could be of concern. Although we found no significant evidence for publication bias, small studies that find no association are less likely to be published.
In conclusion, results from this meta-analysis of prospective cohort studies provide quantitative evidence regarding the inverse association between dietary fiber intake and total mortality. For practical reasons (including time and expense), it is hard to conduct experimental studies to investigate the association between dietary fiber intake and total mortality. Although the results of meta-analyses were based on the observational studies only, our findings may have public health implications, given that the mean dietary fiber intake in the United States and in many European countries (15 g/day) is less than half of recommended levels (28–36 g/day). Further well-designed large prospective cohort studies with repeated measurements of dietary fiber intake from different food sources, long-follow-up periods, and adjustment for all potential confounders are needed to verify the association of dietary fiber intake with total mortality.