In Canada, poor dietary quality is a leading contributor to the burden of chronic diseases, including type 2 diabetes, cardiovascular diseases, and some cancers [1
]. Moreover, the economic cost associated with not meeting national dietary recommendations is estimated to contribute substantially to the Canadian health care system [2
]. Analyses of nationally representative dietary data from the 2004 Canadian Community Health Survey (CCHS) suggest that large gaps exist between reported dietary intakes and national recommendations for healthy eating. For example, in 2004, only 26% of the population aged 2 years and older met the minimum number of daily servings of vegetables and fruit recommended by Eating Well with Canada’s Food Guide (referred to here as the 2007 CFG) for their respective age–sex group [3
]. Meanwhile, approximately 22% of total daily calories consumed by Canadians aged 4 years and older in 2004 came from minimally nutritious foods, such as high fat and/or high sugar foods and sugar-sweetened beverages [4
Between 2007–early 2019, national dietary recommendations advocated by the 2007 CFG (the food guide in use during the time of this study’s writing) [5
] included specific recommendations regarding amounts and types of foods to consume within each of four core food groups: Vegetables and fruit, grain products, milk and alternatives, and meat and alternatives [6
]. Within a food group (e.g., vegetables and fruit), foods can vary in their nutritional content. Therefore, the 2007 CFG included statements on the quality of food choices within each food group to promote nutrient adequacy within a food intake pattern [6
]. For example, the statement “Have vegetables and fruit more often than juice” was issued to maintain the fibre content of the food intake pattern, while the statement “Eat at least one dark green and one orange vegetable each day” aimed at achieving adequate levels of dietary folate and vitamin A [6
]. Within grain products, the 2007 CFG emphasized the consumption of whole grains over non-whole grains to achieve optimal intakes of magnesium and fibre [6
]. The guidance recommending daily consumption of low fat milk (skim, 1% or 2%) was based on “the effectiveness of obtaining adequate calcium and vitamin D while remaining within an appropriate macronutrient profile and total amount of calories” within a food intake pattern [6
]. The Canadian dietary guidelines have recently undergone major revisions [7
] and understanding whether Canadians have changed the average quantity and composition of food choices over time can provide foundational knowledge to help to inform future research and public health practice aimed at improving national nutritional outcomes.
National-level analyses based on dietary data from the 2004 CCHS have previously shown differences in food group intake according to age group. In 2004, 7 out of 10 children aged 4–8 years reportedly consumed less than 5 servings of vegetables and fruit, whereas around half of adults fell short of the 1992 CFG recommended minimum of 5 servings a day (the recommendations in place at the time of the 2004 CCHS) [4
]. Black and Billette showed that in 2004, usual intakes of fruit juices were higher among children and adolescents compared to adults, while adults consumed higher usual intakes of dark green vegetables compared to children [3
]. In 2004, Canadian adolescents had, on average, the highest proportion of their total daily calories (25%) obtained from minimally nutritious foods compared to other age groups [4
]. Whether food intake reported in 2004 across age groups has changed in more recent years remains unknown.
In 2015, Statistics Canada carried out another nationally representative dietary survey (the CCHS-2015 Nutrition), with a population-representative sample and dietary assessment methods similar to the 2004 CCHS [8
]. Recent analyses have examined differences over time in energy and macronutrient intakes reported by Canadian children and adults [9
]. Still, little published work to date has described differences over time in Canadians’ average intakes of foods and beverages. Therefore, the purpose of this study was to examine differences in the quantity and types of foods and beverages reported by Canadians aged 2 years and older from 2004 to 2015. A secondary objective was to test whether dietary differences reported between 2004 and 2015 varied across age groups.
4. Vegetables and Fruit
In 2004, Canadians aged 2 years and older reported consuming, on average, ~5.2 daily servings of total vegetables and fruit. From 2004 to 2015, the amount of daily servings of total vegetables and fruit decreased significantly, both among all energy reporters (−0.7 servings/day) and well as in analyses restricted to only plausible energy reporters (−0.6 servings/day). In 2004, the largest contributor to total vegetables and fruit were other vegetables (i.e., non-dark green and orange vegetables such as cucumbers, tomatoes, celery, corn), which represented ~33% of total daily servings of vegetables and fruit. In 2015, the largest contributors to total vegetables and fruit were both other vegetables and whole fruit (each representing ~28% of all daily servings of vegetables and fruit).
Among all energy reporters (all ages combined), Canadians reported consuming significantly more average daily servings of dark green and orange vegetables (+0.1 servings/day) but fewer other vegetables (−0.4 servings/day), potatoes (−0.1 servings/day), and fruit juices (−0.2 servings/day). Statistically significant differences for other vegetables, potatoes, and fruit juices were found in analyses restricted to plausible energy reporters. In both survey years, average intakes of dark green and orange vegetables and other vegetables were highest among adults and lowest among adolescents and children. Average intake of fruit juices was highest among children and adolescents and lowest among older adults in both 2004 and 2015.
The magnitude of the reduction in total vegetables and fruit varied by age group (p-value for the overall interaction <0.001). Adolescents, adults, and older adults reported, on average, significantly fewer daily servings of total vegetables and fruit in 2015 than in 2004, whereas children reported no difference over time. The changes in daily intakes of other vegetables, whole fruit, and fruit juice also varied by age groups (p-values for the overall interactions <0.05). For example, adolescents, adults and older adults (all energy reporters) reported significantly fewer average daily servings of other vegetables in 2015 compared to 2004, whereas children (all energy reporters) reported no difference over time. Although children (all energy reporters) and adolescents (all energy reporters and only plausible energy reporters) reported, on average, significantly more daily servings of whole fruit in 2015 compared to 2004, no difference was found among adults and older adults.
7. Meat and Alternatives
In 2004, Canadians (all energy reporters, all ages combined) consumed, on average, ~2.2 daily servings of meat and alternatives. The largest contributor to meat and alternatives were meat and poultry (~55% and ~50% of total daily intake of meat and alternatives in 2004 and 2015, respectively). In 2004, the second largest contributors to meat and alternatives were legumes, nuts, and seeds combined and processed meats (each contributing to ~14% of meat and alternatives consumed daily). In 2015, the second largest contributor to meat and alternatives were legumes, nuts, and seeds combined (~17% of total daily servings of meat and alternatives consumed). From 2004 to 2015, Canadians (all age groups combined) reported consuming significantly more meat and alternatives (+0.2 servings/day among all energy reporters and among only plausible reporters). Canadians (all age groups combined) reported significant increases in their average daily intake of legumes, nuts, and seeds (+0.1 servings/day) and eggs (+0.1 servings/day among plausible reporters), with no change in other meat and alternatives subgroups. In both survey cycles, mean daily intake of meat and alternatives was highest among adolescents and adults and lowest among children.
The magnitude of the differences in total daily intakes of meat and alternatives varied by age group (p-value for the overall interaction <0.001). Adults reported the greatest change in intake of meat and alternatives. Adults reported consuming, on average, significantly more meat and alternatives in 2015 compared to 2004 (+0.3 servings/day and +0.5 servings/day among all energy reporters and among only plausible reporters, respectively). Adults (plausible energy reporters) also reported significantly more meat and poultry (+0.2 servings/day) in 2015. By contrast, older adults reported significantly fewer daily servings of meat and poultry (−0.1 servings/day and −0.3 servings/day among all energy reporters and among only plausible intake reporters, respectively). No significant difference in meat and poultry intake was reported among other age groups.
7.1. Mean Daily Intakes from Other Foods and Beverages: 2004 vs. 2015
In 2004, Canadians (all energy reporters, all ages combined) reported consuming, on average, 126 kcal/day from high fat and/or high sugar foods as well as 83 and 22 daily kcal from high- and low-calorie beverages, respectively. From 2004 to 2015, Canadians (all energy reporters, all age groups combined) consumed significantly fewer kcal from high-calorie beverages (for, e.g., non-diet sodas, fruit drinks, sweetened iced tea) (−32 kcal/day), a 39% reduction compared to what was reported a decade earlier. Similar reductions in energy from high-calorie beverages were reported among plausible reporters.
The magnitude of the reduction in high-calorie beverage intake varied by age group (p-value for the overall interaction <0.001), but all age groups reported significantly fewer mean daily calories from these beverages from 2004 to 2015. For example, children aged 6-12 years and adolescents (all energy reporters) decreased their average daily calories from these beverages by 58 kcal/day and 73 kcal/day from 2004 to 2015, respectively. Meanwhile, adults and older adults (all energy reporters) reduced their caloric intake from such beverages by only 30 and 12 kcal/day, respectively.
7.2. Relative Percent Change in Food and Beverages Intake from 2004 to 2015
Average relative percent differences in food group intakes between 2004 and 2015 estimated for all energy reporters and among only plausible energy reporters (all age groups combined) are shown in Figure 2
. For 11 out of 22 dietary variables examined, the relative changes in daily intakes were small (i.e., less than a 10% relative difference). However, daily intakes of dark green and orange vegetables, other milk products (i.e., cheese, yogurt), legumes, nuts and seeds, eggs, and alcoholic beverages all increased, on average, by at least 10% or more from 2004 to 2015 (all energy reporters). By contrast, mean daily intakes of total vegetables and fruit, potatoes, other vegetables, fruit juices, fluid milk, and high-calorie beverages all declined by at least 10% over the same time period. For some food subgroups such as fruit juices and high-calorie beverages, these changes were substantial (>25% relative difference), which translated into a ~0.2 fewer daily servings of fruit juices and ~32 fewer daily calories from sugary beverages. Although the relative percent change was large for legumes, nuts and seeds (a 31% increase), and eggs (a 24% increase), the absolute differences over time were small (a difference of ≤0.1 servings/day of meat and alternatives).
This study compared differences in food and beverage intake estimated from nationally representative surveys of Canadians collected in 2004 and 2015. While some aspects of the Canadian diet improved over time (for example, fewer daily kcal from sugar-sweetened beverages), mean daily intake of several types of foods recommended as part of the 2007 CFG and the newly released 2019 Canada’s Dietary Guidelines [7
] (e.g., vegetables, whole fruit, whole grains) have either stagnated or worsened over time. These findings are in line with Canadian food balance sheet data which suggest that between 2004 and 2013, the availability of vegetables, milk products, sugars, and sweeteners declined while that of eggs, pulses, and legumes increased [29
]. Finally, results found here are also consistent with recent CCHS analyses documenting declines in Canadians’ intake of fluid milk and sugar-sweetened beverages from 2004 to 2015 [11
We found that Canadians aged 2 years and older reported significantly fewer daily calories consumed in 2015 compared to 2004, and this finding was consistent before and after excluding under- and overreporters. These results are in line with other national-level analyses which have reported changes to energy reporting status and average energy intakes by Canadians from 2004 to 2015 [11
]. These findings also echo US research suggesting little change in population-level BMI alongside small declines in population-level energy intakes from 2003–2004 to 2009–2010 [30
From 2004 to 2015, Canadians reported fewer daily servings of vegetables and fruit, and milk and alternatives and less energy from high-calorie beverages, while reporting higher intakes of meat and alternatives. Excluding under- and overreporters and adjusting for covariates (daily energy intake and other sociodemographic shifts from 2004 to 2015) did not change the direction and statistical significance of these differences. The release of the 2007 CFG took place between the survey years examined here. Some of the differences reported over time were consistent with dietary guidelines issued in the 2007 CFG, including increased daily servings of dark green and orange vegetables, legumes, nuts, and seeds and less energy from high-calorie beverages. The shift towards consuming more plant-based proteins (more legumes, nuts, and seeds) was also in line with recent calls to emphasize more plant-based and environmentally sustainable sources of protein, in Canada [7
] and globally [31
]. Still, the average amount consumed of many healthful dietary components emphasized in the 2007 CFG (whole fruit, whole grains, fish and shellfish) did not increase over time.
The shift towards consuming fewer servings of fluid milk could be concerning from a population health perspective if fluid milk is not replaced with other food sources rich in calcium and vitamin D [32
]. In 2004, national-level analyses suggested that a large proportion of Canadians were not meeting their age and sex-based recommended daily servings for milk and alternatives recommended by the 2007 CFG [4
]. Consequently, in 2004, many Canadian adolescents [33
], adults [34
] and older adults [35
] had inadequate intakes of calcium and vitamin D. While our study did not explicitly examine nutrient adequacy (which would have required modelling usual intake distributions), the declines found in population-level mean intakes of milk and alternatives and fluid milk suggests that improvement in the adequacy of these nutrients of concern is unlikely and merits further attention.
Apart from dark green and orange vegetables, estimated intakes of vegetables and fruit generally either stagnated or decreased from 2004 to 2015. In 2004, only 26% of the Canadian population consumed the 2007 CFG minimum number of servings of vegetables and fruit recommended for their respective age–sex group [2
], and there were concerns around low intakes of potassium and fibre among both adults [33
] and children [36
]. The population-level decline in vegetables and fruit intake from 2004 to 2015 suggests that more effective efforts are needed to help Canadians move closer towards national dietary recommendations which, as of 2019, encourage having “plenty of vegetables and fruits” [7
Other international studies examining secular trends in food group intakes demand cautious comparison, as survey methodologies, time frame of data collection, and dietary habits differ across countries. Some international studies drawing from national diet surveys have identified similar shifts in food groups and beverage intakes. In the US, studies drawing from the National Health and Nutrition Examination Surveys suggest that US youth and adults decreased their sugar-sweetened beverages intake by 68 kcal/day and 45 kcal/day from 1999–2000 to 2009–2010, respectively [37
]. Similarly, Rehm et al. [38
] reported that US adults decreased their mean daily intake of sugar-sweetened beverages by, on average, 0.49 servings/day from 1999 to 2012. Since the beginning of the 21st
century, US studies have reported population-level declines in minimally nutritious foods [39
] and trans
] but no improvements in other dietary components such as vegetables, dairy products [39
], or sodium [39
]. The differences over time observed in this study were similar in some ways to the US trends: lower amounts of sugar-sweetened beverages, fruit juices, along with little change or decline in some healthier dietary components, such as vegetables and milk and alternatives.
We conducted sensitivity analyses to examine the impact of energy misreporting on differences over time in food group intakes across age groups. In sensitivity analyses that only included plausible energy reporters (all age groups combined), the direction and statistical significance of the temporal difference for most dietary variables (19 out of 22) were consistent with those differences found in analyses which included all energy intake reporters. This suggests that the changes in the proportion of energy misreporting (i.e., greater energy underreporting in 2015 compared to 2004) had little impact overall on food group intake differences reported over time.
These findings suggest some differences in the magnitude of average dietary intake changes over time between children and adults. For example, children aged 2–12 years did not report significantly fewer servings of vegetables and fruit, whereas older age groups did. Moreover, some dietary improvements were observed among younger age groups (e.g., more whole grains among adolescents, more whole fruit among children) but not among adults. These findings are in line with national [41
] and international [40
] research suggesting modest improvements in dietary habits of children and adolescents since the early 2000s. For example, in Norway [42
] and Scotland [43
], data from the Health Behaviour in School-Aged Children study suggested improvement in children’s intakes of fruit, vegetables, and minimally nutritious foods (sweets and sugary beverages) from ~2000 to 2009–2010. In Canada, a recent analysis drawing from the CCHS 2004 and 2015 reported improved school hour and whole school day dietary quality among Canadian children aged 6–17 years [41
]. Increased vegetable and fruit intakes and decreased energy from minimally nutritious foods and beverages (observed both during school hours and, to a lesser extent, for the whole school day) accounted for most of the improvement in total diet quality scores over time [41
]. This is encouraging, since previous research based on the 2004 CCHS reported that 23–31% of daily calories consumed were derived from other foods and beverages among Canadian children and adolescents [44
Key strengths of this study include the use of large, nationally representative dietary surveys and the use of 24-h dietary recall data which capture detailed information on the quantity and types of foods consumed. However, there are important limitations that deserve consideration. First, these analyses did not model the distribution in usual intakes. Therefore, it is not possible to report whether the proportion of Canadians who met recommended daily intakes for specific 2007 CFG food groups or the proportion of the population who usually consume no foods from a given subgroup changed over time. Second, this study focused on examining change over time and differences among age groups, but further work is needed to more deeply explore if, how, and why specific dietary practices have changed over time among diverse population groups, including by gender, socioeconomic status or geographical context. Third, since income was measured differently between 2004 and 2015 [45
], these analyses did not control for income as a potential confounder in these models. It was also not possible to assess national-level changes in food security status between 2004 and 2015 since the CCHS 2015 did not include household-level survey weights [8
]. Thus, it is possible that income and/or food security may have acted as potential confounders in the change in dietary variables observed over time. Fourth, there were differences in the execution of the survey (e.g., different sample sizes, response rates, changes to the food booklet used to help estimate portion sizes) and data processing (e.g., changes to the nutrient databases used to analyze the 24-h dietary recalls) between survey cycles, which could have implications when comparing dietary intakes between survey years [8
]. For example, a lower response rate (61.6% in 2015 compared to 76% in 2004) increases the potential for non-response bias [8
]. There were also changes to the food model booklets regarding the images used to estimate beverage intakes from 2004 to 2015. Line drawings of the glasses, bowls, and plates in 2004 were replaced in 2015 with actual-size photographs that gave a more realistic three-dimensional view of the items, leading potentially to a downward bias when estimating beverage intakes [47
]. Moreover, differences in energy misreporting (specifically increases in energy underreporting and a decrease in energy overreporting from 2004 to 2015) could alter these findings. However, in sensitivity analyses that only included plausible energy reporters (all age groups combined), the direction and significance of the temporal change for most food/beverage subgroups (19 out of 22) was often consistent with differences found for the full sample. Finally, this analysis did not examine the quality of the food choices over time within each food group and subgroup using the Tier system. For example, we did not examine whether the proportion of Tier 4 foods increased or decreased over time within the food groups examined. Future research is needed to examine how the quality of Canadians’ food choices within food groups have changed over time.