Diet Quality and Cancer Outcomes in Adults: A Systematic Review of Epidemiological Studies

Dietary patterns influence cancer risk. However, systematic reviews have not evaluated relationships between a priori defined diet quality scores and adult cancer risk and mortality. The aims of this systematic review are to (1) describe diet quality scores used in cohort or cross-sectional research examining cancer outcomes; and (2) describe associations between diet quality scores and cancer risk and mortality. The protocol was registered in Prospero, and a systematic search using six electronic databases was conducted through to December 2014. Records were assessed for inclusion by two independent reviewers, and quality was evaluated using a validated tool. Sixty-four studies met inclusion criteria from which 55 different diet quality scores were identified. Of the 35 studies investigating diet quality and cancer risk, 60% (n = 21) found a positive relationship. Results suggest no relationship between diet quality scores and overall cancer risk. Inverse associations were found for diet quality scores and risk of postmenopausal breast, colorectal, head, and neck cancer. No consistent relationships between diet quality scores and cancer mortality were found. Diet quality appears to be related to site-specific adult cancer risk. The relationship with cancer mortality is less conclusive, suggesting additional factors impact overall cancer survival. Development of a cancer-specific diet quality score for application in prospective epidemiology and in public health is warranted.

Eight component score based on eight recommendations from the US National Research Council Diet and Health. Healthiest intakes scored as 0, intermediate scored as 1 and unhealthiest intakes scored as 2. Scores include % energy from total fat; % energy from saturated fat; cholesterol intake; fruit and vegetable intake; breads, cereals and legume intake; protein intake; sodium intake and calcium intake. Comprises 10 components (grains; vegetables; fruit; total fat; saturated fat; cholesterol; iron; calcium; diet diversity; and moderation in added fat and sugar) The total fat, saturated fat, and cholesterol components were calculated as a % total energy and scored categorically as 0, 5, or 10, and the remaining components were scored as continuous variables from 0 to 10, proportional to recommended range of intake. Scores summed across the 10 components for a highest possible score of 100 points.  Similar to the HEI. Comprises 10 equally weighted components with each contributing 10 points toward total score. Components 1-5 based on serve recommendations from the USDA Food Guide Pyramid (vegetables, fruit, meat, milk, grains). Other components based on % energy from fat and sat fat, salt intake and cholesterol intake. Single score developed to assess whether the AHEI predicts chronic disease better than the original HEI.
Differed from the HEI scoring criteria by addressing quality within food groups by removing potatoes (including French fries), including nuts and soy, a ratio of white/red meat, trans fat, PUFA:SAT fat ratio, and adding sub-scores for multivitamin use and alcohol intake. Eight components (vegetables; fruit; nuts and soy; white:red meat ratio; cereal fibre; trans fat; alcohol and PUFA:SAT fat ratio) scored from 0 (worst) to 10 (best) with intermediate intakes scored proportionately between 0 and 10. Multivitamin use scored dichotomously with 2.5 points awarded for non-use and 7.5 points for use. Based on the 12 components of the MyPyramid food patterns. Nine components scored on adequacy, where highest score was assigned for meeting the guidelines: total fruit, including juice (0-5); whole fruit, excluding juice (0-5); total vegetables (0-5); dark green and orange vegetables and legumes (0-5); total grains (0-5); whole grains (0-5); milk (0-10); meat and beans (0-10); and oils (0-10). Three components based on moderation, where lower intakes scored higher: saturated fat (0-10); sodium (0-10); and % energy from fat, alcohol and added sugar (0-20). Total and sub-scores expressed per 1000 calories to account for differences in energy intake.  USA Food and nutrient Similar to above except scoring for sodium changed.
As for the 2005-HEI above except different scoring for sodium. Participants divided into 11 equal groups based on the distribution of reported sodium intakes and assigned corresponding scores of 0-10 (higher score for less sodium intake). 2013 [97].

Food and nutrient
Update the HEI to reflect the changes made in the 2010 Dietary Guidelines for Americans.

USA Food
Single summary score based on adherence to the 1995 US Dietary Guidelines.
Dichotomous score based on intake (Y = 1 N = 0) of 23 food items (apples or pears; oranges; cantaloupe; orange or grapefruit juice; grapefruit; other fruit juices; dried beans; tomatoes; broccoli; spinach, mustard; turnip or collard greens; carrots or mixed vegetables with carrots; green salad; sweet potatoes, yams, other potatoes; baked or stewed chicken or turkey; baked or broiled fish; dark breads; cornbread, tortillas, grits; high fibre cereals; cooked cereals; 2% milk and 2% milk products; 1% skim milk). 1 point given for each item if the food is consumed at least once per week.

Sweden Food
Provide a measure to examine mortality outcomes by poor diet quality.
Scoring based on 16 food items including red meat products (3 items); processed meat products (5 items); high fat dairy (3 items); white bread; sweets; potato chips/popcorn; mayonnaise; ice cream. Consumption of any of these non-recommended products 3 or more times per week assigned a score of 1. Score of 0 assigned for consumption below 3 times per week.

Sweden Food
Provide a measure to examine mortality outcomes by poor diet quality.
Scoring based on 21 items. Consumption of any of these non-recommended products 3 or more times per month assigned a score of 1. Score of 0 assigned for consumption below 3 times per month. nuts; cereals; fish; high monounsaturated:saturated fatty acid ratio) intake above the sex-specific median cut-off assigned score = 1 and below = 0. For components presumed to be detrimental (meat and poultry; and dairy) persons whose intake was above the sex-specific median assigned score = 0 and below = 1. For alcohol, intake for men scored 1 for ethanol intake between 10 and 50 g/d and women scored 1 for intake between 5 and 25 g/d.

USA Food
Single score to asses conformity to the traditional Mediterranean diet.
Nine dietary components used to capture the traditional Mediterranean diet. One point scored for intake above the sex-specific cohort median for the following components considered beneficial to health; whole grains; total vegetables (excluding potatoes); total fruit; fish; legumes; and nuts. One point given for intake below the sex-specific Mediterranean median for red and processed meat and monounsaturated fatty acid to saturated fatty acid ratio. A moderate intake of and 5-25 g/d awarded one point. fruits and nuts; cereals; fish; high monounsaturated and polyunsaturated:saturated fatty acid ratio) intake above the sex-specific median cut-off assigned score = 1 and below = 0. For components presumed to be detrimental (meat and poultry; and dairy) persons whose intake was above the sex-specific median assigned score = 0 and below = 1. For alcohol, intake for men scored 1 for ethanol intake between 10 and 50 g/d and women scored 1 for intake between 5 and 25 g/d.  and nuts; cereals; fish; high monounsaturated:saturated fatty acid ratio) intake above the sex-specific median cut-off assigned score = 1 and below = 0. For components presumed to be detrimental (meat and poultry; and dairy) persons whose intake was above the sex-specific median assigned score = 0 and below = 1. For alcohol, intake for men of ≤2 drinks/d and women intake ≤1 drink/d scored 1. Total of nine component scores. For beneficial components (vegetables; legumes; fruits and nuts; cereals; fish; high monounsaturated and polyunsaturated:saturated fatty acid ratio) intake above the centre and sex-specific median cut-off assigned score = 1 and below = 0. For components presumed to be detrimental (meat and poultry; and dairy) persons whose intake was above the centre and sexspecific median assigned score = 0 and below = 1. For alcohol, intake for men scored 1 for ethanol intake between 10 and 50 g/d and women scored 1 for intake between 5 and 25 g/d. Modification of MMDS to exclude potato products from the vegetable group; separate fruits and nuts into two groups; eliminate the dairy group, include whole-grain products only; and include only red and processed meats in the meat group. For beneficial components (fruits; nuts; vegetables; legumes; whole grains; fish; and MUFA:SAT fat ratio) one point is given for intake above the population median. One point given for intake below the population median for unhealthy components (red and processed meats). For alcohol, intake between 10 and 25 g/d assigned 1 point for both men and women. Modification of MMDS (1) to exclude potato products from the vegetable group; separate fruits and nuts into two groups; eliminate the dairy group, include whole-grain products only; and include only red and processed meats in the meat group. For beneficial components (fruits; nuts; vegetables [excluding potatoes]; legumes; whole grains; fish; and MUFA:SAT fat ratio) one point is given for intake above the population median. One point given for intake below the population median for unhealthy components (red and processed meats). For alcohol, intake between 5 and 15 g/d. USA Food Indicate the degree of adherence to the Mediterranean diet.
As above with the addition of sex-specific cut-offs for alcohol intake. Modification of MMDS to exclude potato products from the vegetable group; separate fruits and nuts into two groups; eliminate the dairy group, include whole-grain products only; and include only red and processed meats in the meat group. For beneficial components (fruits; nuts; vegetables [excluding potatoes]; legumes; whole grains; fish; and MUFA:SAT fat ratio) one point is given for intake above the population median. One point given for intake below the population median for unhealthy components (red and processed meats). For alcohol, intake between 5 and 15 g/d for women and 10-25 g for men assigned 1 point. A variation of the MDS and MMDS. Based on intake of 9 key components. Each component (excluding alcohol) measured as grams/1000 kcal/day and divided into tertile of intake. Participants in tertiles 1, 2 and 3 received a score of 0, 1 and 2 respectively based on their level of intake within the component category. Positive scoring for components presumed to be beneficial (fruit including nuts and seeds, excluding juice; vegetables, excluding potato; legumes; cereals; fish; and olive oil). Scoring was inverted for those components presumed to be non-beneficial to health (total meat and dairy). Alcohol scored as a dichotomous variable. Two points were assigned for moderate intake (5-25 g ethanol/d for women and 10-50 g/day ethanol for men) and 0 points for above and below the sex-specific ranges.  Provide a summary estimate of adherence to the Mediterranean diet with the exclusion of alcohol as alcohol is a known risk factor for BC incidence.
Adapted version of rMED. arMED scoring based on the MDS and MMDS. For the six components presumed to be beneficial to health (fruits, vegetables, legumes, fish, olive oil, and cereals) participants were assigned a score of 0-2 based on country-specific tertiles of intake. This scoring was inverted for components presumed to be detrimental to health (meat and dairy). grains, fish, high monounsaturated:saturated fatty acid ratio) intake above the sex-specific median cut-off assigned score = 1 and below = 0. For components presumed to be detrimental (meat and dairy) persons whose intake was above the sex-specific median assigned score = 0 and below = 1. For alcohol, intake for men scored 1 for ethanol intake between 10 and 50 g/d and women scored 1 for intake between 5 and 25 g/d.

USA Nutrient
Provide a summary score of adherence to a low carbohydrate, low protein eating pattern.
Percent energy from fat, protein, and carbohydrate were divided equally into 11 categories according to percentiles. For fat and protein, those in the highest category received 10 points and those in the next received 9 points and so forth. A reciprocal scoring system was applied for carbohydrate intake.  [98].

Greece Nutrient
Provide a summary score of adherence to a low carbohydrate, low protein eating pattern.
All participants sorted based on deciles of carbohydrate and protein intake. For each participant, ascending decile of protein intake and descending decile of carbohydrate intake were added to create the total LCHP score using, alternatively, absolute and energy-adjusted carbohydrate and protein values. Comprises eight components. For those components considered beneficial to health (fruits; vegetables; nuts and legumes; low-fat dairy products; and wholegrains) participants are categorised into quintiles. Those in quintile 1 receive 1 point and those in quintile 5 receive 5 points. For those components considered detrimental to health (sodium; processed meats; and sweetened beverages), those in quintile 1 received 5 points and those in quintile 5 1 point. Dichotomous score with 1 point awarded for meeting the guideline and 0 awarded otherwise. Total of nine components: total fruit; total vegetables; total whole grains; nuts, seeds, and legumes; meat intake, % energy from sugar; alcohol intake; and % energy from saturated fat.   [101].

USA Nutrient
Totally nutrient-based to assess the target nutrient values from the DASH eating pattern.
Nine components of which are expected to be higher (protein, fibre, magnesium, calcium, and potassium) and lower (total fat, saturated fat, sodium, and cholesterol) on the DASH eating pattern. The method uses nutrient targets based on a 2100-cal diet for both men and women.

EPIC (10 European countries) Foods
Provide a means to improve between-country comparability of fruit and vegetable diversity amongst participants in EPIC the DDS was created and applied.
Based on the baseline questionnaires four different DDS components derived: DDSvegfr counts the total number of different fruits and vegteables eaten over a two week period; DDSveggr one point given for eating each of the following at least once in a two week period (leafy vegetables; fruiting vegetables; root vegetables; cabbage; mushrooms; grain and pod vegetables; onion and garlic; and stalk vegetables); DDSvegpr counts the total number of different vegetable products eaten at least once in two weeks; DDSfr counts the total number of different fruit products eaten at least once in two weeks.

EPIC (10 European countries) Food
Provide a means to improve between-country comparability of fruit and vegetable diversity amongst participants in EPIC the DDS was created and applied. Taiwan Food Simple summary estimate of diet diversity based on 6 core foods.
One point given for ≥0.5 serves per day for each of the following six food groups: dairy; eggs/legumes/meat/fish; grains; fruit; vegetables; oils/fats.  Dash eating pattern cut offs used to assign scores. One point given where intake met the guidelines for 8 components; fruit and vegetables; fish; whole grains; sodium; nuts, seeds, and legumes; added sugar; processed meat; and % energy from saturated fat. Score of 0 given if guideline not met.  The index comprises six components as outlined in the 2005 Swedish Nutrition Recommendations. Components comprise: % energy from saturated fatty acids and polyunsaturated fatty acids; fish and shellfish (g/wk); dietary fibre intake (g/MJ); fruit and vegetable intake (g/d); and % energy from sucrose. Model 1: uses pre-defined cut-offs based on the 2005 SNR and SDG.Cut-offs create dichotomous variables where adherence receives 1 point and non-adherence 0 points. Model 2: uses the study population's median energy-adjusted intakes as cut offs. One point assigned for intake above the population median for PUFA; fish and shellfish; dietary fibre; fruit and vegetables. One point assigned for intake below the median for SFA and sucrose.
Model 3: ranked individuals into quintiles of energyadjusted intake of the components. A proportional scale ranging from 1-5 was assigned to the different quintile groups, with high scores to individuals with higher intakes of PUFA, fish and shellfish; dietary fibre; and fruit and vegetables and lower scores for higher intakes of SFA and sucrose. ODI-R comprises 9 components: amount (10 points) and quality (5 points) of grains and starchy tubers; amount of vegetable (10 points) and fruit (10 points) intake; amount (10 points) and quality (5 points) of eggs/soy/fish/meats; amount of dairy intake (10 points); polyunsaturated:saturated fatty acid ratio (10 points); dietary cholesterol intake (10 points); refined sugar intake (3 points), salt intake (4 points); alcohol intake (3 points); and dietary diversity (at least 0.5 serving from 10 food groups, 0-10 points).

USA food
Develop a food-based diet quality score to investigate relationships to overall and cause-specific mortality.
Based on dichotomous scoring for 34 food groups in a positively rated category, neutral category and a negative rated category. Scores for food groups either positively or negatively rated are categorised into quartiles. The a priori diet quality score was calculated by summing category scores for foods in the positively rated group (0-3) and the negatively rated group (3-0).                Adherence to the GFPI does not have much of an impact on overall or cancer mortality risk * Where a study provided several risk estimates with different degrees of adjustment for confounding the estimate presented is the one adjusting for the largest number of factors.           Up to F/U 16 y Cancer mortality Men HR(tertile 3 V tertile 1) = 0.30 (95% CI: 0.11, 0.83; P trend 0.06) Women HR(tertile 3 V tertile 1) = 0.64 (95% CI: 0.24, 1.68; P trend 0.65) Adherence to the Aussie-DQI in men is associated with reduced cancer mortality * Fully adjusted hazard ratio (HR) or relative risk (RR) reported where given.