Canadian Community Health Survey - Nutrition (CCHS)
Detailed information for 2015
The Canadian Community Health Survey - Nutrition is a national health survey that collected information from Canadians about their eating habits and use of nutritional supplements, as well as other health factors.
Data release - June 20, 2017
This survey will give a detailed and up-to-date picture of not only what people are eating and what vitamins and minerals they take, but the impact this has on health and well-being. It will also evaluate changes in food consumption, nutrition and health since this survey was last done in 2004.
The objectives of the Canadian Community Health Survey - Nutrition are:
(1) To collect detailed data on the consumption of foods and dietary supplements among a representative sample of Canadians at national and provincial levels;
(2) To estimate the distribution of usual dietary intake in terms of nutrients from foods, food groups, dietary supplements and eating patterns;
(3) To gather anthropometric (physical) measurements for accurate body weight and height assessment to interpret dietary intake;
(4) To support the interpretation and analysis of dietary intake data by collecting data on selected health conditions and socio-economic and demographic characteristics; and
(5) To evaluate changes in dietary intake from the 2004 CCHS - Nutrition.
The data collected from the survey will be used by Statistics Canada, Health Canada and the Public Health Agency of Canada, provincial and territorial ministries of health, as well as federal and provincial health planners across the country, industry and researchers and health professionals. Results from our surveys are used extensively for policy-making and program development that affect Canadian communities.
- Diseases and health conditions
- Lifestyle and social conditions
Data sources and methodology
The target population in 2015 covers the population 1 year of age and over living in the ten provinces. Excluded from the survey's coverage are: persons living on reserves and other Aboriginal settlements, full-time members of the Canadian Forces and the institutionalized population.
The Canadian Community Health Survey - Nutrition questionnaire was developed by Statistics Canada in collaboration with stakeholders from Health Canada and the Public Health Agency of Canada, the Provincial Health Ministries, and an expert advisory group consisting of specialists from Health Canada, the Public Health Agency of Canada and academic experts. Content was chosen using the following set of criteria:
- Comparability with previous cycle of CCHS Nutrition survey
- Issues identified as priority for the support/development of programs and policy, surveillance requirements and/or research
- Significant number of people affected by the targeted issue
- Significant impact on family, community, and health care costs
- Data has potential for health improvement with policy intervention
- Issues identified as data gaps from the stakeholder consultations
Nutrition questions are designed for computer-assisted interviewing (CAI), meaning that, as the questions were developed, the associated logical flow into and out of the questions was programmed. This includes specifying the type of answer required, the minimum and maximum values, on-line edits associated with the question and what to do in case of item non-response.
In collaboration with Statistics Canada's Questionnaire Design Resource Centre, the questionnaire was subjected to qualitative testing in May 2013, which consisted of one-on-one interviews. The objective was to evaluate respondent reactions to and understanding of the survey, as well as their willingness to respond to the questions. The qualitative testing was also used to obtain estimates for the various section times.
This is a sample survey with a cross-sectional design.
A stratified three-stage sample of one person from each dwelling in each geographic area (cluster) in 10 provinces.
The frame was created using Census Dissemination Areas (DAs) as building blocks for clusters. Neighbouring DAs were combined in such a way as to ensure as much as possible that clusters contain a minimum of 200 dwellings and not span too large a land area. Information from the Labour Force Survey (LFS) frame was used to identify areas of exclusion and remote areas, as well as to aid in decision-making when clusters did have a large land area.
To select the sample of respondents for the 2015 CCHS - Nutrition Survey, a stratified three-stage design was used in each province except Prince Edward Island. In the first stage, geographical areas called clusters were selected. A list frame of all households in the selected clusters was used to draw a sample of households. In the third stage a person was selected within the household from the household roster created at the start of the interview. In Prince Edward Island, a two-stage design was used, where the households are selected from a list frame in the first stage and persons from the roster in the second stage. For operational reasons, the island was still divided up into areas however each of these areas appears in the sample.
In all provinces except PEI, clusters were selected by province with systematic probability proportional to size sampling. In PEI, no sampling of clusters was done (every cluster appears with certainty). To ensure that seasonality was taken into account, the selected clusters were allocated to collection periods to ensure an even distribution of sample and an even distribution of urban and rural clusters throughout the year. Households were then selected in each cluster. Upon visiting a selected dwelling, a roster of members in the household was created and a person was randomly selected among eligible members. The probability of selection of persons varied in order to attain the targeted sample sizes in each domain. Finally, each respondent was given a probability of selection for a 2nd dietary recall, which also varied according to the respondent's DRI group.
SAMPLING AND SUB-SAMPLING
A sample of 24,000 respondents was desired at the national level. A minimum sample size in each province for each of twelve age-sex groups corresponding to Dietary Reference Intake (DRI) groups, was also targeted. These DRI groups are: ages 1-3, 4-8, 9-13 M, 9-13 F, 14-18 M, 14-18 F, 19-50 M, 19-50 F, 51-70 M, 51-70 F, 71+ M, and 71+ F.
A two-step strategy was used to allocate the sample to the provinces. First, 80 sample units were allocated to each DRI group in each province, accounting for 9600 units. The remaining 14,400 units were allocated to the provinces using a power allocation method with power q=0.7, based on the estimated population in each province. The total sample size of any given province was found by adding the sizes obtained in the two steps. The provincial sample was then allocated to the DRI groups based on a power allocation using q=0.5, to attain the targeted sample size in each DRI group in each province.
Sample sizes were enlarged to account for out-of-scope dwellings and anticipated non-response, for a total of 37,694 selected dwellings.
Data collection for this reference period: 2015-01-02 to 2015-12-31
Responding to this survey is voluntary.
Data are collected directly from survey respondents.
Before the start of each collection period, introductory letters and brochures explaining the purpose of the survey were sent to the sampled households.
Respondents were interviewed using a computer assisted personal interview (CAPI). Approximately 37% of respondents did a second dietary recall using a computer assisted telephone interview (CATI).
The format of the interview was determined by the age of the selected respondent:
- Selected respondent ages 1 to 5: Proxy only (The parent or guardian was asked to provide the child's information.)
- Selected respondent ages 6 to 11: Parent-assisted (The child was asked to provide his/her information with the help of his/her parent or guardian.)
- Selected respondent ages 12 and up: Non-proxy (The respondent was asked to provide his/her own information.)
It should be noted that proxy interviews for respondents aged 6 and older were to occur only if the mental or physical health of the selected member makes it impossible to complete the interview during the collection period.
The Licensed Natural Health Products Database (LNHPD) was used to code the nutritional supplements collected and to determine their nutrient strengths. This administrative database was downloaded from Health Canada's (HC) website. Although this a publicly available database, Statistics Canada did extensive processing to transform text fields into data fields usable for calculations. With the collaboration of HC nutrition experts, the version from December 31st 2015 was used to produce the nutrient profile per dosage unit of every nutritional supplement coded to a Natural Product Number (NPN).
View the Questionnaire(s) and reporting guide(s).
A number of edits were built into the CAI application used to collect the 2015 data. Incorrect entries were minimized due to the automated routing of questions and other logical consistency checks. Edits were also performed during processing. Inconsistencies were usually corrected by setting one or both of the variables in question to "Not stated". The percentage of records corrected is quite low: for example, for measured height and weight, less than 0.5% of records required edit corrections.
If household income was missing, this variable was imputed on the 2015 CCHS - Nutrition. Missing values due to either respondent refusal or respondent's lack of knowledge of household income were completed using statistical techniques. Approximately 25% of the values were imputed.
If a respondent's postal code is missing or invalid, it was imputed through a donor imputation process. Postal codes are used to derive the geographical variables on the 2015 CCHS - Nutrition data files. Approximately 3% of postal codes were imputed.
In order for estimates produced from survey data to be representative of the covered population, and not just the sample itself, users must incorporate the survey weights in their calculations. A survey weight is given to each respondent included in the final sample. This weight corresponds to the number of persons in the entire population that are represented by the respondent.
As described above, the CCHS - Nutrition uses the Household Survey Frame to select dwellings to be contacted for an interview.
The weighting strategy comes up with a person-level weight. The adjustments applied to the initial weights are based on modeling probabilities of response (at the household level and person level). Variables derived from the collection paradata as well as characteristics of the units are used to create the models. Then these probabilities are used to create groups of respondents and nonrespondents in which to transfer the weights of the nonrespondents to the respondents. The person-level weights then undergo two more adjustments (Winsorization and Calibration to known population totals such as by geography and age and sex), and become the final person-level weights.
Bootstrap weights are created through resampling the original sample and applying similar adjustments to the bootstrap weights as to the sample weights.
The steps for weighting are described in chapter 8 of the CCHS - Nutrition User Guide.
The sample design used for this survey was not self-weighting. That is to say, the sampling weights are not identical for all individuals in the sample. When producing simple estimates, including the production of ordinary statistical tables, users must apply the proper sampling weight.
Estimates of the number of people with a certain characteristic are obtained from the data file by summing the final weights of all records possessing the characteristic of interest.
Proportions and ratios are obtained by summing the final weights of records having the characteristic of the numerator and the denominator, and then dividing the first estimate by the second.
After the 2015 data were compiled, the data were reviewed in terms of the accuracy, coherence and overall reasonableness. A variety of summary indicators were calculated from the 2015 data and compared with the 2004 Nutrition data, 2015 CCHS-Annual data or other surveys such as the American National Health and Nutrition Examination Survey (NHANES), as appropriate. Any statistical differences were reviewed for reasonableness with respect to changes in the survey collection methodology, trends in food availability, supply, and consumption; and demographic and health trends.
Statistics Canada is prohibited by law from releasing any information it collects that could identify any person, business, or organization, unless consent has been given by the respondent or as permitted by the Statistics Act. Various confidentiality rules are applied to all data that are released or published to prevent the publication or disclosure of any information deemed confidential. If necessary, data are suppressed to prevent direct or residual disclosure of identifiable data.
For all CCHS data files, variables that directly identify a respondent have been removed from the master data files. For tabular data, small cells have been suppressed.
Revisions and seasonal adjustment
The quality of estimates produced with CCHS data is measured with the coefficient of variation (CV), produced using bootstrap weights. The CV magnitude will depend on the domain of interest and the prevalence of the characteristic.
In 2015, a high prevalence characteristic 'Very good or excellent self-perceived health' had a CV of 1.3% at the Canada level and no higher than 4.6% at the provincial level. A low prevalence characteristic 'Excludes gluten from diet' had a CV of 9.1% at the Canada level and up to 29.5% in some provinces. Disaggregating estimates further by age group or sex will increase the coefficient of variation.
In 2015, around 25% of respondents were imputed for total household income.
The overall estimation response rate at the Canada level is 62%. Response rates by province are provided in Chapter 9 of the 2015 CCHS-Nutrition User Guide.
Much time and effort was devoted to reducing non-sampling errors in the survey. Quality assurance measures were applied at each stage of the data collection and processing cycle to control the quality of the data. For example, within the computer-assisted questionnaire, there were edits that prompted the interviewer to confirm with the respondent any values of high amounts of food consumed.
The effect of non-response on survey results is a major source of non-sampling error in surveys. The scope of non-response varies from partial non-response (where the respondent does not respond to one or more questions) to total non-response.
Partial nonresponse to the income and postal code questions were treated by imputation, as described above. Partial nonresponse to all other variables is given as its own value in the data file, enabling users to tailor the treatment of partial nonresponse in their particular analysis.
There was total non-response when the person selected to participate in the survey refused to do so or could not be contacted by the interviewer. Cases of total non-response were taken into account during weighting by correcting the weights of persons who responded to the survey in order to compensate for those who did not respond.
In addition to increasing variance (accuracy), non-response can result in biased estimates if non-respondents have different characteristics from respondents. To reduce the number of non-response cases and to ensure that procedures are followed consistently, the interviewers were all extensively trained by Statistics Canada and provided with detailed interviewer manuals. Refusals are followed up by an interviewer manager to encourage respondents to participate in the survey. Total non-response is handled by adjusting the weight of households that responded to the survey to compensate for those who do not respond.
The target population of the 2015 CCHS-Nutrition covers the population 1 year of age and over living in private dwellings in the ten provinces. Excluded from the survey's coverage are: persons living on reserves and other Aboriginal settlements, full-time members of the Canadian forces and the institutionalized population. Altogether these exclusions represent less than 10% of the target population.