International Journal of Sport Nutrition and Exercise Metabolism, 2015, 25, 243 -251 http://dx.doi.org/10.1123/ijsnem.2014-0034 © 2015 Human Kinetics, Inc.
Original Research
Relationship Between General Nutrition Knowledge and Dietary Quality in Elite Athletes Inge Spronk, Susan E. Heaney, Tania Prvan, and Helen T. O’Connor This study investigated the association between general nutrition knowledge and dietary quality in a convenience sample of athletes (≥ state level) recruited from four Australian State Sport Institutes. General nutrition knowledge was measured by the validated General Nutrition Knowledge Questionnaire and diet quality by an adapted version of the Australian Recommended Food Score (A-ARFS) calculated from food frequency questionnaire data. Analysis of variance and linear modeling were used to assess relationships between variables. Data: mean (Standard Deviation). A total of 101 athletes (Males: 37; Females: 64), 18.6 (4.6) years were recruited mainly from team sports (72.0%). Females scored higher than males for both nutrition knowledge (Females: 59.9%; Males: 55.6%; p = .017) and total A-ARFS (Females: 54.2% Males: 49.4%; p = .016). There was no significant influence of age, level of education, athletic caliber or team/individual sport participation on nutrition knowledge or total A-ARFS. However, athletes engaged in previous dietetic consultation had significantly higher nutrition knowledge (61.6% vs. 56.6%; p = .034) but not total A-ARFS (53.6% vs. 52.0%; p = .466). Nutrition knowledge was weakly but positively associated with total A-ARFS (r = .261, p= .008) and A-ARFS vegetable subgroup (r = .252, p = .024) independently explaining 6.8% and 5.1% of the variance respectively. Gender independently explained 5.6% of the variance in nutrition knowledge (p= .017) and 6.7% in total A-ARFS (p = .016). Higher nutrition knowledge and female gender were weakly but positively associated with better diet quality. Given the importance of nutrition to health and optimal sports performance, intervention to improve nutrition knowledge and healthy eating is recommended, especially for young male athletes. Keywords: nutrition assessment, nutritional status, sport nutrition, nutrition knowledge Numerous factors including taste, convenience, cultural and religious beliefs, food security and availability (Parmenter & Wardle, 1999, Wardle et al., 2000, Hendrie et al., 2008a, Heaney et al., 2011) are reported to impact dietary intake. In athletes, a desire to increase lean and reduce fat mass or periodise fuels for training or competition are also known to be influential (American Dietetic Association et al., 2009). Although nutrition knowledge is considered to be a pivotal factor (Spendlove et al., 2012), its influence on dietary intake surprisingly remains relatively unexplored in athletes (Heaney et al., 2011). Nutrition education programs are designed to increase nutrition knowledge with the expectation that this translates into improved dietary intake, better health and for athletes, enhanced sports performance. Spronk is with the Division of Human Nutrition, Wageningen University, Wageningen, Netherlands. Heaney and O’Connor are with the Discipline of Exercise and Sport Science, University of Sydney, Sydney, Australia. Prvan is with the Dept. of Statistics, Faculty of Science, Macquarie University, Sydney, Australia. Address author correspondence to Helen T. O’Connor at [email protected].
Nutrition knowledge in athletes is generally reported to be similar or better than the general population but study quality is poor and use of validated instruments rare (Heaney et al., 2011). Demographic factors such as female gender, higher athletic caliber, and engagement in physique-oriented sports have been associated with higher nutrition knowledge (Raymond-Barker et al., 2007). In nonathletic groups, better-educated, middle aged (compared with younger or older age) participants tend have higher nutrition knowledge (Parmenter et al., 2000, Wardle et al., 2000, Hendrie et al., 2008a). In studies examining the association between nutrition knowledge and dietary intake in either athletes or the general population, the relationship is reported to be either null or weak (r < .5) (Spronk et al., 2014). Nutrition knowledge is regarded as a difficult construct to measure (Worsley, 2002). In athlete studies, assessment has included general concepts relevant to maintaining health (e.g., eating more dietary fiber and less saturated fat) (Abood et al., 2004, Spendlove et al., 2012) through to sports specific nutrition knowledge (e.g., an understanding of higher energy, protein, carbohydrate, and fluid needs for sports performance) (Rash et al., 2008). Assessment of sports specific nutrition
243
Downloaded by Purdue Univ on 09/17/16, Volume 25, Article Number 3
244 Spronk et al.
knowledge can be difficult and although the general principles are similar, there are differences in dietary requirements between sports (Burke & Deakin, 2006). However, general nutrition knowledge principles which encompass healthy eating are relevant to all athletes and these are also consistent with public health guidelines (National Health & Medical Research Council, 2003). General nutrition knowledge which examines understanding of healthy eating guidelines would also be expected to be more strongly related to overall dietary intake or quality than specific sports nutrition knowledge such as hydration requirements, preevent or recovery eating. To optimize performance, athletes must know how to consume a healthy, balanced diet and combine this with specific sports nutrition strategies relevant to their training and competition situation (American Dietetic Association et al., 2009). Examining associations between nutrition knowledge and diet is challenging as dietary intake is also difficult to measure accurately. Dietary assessment can be more problematic in athletes due to larger energy intakes and portion sizes, wide daily variation due to periodised training programs and consumption of food and beverages (often specific sports foods) during training sessions or competition where precise recording and then nutrient analysis is more difficult (Heaney et al., 2010). Although numerous studies have reported on inadequate nutrient intake in athletes, those with higher energy intake often meet micronutrient requirements due to the sheer volume of food consumed (Heaney et al., 2010). Recently, examination of dietary intake via a diet quality score has been employed to assess positive dietary attributes (e.g., intake of a greater variety of vegetables, fruits, wholegrains and reduced fat dairy foods) (Collins et al., 2008, Guenther et al., 2008). An Australian Recommended Food Score (ARFS), derived primarily from dietary intakes measured by a Food Frequency Questionnaire (FFQ; Cancer Council Victoria, 2011) has been used to assess diet quality in the community (Collins et al., 2008) but has not been applied to assess the diet of athletes. Therefore, the aim of this study was to evaluate the relationship between nutrition knowledge and the quality of dietary intake in a sample of elite Australian athletes. The study also aimed to examine the association and underpinning demographic factors linking nutrition knowledge and quality of dietary intake in this population.
Methods Participants Participants were a convenience sample of elite athletes (≥ state level; ≥ 16 years) who had a monetary scholarship from the Australian government to support development of their athletic careers at one of four nonresidential state institutes/academies of sport in Australia. To increase participant numbers, elite rugby league players (18–20 years) competing in the Australian National Youth Competition
were also recruited. Participants were a subset of a larger study reporting on the nutrition knowledge of athletes (Spendlove et al., 2012). The study was approved by the Human Ethics Committee of the University of Sydney. Participation was anonymous and informed written consent before study commencement was obtained (for participants 1/week for each of the following; potato (boiled, mashed, baked); pumpkin; cauliflower; green beans; spinach; cabbage or brussel sprouts; peas; broccoli; carrots; zucchini, eggplant, squash; capsicum; corn, sweet corn, corn on the cob; mushrooms; tomatoes; lettuce; celery, cucumber; avocado; onion, spring onion, leek
Fruits
≥ 2 pieces of fruit/day; >1/week for each of the following; canned fruit, e.g., peaches, two fruits; fresh fruit salad; dried fruit, e.g., sultanas, dried apricots; apple or pear; orange, mandarin, grapefruit; banana; peach, nectarine, plum or apricot; pineapple; grapes, strawberries, blueberries; melon, e.g., watermelon, rockmelon, honeydew melon
11
Protein foods Meat/ fish/chicken
1–4/week of each of the following; mince dish, e.g., spaghetti bolognese, rissoles, shepherd’s pie, lasagne; meat (beef or lamb), e.g., roast, chops, steak, beef stroganoff, stir fry (with or without sauce); chicken, e.g., roast, BBQ, satay, stir fry (with or without sauce); pork, e.g., roast, chops, sweet & sour, stir fry (with or without sauce); >1/week of each of fresh fish not crumbed or battered; canned tuna, salmon, sardines including patties; other seafood, e.g., prawns, lobster
7
Protein foods Vegetarian alternatives
>1/week of each of the following; nuts, e.g., peanuts, almonds; peanut/nut butter, ; eggs, e.g., boiled, scrambled; baked beans; other beans, lentils, e.g., chickpeas, split peas
5
Grains
Usually eating ‘other’ bread; >1/week of each of muesli; porridge; breakfast cereal; bread, pita bread, roll or toast all types; English muffin, bagel or crumpet; rice; pasta, e.g., spaghetti, lasagna, pasta bake
Usually eating brown (multigrain, wholemeal) bread (2 points)
9
Dairy
≥ 2 serves of a glass of milk, a tub of yoghurt or a slice of cheese per day; usually drinking reduced fat milk, skim milk or soy milk; 1/week—1/day of each of flavored the following; milk, hot chocolate, milkshake, thickshake, smoothie; ice cream—vanilla, chocolate, strawberry, sundaes, cones; >1/day of each of plain milk—glass or with cereal; 1/day—3/day of each of cheese, including cheese on sandwiches, biscuits or on toast
Italics: combination of these answers (additional 1 point)
9
Water
≥ 4 glasses of water/day
1
Other
>1/week of each of Vegemite, Mighty Mite, Promite, Marmite; tomato sauce, barbecue sauce
2
Total
64
†Adapted from Australian Recommended Food Score (Collins et al., 2008)
Linear Model to the data, assumptions were checked by examining the residuals versus fitted values plot and the normal probability plot of the standardized residuals. When variables were not normally distributed, Spearman’s Rank Correlation coefficient (rs) was used instead of Pearson’s. Results were reported as mean (SD) unless otherwise indicated. If the normal distribution assumption was violated the Mann-Whitney test was performed instead of the independent t test when comparing two groups. Significance was accepted at p < .05 level. Analysis of dietary intake data consisted primarily of one-way analysis of variance (ANOVA) or Kruskall-Wallis if the normality assumption did not hold. Independent t tests were conducted to identify any gender-related mean differences when data were normally distributed and Mann Whitney test if data were not. Chi square tests for asso-
ciations between variables and groups were conducted when data were categorical.
Results Participant Characteristics Demographic characteristics of participants (n = 101) are summarized in Table 2. Females comprised 63.4% of the sample with a high representation of younger (16–18 years; 68.3%) and team sport athletes (72.3%) and low representation from physique oriented sports (diving; n = 1). Participants mean age was 18.6 (4.6) years with no difference between genders (p = .137). Most participants were competing at national or international level (79.2%).
IJSNEM Vol. 25, No. 3, 2015
246 Spronk et al.
Table 2 Participant Characteristics Male (n = 37) (36.6%) Variable age (y)
Mean (%)
SD
17.6
1.2
Female (n = 64) (63.4%) Mean (%)
SD
19.1
5.6
Total (N = 101) (100%) Mean (%)
SD
18.6
4.6
athletes >18
10 (9.9%)
22 (21.8%)
32 (31.7%)
athletes 16–18
27 (26.7%)
42 (41.6%)
69 (68.3%)
30 (29.7%)
43 (42.6%)
73 (72.3%)
7 (6.9%)
21(20.8%)
28 (27.7%)
first year
21 (20.8%)
33 (32.7%)
54 (53.5%)
1–3 years
11 (10.9%)
20 (19.8%)
31 (30.7%)
4+ years
5 (5.0%)
11 (10.9%)
16 (15.8%)
Sport team† individual‡ Years on scholarship
Downloaded by Purdue Univ on 09/17/16, Volume 25, Article Number 3
Highest level of competition state
9 (8.9%)
12 (11.9%)
21 (20.8%)
national
15 (14.9%)
26 (25.7%)
41 (40.6%)
international (junior or open)
13 (12.9%)
26 (25.7%)
39 (38.6%)
high school
26 (25.7%)
43 (42.6%)
69 (68.3%)
tertiary
11 (10.9%)
21 (20.8%)
32 (31.7%)
Highest level of education
Living situation parents (including other family situation)
30 (29.7%)
51 (50.5%)
81 (79.2%)
alone
0 (0%)
2 (2.0%)
2 (2.0%)
peers
2 (2.0%)
2 (2.0%)
4 (4.0%)
partner
1 (1.0%)
3 (3.0%)
4 (4.0%)
other§
4 (4.0%)
6 (5.9%)
10 (9.9%)
Average hr spent (per week) working¶
0
12
2
8
2
9
training
15.6*
7.0
12.4*
7.1
13.6
7.2
studying¶
5
27
10
30
8
28
traveling
8.4
5.6
8.2
4.6
8.2
5.0
†Includes:
baseball, basketball, football, hockey, netball, rugby league. aerial skiing, diving, equestrian, golf, kayaking/rowing/canoe, sailing, track and field. §Includes living in a dorm or living part time in one situation e.g., At home with parents and part time elsewhere e.g., with peers at training facility or with partner. ¶Data = median and interquartile range. *t=2.17, p18 y (n = 32)
Part C/10 points
Mean % 60.9%
63.8% Female (n = 64)
Part B/70 points
65.7
13.6
58.1% 3.0
65.5
11.7
58.0%
Note. Part A: Dietary recommendations, Part B: Sources of nutrients, Part C: Choosing foods, Part D: Diet-disease relationship. Parts A, C and D used Mann-Whitney test to compare scores on sex, age group, education category and individual/team sport. Part B and Total GNKQ used independent t test. †Includes: aerial skiing, diving, equestrian, golf, kayaking/rowing/canoe, sailing, track and field. ‡Includes: baseball, basketball, football, hockey, netball, rugby league. *p = .03 (Independent t test analysis) **p = .017(Independent t test analysis)
observed between participants below or above 18 years (p = .230–.846), however the total knowledge score increased on average by 0.4 of a point for each year increase in age. No differences were observed between participants of state, national or international caliber (p = .357–0.791) or those competing in individual or team sports (p = .293–0.946) nor between participants with differing levels of education (p = .437–.830). As most (80.2%) participants lived with at least one parent, the influence of living situation was not able to be assessed. Participants who reported having previous professional contact with a dietitian (29.7%) had a higher overall score on the GNKQ (61.6% versus 56.6%; t=2.16, p = .034) than for those who had not, however differences on subsection scores for Parts A-D (p = .372–.602) were not significant.
Dietary Intake Overall A-ARFS was 33.6 (6.4) out of the possible 64 points (52.5%). Highest scoring domains were intake of water (70.0%), other (condiments and sauces; 70.0%), fruit (61.8%) and grains (60.0%) with dairy (52.2%)
and protein intake (35.7% from animal and 34.0% from vegetarian based) lowest (Table 4). A significant effect of gender was found for the A-ARFS with females (54.2%) scoring higher than males (49.4%) (t=-2.45, p = .016). Females also scored higher for vegetables (t=-3.55, p = .001), fruit (t=-2.80, p = .006) and water (t=-2.19, p = .031). No differences were observed between genders for grains (Mann-Whitney; p = .239), dairy (Mann-Whitney; p = .963), animal protein (Mann-Whitney; p = .495) or other (Chi-squared = 0.106, df = 2, p = .949). Males scored better for the vegetarian protein group (MannWhitney; p = .008) (Table 4). Gender independently explained 5.7% of the variance in A-ARFS (p = .016) and 11% for vegetable group (p = .001). No effect of age, level of education, sport type, athletic caliber or professional dietetic contact was observed for the overall or food group A-ARFS.
Relationship Between Nutrition Knowledge and Diet Quality A significant positive weak association between overall GNKQ score and total A-ARFS (r = .261, p = .008)
IJSNEM Vol. 25, No. 3, 2015
248 Spronk et al.
Table 4 Mean Scores on the Adapted—Australian Recommended Food Score Male (n = 37) Mean A-ARFS
31.6*
/64 points
49.4%
Female (n = 64)
SD
Mean
SD
7.4
34.7*
5.5
54.2%
Total (n = 101) Mean
SD
33.6
6.4
52.5%
Scoring domains 11.3**
9.8**
/20 points
49.0%
56.5%
6.3**
7.3**
fruit /11 points protein meat/fish/chicken) /7 points protein vegetarian Downloaded by Purdue Univ on 09/17/16, Volume 25, Article Number 3
3.7
vegetables
/5 points grains /9 points dairy /9 points water /1 point other /2 points
2.6
57.3% 2.6
1.0 1.0 1.4
75.0%
5.5
1.5
4.7
0.5
0.8*
1.1
1.4 70.0%
2.3
2.5
1.1
1.7
1.1
34.0% 1.2
5.4
1.3
60.0% 1.6
4.7
1.6
52.2% 0.4
80.0% 0.6
6.8
35.7%
52.2%
50.0% 1.5
1.1
61.1%
53.3% 0.5*
1.4**
3.3
61.8%
28.0%
57.8% 4.8
2.0
34.3%
42.0% 5.2
2.4
10.5 52.5%
66.4%
37.1% 2.1**
3.0
0.7
0.5
70.0% 0.7
1.4
0.7
70.0%
Note. A-ARFS= Adapted Australian Recommended Food Score *p
Original Research
Relationship Between General Nutrition Knowledge and Dietary Quality in Elite Athletes Inge Spronk, Susan E. Heaney, Tania Prvan, and Helen T. O’Connor This study investigated the association between general nutrition knowledge and dietary quality in a convenience sample of athletes (≥ state level) recruited from four Australian State Sport Institutes. General nutrition knowledge was measured by the validated General Nutrition Knowledge Questionnaire and diet quality by an adapted version of the Australian Recommended Food Score (A-ARFS) calculated from food frequency questionnaire data. Analysis of variance and linear modeling were used to assess relationships between variables. Data: mean (Standard Deviation). A total of 101 athletes (Males: 37; Females: 64), 18.6 (4.6) years were recruited mainly from team sports (72.0%). Females scored higher than males for both nutrition knowledge (Females: 59.9%; Males: 55.6%; p = .017) and total A-ARFS (Females: 54.2% Males: 49.4%; p = .016). There was no significant influence of age, level of education, athletic caliber or team/individual sport participation on nutrition knowledge or total A-ARFS. However, athletes engaged in previous dietetic consultation had significantly higher nutrition knowledge (61.6% vs. 56.6%; p = .034) but not total A-ARFS (53.6% vs. 52.0%; p = .466). Nutrition knowledge was weakly but positively associated with total A-ARFS (r = .261, p= .008) and A-ARFS vegetable subgroup (r = .252, p = .024) independently explaining 6.8% and 5.1% of the variance respectively. Gender independently explained 5.6% of the variance in nutrition knowledge (p= .017) and 6.7% in total A-ARFS (p = .016). Higher nutrition knowledge and female gender were weakly but positively associated with better diet quality. Given the importance of nutrition to health and optimal sports performance, intervention to improve nutrition knowledge and healthy eating is recommended, especially for young male athletes. Keywords: nutrition assessment, nutritional status, sport nutrition, nutrition knowledge Numerous factors including taste, convenience, cultural and religious beliefs, food security and availability (Parmenter & Wardle, 1999, Wardle et al., 2000, Hendrie et al., 2008a, Heaney et al., 2011) are reported to impact dietary intake. In athletes, a desire to increase lean and reduce fat mass or periodise fuels for training or competition are also known to be influential (American Dietetic Association et al., 2009). Although nutrition knowledge is considered to be a pivotal factor (Spendlove et al., 2012), its influence on dietary intake surprisingly remains relatively unexplored in athletes (Heaney et al., 2011). Nutrition education programs are designed to increase nutrition knowledge with the expectation that this translates into improved dietary intake, better health and for athletes, enhanced sports performance. Spronk is with the Division of Human Nutrition, Wageningen University, Wageningen, Netherlands. Heaney and O’Connor are with the Discipline of Exercise and Sport Science, University of Sydney, Sydney, Australia. Prvan is with the Dept. of Statistics, Faculty of Science, Macquarie University, Sydney, Australia. Address author correspondence to Helen T. O’Connor at [email protected].
Nutrition knowledge in athletes is generally reported to be similar or better than the general population but study quality is poor and use of validated instruments rare (Heaney et al., 2011). Demographic factors such as female gender, higher athletic caliber, and engagement in physique-oriented sports have been associated with higher nutrition knowledge (Raymond-Barker et al., 2007). In nonathletic groups, better-educated, middle aged (compared with younger or older age) participants tend have higher nutrition knowledge (Parmenter et al., 2000, Wardle et al., 2000, Hendrie et al., 2008a). In studies examining the association between nutrition knowledge and dietary intake in either athletes or the general population, the relationship is reported to be either null or weak (r < .5) (Spronk et al., 2014). Nutrition knowledge is regarded as a difficult construct to measure (Worsley, 2002). In athlete studies, assessment has included general concepts relevant to maintaining health (e.g., eating more dietary fiber and less saturated fat) (Abood et al., 2004, Spendlove et al., 2012) through to sports specific nutrition knowledge (e.g., an understanding of higher energy, protein, carbohydrate, and fluid needs for sports performance) (Rash et al., 2008). Assessment of sports specific nutrition
243
Downloaded by Purdue Univ on 09/17/16, Volume 25, Article Number 3
244 Spronk et al.
knowledge can be difficult and although the general principles are similar, there are differences in dietary requirements between sports (Burke & Deakin, 2006). However, general nutrition knowledge principles which encompass healthy eating are relevant to all athletes and these are also consistent with public health guidelines (National Health & Medical Research Council, 2003). General nutrition knowledge which examines understanding of healthy eating guidelines would also be expected to be more strongly related to overall dietary intake or quality than specific sports nutrition knowledge such as hydration requirements, preevent or recovery eating. To optimize performance, athletes must know how to consume a healthy, balanced diet and combine this with specific sports nutrition strategies relevant to their training and competition situation (American Dietetic Association et al., 2009). Examining associations between nutrition knowledge and diet is challenging as dietary intake is also difficult to measure accurately. Dietary assessment can be more problematic in athletes due to larger energy intakes and portion sizes, wide daily variation due to periodised training programs and consumption of food and beverages (often specific sports foods) during training sessions or competition where precise recording and then nutrient analysis is more difficult (Heaney et al., 2010). Although numerous studies have reported on inadequate nutrient intake in athletes, those with higher energy intake often meet micronutrient requirements due to the sheer volume of food consumed (Heaney et al., 2010). Recently, examination of dietary intake via a diet quality score has been employed to assess positive dietary attributes (e.g., intake of a greater variety of vegetables, fruits, wholegrains and reduced fat dairy foods) (Collins et al., 2008, Guenther et al., 2008). An Australian Recommended Food Score (ARFS), derived primarily from dietary intakes measured by a Food Frequency Questionnaire (FFQ; Cancer Council Victoria, 2011) has been used to assess diet quality in the community (Collins et al., 2008) but has not been applied to assess the diet of athletes. Therefore, the aim of this study was to evaluate the relationship between nutrition knowledge and the quality of dietary intake in a sample of elite Australian athletes. The study also aimed to examine the association and underpinning demographic factors linking nutrition knowledge and quality of dietary intake in this population.
Methods Participants Participants were a convenience sample of elite athletes (≥ state level; ≥ 16 years) who had a monetary scholarship from the Australian government to support development of their athletic careers at one of four nonresidential state institutes/academies of sport in Australia. To increase participant numbers, elite rugby league players (18–20 years) competing in the Australian National Youth Competition
were also recruited. Participants were a subset of a larger study reporting on the nutrition knowledge of athletes (Spendlove et al., 2012). The study was approved by the Human Ethics Committee of the University of Sydney. Participation was anonymous and informed written consent before study commencement was obtained (for participants 1/week for each of the following; potato (boiled, mashed, baked); pumpkin; cauliflower; green beans; spinach; cabbage or brussel sprouts; peas; broccoli; carrots; zucchini, eggplant, squash; capsicum; corn, sweet corn, corn on the cob; mushrooms; tomatoes; lettuce; celery, cucumber; avocado; onion, spring onion, leek
Fruits
≥ 2 pieces of fruit/day; >1/week for each of the following; canned fruit, e.g., peaches, two fruits; fresh fruit salad; dried fruit, e.g., sultanas, dried apricots; apple or pear; orange, mandarin, grapefruit; banana; peach, nectarine, plum or apricot; pineapple; grapes, strawberries, blueberries; melon, e.g., watermelon, rockmelon, honeydew melon
11
Protein foods Meat/ fish/chicken
1–4/week of each of the following; mince dish, e.g., spaghetti bolognese, rissoles, shepherd’s pie, lasagne; meat (beef or lamb), e.g., roast, chops, steak, beef stroganoff, stir fry (with or without sauce); chicken, e.g., roast, BBQ, satay, stir fry (with or without sauce); pork, e.g., roast, chops, sweet & sour, stir fry (with or without sauce); >1/week of each of fresh fish not crumbed or battered; canned tuna, salmon, sardines including patties; other seafood, e.g., prawns, lobster
7
Protein foods Vegetarian alternatives
>1/week of each of the following; nuts, e.g., peanuts, almonds; peanut/nut butter, ; eggs, e.g., boiled, scrambled; baked beans; other beans, lentils, e.g., chickpeas, split peas
5
Grains
Usually eating ‘other’ bread; >1/week of each of muesli; porridge; breakfast cereal; bread, pita bread, roll or toast all types; English muffin, bagel or crumpet; rice; pasta, e.g., spaghetti, lasagna, pasta bake
Usually eating brown (multigrain, wholemeal) bread (2 points)
9
Dairy
≥ 2 serves of a glass of milk, a tub of yoghurt or a slice of cheese per day; usually drinking reduced fat milk, skim milk or soy milk; 1/week—1/day of each of flavored the following; milk, hot chocolate, milkshake, thickshake, smoothie; ice cream—vanilla, chocolate, strawberry, sundaes, cones; >1/day of each of plain milk—glass or with cereal; 1/day—3/day of each of cheese, including cheese on sandwiches, biscuits or on toast
Italics: combination of these answers (additional 1 point)
9
Water
≥ 4 glasses of water/day
1
Other
>1/week of each of Vegemite, Mighty Mite, Promite, Marmite; tomato sauce, barbecue sauce
2
Total
64
†Adapted from Australian Recommended Food Score (Collins et al., 2008)
Linear Model to the data, assumptions were checked by examining the residuals versus fitted values plot and the normal probability plot of the standardized residuals. When variables were not normally distributed, Spearman’s Rank Correlation coefficient (rs) was used instead of Pearson’s. Results were reported as mean (SD) unless otherwise indicated. If the normal distribution assumption was violated the Mann-Whitney test was performed instead of the independent t test when comparing two groups. Significance was accepted at p < .05 level. Analysis of dietary intake data consisted primarily of one-way analysis of variance (ANOVA) or Kruskall-Wallis if the normality assumption did not hold. Independent t tests were conducted to identify any gender-related mean differences when data were normally distributed and Mann Whitney test if data were not. Chi square tests for asso-
ciations between variables and groups were conducted when data were categorical.
Results Participant Characteristics Demographic characteristics of participants (n = 101) are summarized in Table 2. Females comprised 63.4% of the sample with a high representation of younger (16–18 years; 68.3%) and team sport athletes (72.3%) and low representation from physique oriented sports (diving; n = 1). Participants mean age was 18.6 (4.6) years with no difference between genders (p = .137). Most participants were competing at national or international level (79.2%).
IJSNEM Vol. 25, No. 3, 2015
246 Spronk et al.
Table 2 Participant Characteristics Male (n = 37) (36.6%) Variable age (y)
Mean (%)
SD
17.6
1.2
Female (n = 64) (63.4%) Mean (%)
SD
19.1
5.6
Total (N = 101) (100%) Mean (%)
SD
18.6
4.6
athletes >18
10 (9.9%)
22 (21.8%)
32 (31.7%)
athletes 16–18
27 (26.7%)
42 (41.6%)
69 (68.3%)
30 (29.7%)
43 (42.6%)
73 (72.3%)
7 (6.9%)
21(20.8%)
28 (27.7%)
first year
21 (20.8%)
33 (32.7%)
54 (53.5%)
1–3 years
11 (10.9%)
20 (19.8%)
31 (30.7%)
4+ years
5 (5.0%)
11 (10.9%)
16 (15.8%)
Sport team† individual‡ Years on scholarship
Downloaded by Purdue Univ on 09/17/16, Volume 25, Article Number 3
Highest level of competition state
9 (8.9%)
12 (11.9%)
21 (20.8%)
national
15 (14.9%)
26 (25.7%)
41 (40.6%)
international (junior or open)
13 (12.9%)
26 (25.7%)
39 (38.6%)
high school
26 (25.7%)
43 (42.6%)
69 (68.3%)
tertiary
11 (10.9%)
21 (20.8%)
32 (31.7%)
Highest level of education
Living situation parents (including other family situation)
30 (29.7%)
51 (50.5%)
81 (79.2%)
alone
0 (0%)
2 (2.0%)
2 (2.0%)
peers
2 (2.0%)
2 (2.0%)
4 (4.0%)
partner
1 (1.0%)
3 (3.0%)
4 (4.0%)
other§
4 (4.0%)
6 (5.9%)
10 (9.9%)
Average hr spent (per week) working¶
0
12
2
8
2
9
training
15.6*
7.0
12.4*
7.1
13.6
7.2
studying¶
5
27
10
30
8
28
traveling
8.4
5.6
8.2
4.6
8.2
5.0
†Includes:
baseball, basketball, football, hockey, netball, rugby league. aerial skiing, diving, equestrian, golf, kayaking/rowing/canoe, sailing, track and field. §Includes living in a dorm or living part time in one situation e.g., At home with parents and part time elsewhere e.g., with peers at training facility or with partner. ¶Data = median and interquartile range. *t=2.17, p18 y (n = 32)
Part C/10 points
Mean % 60.9%
63.8% Female (n = 64)
Part B/70 points
65.7
13.6
58.1% 3.0
65.5
11.7
58.0%
Note. Part A: Dietary recommendations, Part B: Sources of nutrients, Part C: Choosing foods, Part D: Diet-disease relationship. Parts A, C and D used Mann-Whitney test to compare scores on sex, age group, education category and individual/team sport. Part B and Total GNKQ used independent t test. †Includes: aerial skiing, diving, equestrian, golf, kayaking/rowing/canoe, sailing, track and field. ‡Includes: baseball, basketball, football, hockey, netball, rugby league. *p = .03 (Independent t test analysis) **p = .017(Independent t test analysis)
observed between participants below or above 18 years (p = .230–.846), however the total knowledge score increased on average by 0.4 of a point for each year increase in age. No differences were observed between participants of state, national or international caliber (p = .357–0.791) or those competing in individual or team sports (p = .293–0.946) nor between participants with differing levels of education (p = .437–.830). As most (80.2%) participants lived with at least one parent, the influence of living situation was not able to be assessed. Participants who reported having previous professional contact with a dietitian (29.7%) had a higher overall score on the GNKQ (61.6% versus 56.6%; t=2.16, p = .034) than for those who had not, however differences on subsection scores for Parts A-D (p = .372–.602) were not significant.
Dietary Intake Overall A-ARFS was 33.6 (6.4) out of the possible 64 points (52.5%). Highest scoring domains were intake of water (70.0%), other (condiments and sauces; 70.0%), fruit (61.8%) and grains (60.0%) with dairy (52.2%)
and protein intake (35.7% from animal and 34.0% from vegetarian based) lowest (Table 4). A significant effect of gender was found for the A-ARFS with females (54.2%) scoring higher than males (49.4%) (t=-2.45, p = .016). Females also scored higher for vegetables (t=-3.55, p = .001), fruit (t=-2.80, p = .006) and water (t=-2.19, p = .031). No differences were observed between genders for grains (Mann-Whitney; p = .239), dairy (Mann-Whitney; p = .963), animal protein (Mann-Whitney; p = .495) or other (Chi-squared = 0.106, df = 2, p = .949). Males scored better for the vegetarian protein group (MannWhitney; p = .008) (Table 4). Gender independently explained 5.7% of the variance in A-ARFS (p = .016) and 11% for vegetable group (p = .001). No effect of age, level of education, sport type, athletic caliber or professional dietetic contact was observed for the overall or food group A-ARFS.
Relationship Between Nutrition Knowledge and Diet Quality A significant positive weak association between overall GNKQ score and total A-ARFS (r = .261, p = .008)
IJSNEM Vol. 25, No. 3, 2015
248 Spronk et al.
Table 4 Mean Scores on the Adapted—Australian Recommended Food Score Male (n = 37) Mean A-ARFS
31.6*
/64 points
49.4%
Female (n = 64)
SD
Mean
SD
7.4
34.7*
5.5
54.2%
Total (n = 101) Mean
SD
33.6
6.4
52.5%
Scoring domains 11.3**
9.8**
/20 points
49.0%
56.5%
6.3**
7.3**
fruit /11 points protein meat/fish/chicken) /7 points protein vegetarian Downloaded by Purdue Univ on 09/17/16, Volume 25, Article Number 3
3.7
vegetables
/5 points grains /9 points dairy /9 points water /1 point other /2 points
2.6
57.3% 2.6
1.0 1.0 1.4
75.0%
5.5
1.5
4.7
0.5
0.8*
1.1
1.4 70.0%
2.3
2.5
1.1
1.7
1.1
34.0% 1.2
5.4
1.3
60.0% 1.6
4.7
1.6
52.2% 0.4
80.0% 0.6
6.8
35.7%
52.2%
50.0% 1.5
1.1
61.1%
53.3% 0.5*
1.4**
3.3
61.8%
28.0%
57.8% 4.8
2.0
34.3%
42.0% 5.2
2.4
10.5 52.5%
66.4%
37.1% 2.1**
3.0
0.7
0.5
70.0% 0.7
1.4
0.7
70.0%
Note. A-ARFS= Adapted Australian Recommended Food Score *p
