- The document summarizes findings from a review of literature on the effects of resistant starch on health outcomes. It found that the majority of studies examined the effect on colonic fermentation and short-chain fatty acid production. Most studies were randomized crossover designs with sample sizes under 50 subjects and conducted in healthy populations in Europe. The review captured data on exposures to different types of resistant starch but noted a lack of research on some types and certain health outcomes.

1. Evidence Map Review: Resistant Starch on Health Outcomes
Background
Conclusions
Dietary fiber research has focused on promising outcomes associated with
exposure to resistant starch. There are four different kinds of resistant starch:
physically inaccessible (RS1), native starch granules (RS2), retrograded starch
(RS3), and chemically modified starch (RS4). Each of these is derived from a
variety of different foods with different chemical properties. RS1 is also known
as indigestible resistant starch and can typically be found in legumes, seeds,
and whole grains. RS2, on the other hand, occurs in its native, granular form
such as unripened bananas, high-amylose starch, and raw potatoes. RS3 is a
type of starch known as retrograded starch, which has been heated and then
cooled leading to a less soluble product, found in certain kinds of bread and
rice. All three of these starches are classified as insoluble fibers, as they do not
dissolve in water and primarily function to provide bulking throughout the
process of digestion. In contrast, RS4, or starch that has been chemically
modified to resist digestion, is considered a soluble fiber due to its ability to
dissolve in water and to be readily fermented in the colon by microbiota. This
starch includes foods that would not be found in nature, such as modified food
starches or processed grains.
• Large number of literature on general RS and RS types 1-3, but
lacking on some fiber types, including corn fiber, high-amylose
starch, and resistant wheat dextrin.
• Lack of research on RS and blood pressure, lipids, GI symptoms,
and transit time. Further examination required.
• Strength of database allows it to capture specific RS exposures
based on how authors reported it. Weakness highlights need for
database users to be familiar with database exposures to properly
identify exposures relevant to fiber of interest.
Vrushabh Shah1,2, Kara A. Livingston1, Caleigh M. Sawicki1, Susan B. Roberts1, Nicola M. McKeown1, Mei Chung3
1Jean Mayer USDA Human Nutrition Research Center on Aging, Tufts University, Boston, MA; 2 Boston University School of Public Health, Boston, MA;
3 Nutrition/Infection Unit, Department of Public Health and Community Medicine, Tufts University School of Medicine, Boston, MA
Figure 3. Database Captured Distribution of Health Outcomes
Methods
Results
Summary of Results
• The Fiber Database had 76 Papers that had at least one exposure
relevant to a resistant starch, and 104 Exposures.
• The largest proportion of outcomes studied was colonic
fermentation/SCFA fermentation, with 17% of papers reporting this
outcome. [Figure 3]
• A majority of the studies were randomized crossover design (74%) and
had a sample size in between 1 and 49 subjects [Table 2]
• 60 Studies were performed on healthy subjects and 51% of studies were
conducted in Europe. [Table 3]
• Multiple descriptions of resistant starch exposure in database, from broad
“resistant starch category,” to specific types 1-4, to food level (i.e. wheat
dextrin) [Figure 4]
• Only 6 studies included adult and adolescent participants and 1 study was
a weight loss intervention. [Table 3, Table 2]
• Studies with exposure to corn fiber had the smallest sample size, while
exposure to wheat dextrin had the largest study sample. [Figure 4]
Development of the Evidence Map:
1) Conducted a systematic literature search in OVID
Medline for fiber interventions with at least one of 9
health outcomes identified at the Ninth Vahouny Fiber
Symposium by ILSI North America & ILSI Europe in
2010 (listed in Table 1)
2) Screened abstracts applying broad inclusion criteria
(see Table 1)
3) Screened full texts of accepted abstracts, applying
additional exclusion criteria (see Table 1)
4) Extracted data from articles into web-based
repository, SRDRTM (http://srdr.ahrq.gov)
5) Pulled all publications in SRDR database with the
terms classifying Resistant Starch as a fiber
exposure.
Inclusion Criteria Exclusion Criteria
-1946 to September 2013
-Published in English
-Abstract contained a fiber term and one of
the following Vahouny heath outcomes:
1. Total & LDL cholesterol
2. Post-prandial glucose & insulin
3. Blood pressure
4. Fecal bulk & laxation
5. Transit time
6. Colonic fermentation & SCFA production
7. Modulation of colonic microflora
8. Weight/Adiposity
9. Increased satiety
-Reviews, bibliographies, case reports
-Observational studies
-Intervention was via tube feeding or enteral
nutrition
-Population was infants (<3 years), pregnant
or breastfeeding women, has any type of
cancer, bowel disease, renal failure, or other
chronic disease condition
-Intervention has no concurrent control
-Fiber dose not clearly reported
-Intervention not sufficiently controlled to
isolate the effect of fiber
-Animal-only studies
-In vitro studies
Data Analysis:
• Descriptive analyses were performed to characterize
study design and outcomes examined
• Fibers and outcomes were reviewed and grouped
into meaningful categories
• Weighted scatter plots were used to visualize
resistant starch exposures by outcome group
and sample size.
11,833 abstracts
identified in OVID
7,257 abstracts
screened
2,047 full text
screened
Fiber Database
N=868
76 Resistant
Starch Papers
Table 1. Inclusion and Exclusion Criteria
Funded by International Life Sciences Institute (ILSI) North America Branch
Contact: vshah923@bu.edu
Figure 2. Database Captured Distribution of Resistant Starch1
Figure 4. Health Outcomes Examined with Resistant Starch Fiber Types2
Figure 1. Classification of Resistant Starch
1Within the Combination/Mixed group, there were 9 combinations that contained resistant starch and 1 combination that did not.
This one combination included barley beta-glucan and oat fiber, and was thus excluded from the figure.
Table 2. Study Design Characteristics (n=76) Table 3. Study Population (n=76) 2Each bubble represents a single publication, and the size corresponds to the study sample size. Publications may
be represented more than once if multiple fiber types or outcome groups were examined but are only represented
once within any given cross-sectional category. Actual bubble placement within each category is random.Characteristic n (%)
Design
Randomized Crossover 56 (74)
Randomized Parallel 15 (20)
Randomized Parallel &
Crossover
1 (1)
Non-Randomized/Unspecified 4 (5)
Sample Size
0 ≤ n ≤ 49 69 (91)
n ≥ 50 7 (9)
Diet Type
Isocaloric/Maintenance 35 (46)
Acute Feeding 27 (36)
Unspecified 12 (16)
Weight Loss 1 (1)
Overfeeding 1 (1)
Characteristic n (%)
Age
Adults (20+ Years) 70 (92)
Adults (20+ Years) + Adolescents (12-
19 years)
6 (8)
Baseline Health
Healthy 60 (79)
Obese/Overweight 9 (12)
Diabetes or At Risk of Diabetes 3 (4)
Metabolic Syndrome 2 (3)
Other 2 (3)
Region
Asia 9 (12)
Australia 9 (12)
Europe 39 (51)
North America 19 (25)
GI Symptoms
Glucose & Insulin
Metabolism
Blood Pressure
Blood
Lipids
Colonic Fermentation/
SCFA Production
Postprandial
Glycemia/Insulinemia
Satiety
Transit Time
Weight/Adiposity
Total & LDL Cholesterol
Modulation Colonic
Microflora
Fecal Bulk/Laxation