This paper is in the following e-collection/theme issue:

Diabetes Self-Management (495) Participatory Medicine & E-Patients (787) Health Coaching for Diabetes Patients (150) Diabetes Education and Elearning for Patients (117)

Published on in Vol 8 (2023)

Preprints (earlier versions) of this paper are available at https://preprints.jmir.org/preprint/35682, first published .
The Influence of Age, Sex, and Socioeconomic Status on Glycemic Control Among People With Type 1 and Type 2 Diabetes in Canada: Patient-Led Longitudinal Retrospective Cross-sectional Study With Multiple Time Points of Measurement

The Influence of Age, Sex, and Socioeconomic Status on Glycemic Control Among People With Type 1 and Type 2 Diabetes in Canada: Patient-Led Longitudinal Retrospective Cross-sectional Study With Multiple Time Points of Measurement

The Influence of Age, Sex, and Socioeconomic Status on Glycemic Control Among People With Type 1 and Type 2 Diabetes in Canada: Patient-Led Longitudinal Retrospective Cross-sectional Study With Multiple Time Points of Measurement

Authors of this article:

Seyedmostafa Mousavi1, 2, 3 Author Orcid Image ;   Dana Tannenbaum Greenberg1 Author Orcid Image ;   Ruth Ndjaboué1, 4, 5 Author Orcid Image ;   Michelle Greiver1, 6, 7 Author Orcid Image ;   Olivia Drescher1 Author Orcid Image ;   Selma Chipenda Dansokho1, 8 Author Orcid Image ;   Denis Boutin1, 9 Author Orcid Image ;   Jean-Marc Chouinard1 Author Orcid Image ;   Sylvie Dostie1 Author Orcid Image ;   Robert Fenton1, 10 Author Orcid Image ;   Marley Greenberg1 Author Orcid Image ;   Jonathan McGavock1, 11 Author Orcid Image ;   Adhiyat Najam1 Author Orcid Image ;   Monia Rekik1, 12, 13, 14 Author Orcid Image ;   Tom Weisz1, 15, 16 Author Orcid Image ;   Donald J Willison1, 17 Author Orcid Image ;   Audrey Durand18, 19, 20, 21 Author Orcid Image ;   Holly O Witteman1, 2, 8, 22 Author Orcid Image ;   Diabetes Action Canada Research Questions Prioritization Study23
Article Authors Cited by (3) Tweetations (2) Metrics

    Original Paper

    1Diabetes Action Canada, Toronto, ON, Canada

    2VITAM Research Centre in Sustainable Health, Québec, QC, Canada

    3Université Laval, Québec, QC, Canada

    4Centre de recherche sur le Vieillissement, Sherbrooke, QC, Canada

    5School of Social Work, Faculty of Letters and Human Sciences, Université de Sherbrooke, Sherbrooke, QC, Canada

    6Department of Family and Community Medicine, North York General Hospital, Toronto, ON, Canada

    7Department of Family and Community Medicine, Temerty Faculty of Medicine, University of Toronto, Toronto, ON, Canada

    8Office of Education and Professional Development, Faculty of Medicine, Université Laval, Québec, QC, Canada

    9Centre de recherche du CHUS, Sherbrooke, QC, Canada

    10National Indigenous Diabetes Association, Winnipeg, MB, Canada

    11Pediatrics and Child Health, Rady Faculty of Health Sciences, University of Manitoba, Winnipeg, MB, Canada

    12Department of Operations and Decision Systems, Faculty of Business Administration, Université Laval, Québec, QC, Canada

    13Cardiometabolic Health, Diabetes and Obesity Research Network (CMDO), Sherbrooke, QC, Canada

    14Interuniversity Research Centre of Enterprise Networks (CIRRELT), Montréal, QC, Canada

    15Wounds Canada, North York, ON, Canada

    16Diabetes Canada, Toronto, ON, Canada

    17Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, ON, Canada

    18Canada CIFAR AI Chair, Québec, QC, Canada

    19Institute Intelligence and Data, Université Laval, Québec, QC, Canada

    20Department of Computer Science and Software Engineering, Faculty of Science and Engineering, Université Laval, Québec, QC, Canada

    21Department of Electrical Engineering and Computer Engineering, Faculty of Science and Engineering, Faculty of Science and Engineering, Québec, QC, Canada

    22Department of Family and Emergency Medicine, Faculty of Medicine, Université Laval, Québec, QC, Canada

    23See Acknowledgments

    Corresponding Author:

    Holly O Witteman, PhD

    Department of Family and Emergency Medicine

    Faculty of Medicine

    Université Laval

    Pavillon Ferdinand Vandry

    1050 Ave. de la Médecine

    Québec, QC, G1V 0A6

    Canada

    Phone: 1 418 656 2131 ext 403981

    Email: holly.witteman@fmed.ulaval.ca


    Background: Clinical guidelines for most adults with diabetes recommend maintaining hemoglobin A1c (HbA1c) levels ≤7% (≤53 mmol/mol) to avoid microvascular and macrovascular complications. People with diabetes of different ages, sexes, and socioeconomic statuses may differ in their ease of attaining this goal.

    Objective: As a team of people with diabetes, researchers, and health professionals, we aimed to explore patterns in HbA1c results among people with type 1 or type 2 diabetes in Canada. Our research question was identified by people living with diabetes.

    Methods: In this patient-led retrospective cross-sectional study with multiple time points of measurement, we used generalized estimating equations to analyze the associations of age, sex, and socioeconomic status with 947,543 HbA1c results collected from 2010 to 2019 among 90,770 people living with type 1 or type 2 diabetes in Canada and housed in the Canadian National Diabetes Repository. People living with diabetes reviewed and interpreted the results.

    Results: HbA1c results ≤7.0% represented 30.5% (male people living with type 1 diabetes), 21% (female people living with type 1 diabetes), 55% (male people living with type 2 diabetes) and 59% (female people living with type 2 diabetes) of results in each subcategory. We observed higher HbA1c values during adolescence, and for people living with type 2 diabetes, among people living in lower income areas. Among those with type 1 diabetes, female people tended to have lower HbA1c levels than male people during childbearing years but higher HbA1c levels than male people during menopausal years. Team members living with diabetes confirmed that the patterns we observed reflected their own life courses and suggested that these results be communicated to health professionals and other stakeholders to improve the treatment for people living with diabetes.

    Conclusions: A substantial proportion of people with diabetes in Canada may need additional support to reach or maintain the guideline-recommended glycemic control goals. Blood sugar management goals may be particularly challenging for people going through adolescence or menopause or those living with fewer financial resources. Health professionals should be aware of the challenging nature of glycemic management, and policy makers in Canada should provide more support for people with diabetes to live healthy lives.

    JMIR Diabetes 2023;8:e35682

    doi:10.2196/35682

    Keywords

    adolescent; adult; cohort studies; co-design; diabetes mellitus; diabetes mellitus type 1; diabetes mellitus type 2; glycated hemoglobin; menopause; participatory medicine; patient engagement; postmenopause; premenopause; public and patient involvement; sex factors; socioeconomic disparities in health; user design; patient engagement; public and patient involvement 


    Background

    Diabetes is a chronic condition with 2 common types: type 1 and type 2 [1]. Both types of diabetes are marked by elevated blood glucose levels, but their causes of onset and treatments are generally different [2], and type 1 is less common than type 2 [3]. A standard laboratory test for people with all types of diabetes is the hemoglobin A1c (HbA1c) test. Unlike a single blood glucose measure, HbA1c is a measure of approximate mean blood glucose over 2-3 months [4] that can be measured at any time of day [5] and is often used as a marker of overall glycemic control. Higher HbA1c levels are associated with an increased prevalence of complications of diabetes, affecting the eyes, kidneys, heart, and nerves [1]. Clinical guidelines in Canada recommend maintaining HbA1c levels ≤7% (≤53 mmol/mol) for most adults with diabetes and ≤7.5% (≤59 mmol/mol) for most children with diabetes [6].

    Among people with diabetes, differences in HbA1c levels are associated with sociodemographic characteristics, including age [7-10], socioeconomic status [11,12], and sex [9,13-23]. Specifically, HbA1c levels tend to be higher among adolescents compared with other age groups [7-10] and lower among people with higher socioeconomic status (measured using postal codes) compared with those with lower socioeconomic status [11,12]. Studies comparing HbA1c levels by sex have shown mixed results across countries, sometimes showing higher HbA1c levels among those who are male [9], sometimes showing higher HbA1c levels among those who are female [13-20], and sometimes showing no difference [21-23].

    The relationships between HbA1c levels and individual and social characteristics are not only apparent in the literature [24]; they are also tangible in the lives of people with diabetes. As noted by team members living with diabetes reflecting on their own lives and on comparisons with peers, HbA1c goals may be easier to attain in some situations than others. Such expertise and perspective gleaned from the lived experience of diabetes can add insight and nuances to diabetes research. Including this expertise in health research is a central tenet of patient partnership. Patient-partnered research involves people living with conditions as full members of the research team. In addition to ethical reasons that people living with conditions should be included in research decisions that affect them, such inclusion can also improve the relevance and quality of the studies [25-27].

    Objective

    This study’s research question was developed by people living with type 1 or type 2 diabetes who were involved in a larger project in which people living with diabetes developed and prioritized research questions about diabetes. This study aimed to answer one of the identified and prioritized questions: Accounting for socioeconomic status, what patterns exist in HbA1c results among people in Canada with type 1 or type 2 diabetes of different sexes at different ages? Although some data already exist regarding HbA1c results in Canada [24,28], there has not yet been a national analysis like this stratified by the type of diabetes and including variables of sex, age, and socioeconomic status. Furthermore, few or no studies have included people living with diabetes as members of the research team.


    Study Design

    This study used a longitudinal retrospective cross-sectional study design with multiple time points of measurement.

    Study Data and Inclusion Criteria

    To answer our research question, we used data from the Canadian National Diabetes Repository. As of July 1, 2021, this repository consisted of electronic medical records of 123,543 people with type 1 and type 2 diabetes living in Canada. The repository continues to grow, and as of April 23, 2022, it included records of 147,459 such people. Records describe patients treated in family medicine practices in 5 Canadian provinces: Ontario (57% of HbA1c results), Alberta (25% of HbA1c results), Manitoba (15% of HbA1c results), Quebec (3% of HbA1c results), and Newfoundland and Labrador (0.4% of HbA1c results) [29,30]. These data originate from a larger data repository, the Canadian Primary Care Sentinel Surveillance Network. Previous analyses of the representativeness of the Canadian Primary Care Sentinel Surveillance Network have noted that, compared with the Canadian population, the data describe more people who are older and more people who are female, which may reflect patterns of greater health care seeking among these groups. It is therefore recommended to include age and sex in analytical models when analyzing these data [31].

    In the National Diabetes Repository, people are identified as having diabetes if their record includes an HbA1c result of ≥6.5% or at least 2 fasting blood glucose results >7 mmol/L recorded on different dates ≤1 year apart. However, medical records in Canada do not currently specify the type of diabetes. Therefore, we distinguished between people with type 1 and type 2 diabetes using an algorithm recently developed and validated using Canadian data. The machine learning algorithm analyzed 21 variables (eg, insulin use, nonmetformin antihyperglycemic agent use, and insulin pump use) and demonstrated a sensitivity (ie, ability to correctly identify that someone with type 1 diabetes has type 1 diabetes) of 80.6% and a specificity (ie, ability to correctly identify that someone without type 1 diabetes does not have type 1 diabetes) of 99.8% [32].

    For this study, we included all HbA1c results from the Canadian National Diabetes Repository from individuals with diabetes who had at least one HbA1c measurement between 2010 and 2019 and for whom our other variables of interest (age, sex, and socioeconomic status) were available. In other words, we excluded records that lacked one or more of these data elements. We also excluded records with HbA1c results <3.5% (15 mmol/mol) or >20% (195 mmol/mol), as clinical experts on our team deemed these to be likely laboratory errors, data entry errors, or rare outliers. We derived individuals’ ages by subtracting each person’s date of birth from the date when each HbA1c measurement was performed. We excluded data from individuals aged <10 years, as they contributed such a small proportion of the data in the repository (0.04%) that we would be unable to conduct robust analyses for this subpopulation. As a proxy measure of socioeconomic status, the Canadian National Diabetes Repository applies an established index developed by the Canadian Institute for Health Information [33] to derive neighborhood before-tax income quintiles using the first 3 digits of individuals’ most recent residential postal codes. Quintile 1 denotes people living in lower-income neighborhoods, quintiles 2 to 4 denote middle-income neighborhoods, and quintile 5 is assigned to the highest-income neighborhoods [34].

    Statistical Analysis

    Our unit of analysis was each HbA1c result. We graphically inspected HbA1c results across age groups using locally estimated scatterplot smoothing curves for type 1 diabetes. We used generalized additive model curves for type 2 diabetes to account for the large amount of data in this category. We then analyzed HbA1c results using generalized estimating equations to account for dependency structure at the between-subject and within-subject levels. As the correlations between HbA1c measurements in different years were similar, we used an exchangeable correlation structure to account for correlations between HbA1c measurements of each individual [35]. As the distribution of HbA1c levels was highly skewed, we used log transformation to reduce the skewness and stabilize the variance [36]. Owing to nonlinearity, we categorized age, dividing it into ordinal categories at 10- or 20-year intervals. The age range of 10 to 19 years included puberty and menarche for a large proportion of people who experience puberty and menarche [37,38]. The age range of 20 to 39 years included childbearing years for a large proportion of people who experience pregnancy [39-41]. The age range of 40 to 59 years included perimenopause and menopause for a large proportion of people who experience perimenopause and menopause [42-45]. The age range of 60 to 79 years included postmenopause for a large proportion of people who experience postmenopause [46]. The age range of ≥80 years included advanced adulthood.

    We performed 2-way ANOVA to determine the relationship between HbA1c levels and independent variables. We then analyzed potential interactions between age and sex for type 1 and type 2 diabetes, reporting pairwise interaction contrasts. We set our threshold for statistical significance at P<.05. We conducted 2 sets of sensitivity analyses. First, to explore whether practice and guideline changes over time might have influenced HbA1c results and our findings, we conducted sensitivity analyses by rerunning all plots and models on 3 subsets of data: 2010 to 2012, 2013 to 2016, and 2017 to 2019. Second, we wished to check for potential biases in our results associated with unstable blood glucose levels shortly after diabetes diagnosis. In other words, if diagnostic HbA1c values were present in the data set, higher HbA1c values may not have anything to do with diabetes care or self-management but rather, may simply be reflections of a new condition. We did not have dates of diagnoses in the data set; therefore, to explore this issue, we removed the earliest HbA1c result from the data set for each person and reran our analytical models. The details of these sensitivity analyses are provided in Multimedia Appendix 1.

    We performed all analyses using R 3.6.1 (R Foundation for Statistical Computing) and associated packages, including ggplot2 for graphics, GEE for generalized estimating equations, and estimated marginal means for contrasts [47-57].

    Interpreting Results With Team Members With Diabetes

    Following our analyses, we organized meetings with members of the research team living with type 1 and type 2 diabetes. These team members had already been involved throughout the yearlong larger project, including a series of early meetings to orient all team members to epidemiological cohort studies and subcommittee meetings led and attended only by patient partners to generate research questions.

    To present and discuss the results of this study, our bilingual team held 1 meeting in English and 1 meeting in French. We invited all team members living with diabetes to attend, presented the results for approximately 15 minutes, and then held an open discussion for the remaining 45 minutes. We recorded the meetings to ensure that we accurately noted team members’ comments, and we invited all team members to review the manuscript to ensure that we conveyed meaning correctly. We invited all team members to fulfill the authorship criteria as defined by the International Committee of Medical Journal Editors and, accordingly, be coauthors of the manuscript. A total of 9 patient partner team members accepted the invitation to serve as coauthors of the manuscript.

    Because team members were identified as such; that is, they were members of the research team, not study participants, neither our research team nor the research ethics committee that approved the study wished to treat these team members any differently than the team members who were professors, scientists, health professionals, students, and research staff. For this reason, we did not treat discussions among patient partners as research data with study participants, but rather, used the same approaches that one might use with any group of research team members. Specifically, we recorded meetings to ensure accurate note-taking, wrote summaries of discussions, and invited meeting attendees to review and comment on the summaries as we collaboratively drafted this manuscript. Although scientific papers do not typically report the identity characteristics of authors and it is rare to read concerns about whether a group of scientist authors can adequately represent the perspectives of the scientific community as a whole, we will also note that Diabetes Action Canada’s methods for recruitment of patient partners explicitly focuses on making extra efforts to ensure broad representation [27]. For this project, as in other projects [58], we deliberately constructed a team of people across all project roles who had a wide variety of backgrounds, with a particular focus on ensuring diversity in type of diabetes and in ethnocultural and socioeconomic backgrounds.

    Ethics Approval

    Our study received ethics approval from the Laval University Research Ethics Committee, 2020-373/20-01-2021. The original data collection was also conducted with ethics approval. This study was approved by the National Diabetes Repository governance committee. The governance committee comprises a minimum of 50% of people living with diabetes.


    Population

    The resulting data set consisted of 2 groups of data: one for people identified by the algorithm as having type 1 diabetes and one for people identified by the algorithm as having type 2 diabetes. Table 1 summarizes the HbA1c results from people with type 1 and type 2 diabetes according to age, sex, and socioeconomic status. Because HbA1c continues to be reported as a percentage in Canada, we use percentages to facilitate understanding by people living with diabetes in Canada. Of the 1950 and 946,931 available records, we excluded 1 record from the data available for people with type 1 diabetes and 1337 records from the data available for people with type 2 diabetes owing to HbA1c values <3.5% (15 mmol/mol) or >20% (195 mmol/mol). Among people living with type 1 diabetes, individuals contributed data for a median of 3 (Q1-Q3 2-6) years. Most (244/296, 82.4%) participants contributed data to only 1 of the age groups used in our analyses. The others (52/296, 17.6%) contributed data to 2 age groups. Among people living with type 2 diabetes, individuals contributed data for a median of 5 (Q1-Q3 3-7) years. Most (68,437/90,417, 75.7%) participants contributed data to only 1 of the age groups used in our analyses. The others (21,980/90,417, 24.3%) contributed data to 2 age groups.

    Mean HbA1c was 8.3% (SD 1.7%) for female people with type 1 diabetes, 8.0% (SD 1.7%) for male people with type 1 diabetes, 7.1% (SD 1.3%) for female people with type 2 diabetes, and 7.2% (SD 1.3%) for male people with type 2 diabetes. In both types of diabetes, people who are female more often lived in geographic areas with lower socioeconomic status compared with people who are male.

    Table 1. Characteristics of hemoglobin A1c resultsa among people with type 1 and type 2 diabetes.
    VariableType 1 diabetes (1949 HbA1cb results from 296 people)Type 2 diabetes (945,262 HbA1c results from 90,417 people)

    		Male (n=155)Female (n=141)Male (n=47,286)Female (n=43,131)
    Population

    		People with HbA1c results in 2 age groups, n (%)22 (14.2)30 (21.3)11,630 (24.6)10,350 (24)
    HbA1c resultsa

    		Mean (SD), %8 (1.7)8.3 (1.7)7.2 (1.3)7.1 (1.3)

    		Mean (SD), mmol/mol64 (16)67 (16)55 (12)54 (12)

    		Median (Q1-Q3c; ranged), %7.6 (6.9-8.5; 4.8-16.3)8 (7.2-9.1; 5-16)6.9 (6.3-7.8; 3.5-19.6)6.8 (6.3-7.6; 3.5-19.5)

    		Median (Q1-Q3c; ranged), mmol/mol60 (52-69; 29-155)64 (55-76; 31-151)52 (45-62; 15-191)51 (45-60; 15-190)

    		Results per person, median (Q1-Q3c; ranged)4 (2-9; 1-42)5 (2-9; 1-48)8 (4-15; 1-112)8 (4-15; 1-115)

    		HbA1c results ≤7%, %30.5215559

    		Total, n (%)970 (49.7)979 (50.2)505,364 (53.4)439,898 (46.5)
    Age (years), n (%)a

    		10-1944 (4.5)37 (3.7)1020 (0.2)1097 (0.2)

    		20-39454 (46.8)559 (57)13,744 (2.7)18,675 (4.2)

    		40-59305 (31.4)277 (28.2)145,393 (28.7)123,621 (28.1)

    		60-79167 (17.2)106 (10.8)287,045 (56.7)235,044 (53.4)

    		≥800 (0)0 (0)58,162 (11.5)61,461 (13.9)
    Socioeconomic status, n (%)a

    		1 (lowest income)199 (20.5)336 (34.3)115,565 (22.8)116,715 (26.5)

    		2141 (14.5)151 (15.4)108,374 (21.4)98,284 (22.3)

    		3223 (22.9)196 (20)94,672 (18.7)80,026 (18.1)

    		4222 (22.8)182 (18.5)93,043 (18.4)74,537 (16.9)

    		5 (highest income)185 (19)114 (11.6)93,710 (18.5)70,336 (15.9)

    aAs noted in the Methods section, our unit of analysis is each HbA1c result. These summary statistics are therefore calculated across HbA1c results in the data set, meaning that each HbA1c result within a given category contributes 1 data point.

    bHbA1c: hemoglobin A1c.

    cQ1-Q3: quartile 1-quartile 3.

    dRange=minimum value-maximum value.

    As shown in Figure 1, we observed a relationship between age and HbA1c levels among people with type 1 or type 2 diabetes. We also observed that these relationships may differ between people who are male and people who are female. As shown in Figure 2, socioeconomic status may also be associated with HbA1c levels during much of adulthood, with somewhat higher HbA1c values among adults aged 30 or 40 years through 70 years and living in geographic areas with lower mean income.

    Table 2 shows the results of ANOVA fitted by generalized estimating equations for both types of diabetes.

    According to our analyses of 296 people in the Canadian National Diabetes Repository with type 1 diabetes, there was a statistically significant relationship between age and HbA1c levels, with overall lower HbA1c levels among older people with type 1 diabetes. We observed no statistically significant relationships between sex and HbA1c, socioeconomic status and HbA1c, and the interaction term showed no statistically significant differences between people who are male and female at different ages. For 90,417 people in the Canadian National Diabetes Repository with type 2 diabetes, all variables demonstrated statistically significant relationships with HbA1c levels. HbA1c results are lower among people who are older, higher among people who are male, and higher among people living in geographic areas with lower mean income. There was a significant interaction between age and sex, suggesting that the pattern of lower HbA1c values among older people with type 2 diabetes was somewhat different between those who are male and female. Specifically, those who are male appear to reach lower HbA1c values later in their life span compared with those who are female. Multimedia Appendix 1 shows pairwise interaction contrasts for this interaction.

    Sensitivity analyses (Multimedia Appendix 1) demonstrated that the results were similar for data from 2010 to 2012, 2013 to 2016, and 2017 to 2019, suggesting that practice and guideline changes during the data collection period did not substantially influence our findings. Sensitivity analyses also suggested no changes in our findings when the first HbA1c value for each person was removed from the data set, suggesting that patterns observed are not a reflection of new diagnoses.

    ‎
    Figure 1. Mean HbA1c levels by sex (male, female) across all ages of people with type 1 and type 2 diabetes in Canada. gam: generalized additive model; HbA1c: hemoglobin A1c; loess: locally estimated scatterplot smoothing.
    ‎
    Figure 2. Mean HbA1c levels of people with type 1 and type 2 diabetes in Canada across all ages by socioeconomic status. gam: generalized additive model; HbA1c: hemoglobin A1c; loess: locally estimated scatterplot smoothing.
    Table 2. Relationship between age, sex, socioeconomic status, and hemoglobin A1c (HbA1c) among people with type 1 and type 2 diabetes in Canada.

    		Type 1 diabetes (1949 HbA1c results from 296 people)Type 2 diabetes (945,262 HbA1c results from 90,417 people)

    		F statisticP valueF statisticP value
    Age3.74.01635.70<.001
    Sex2.26.13186.85<.001
    Socioeconomic status0.77.54218.78<.001
    Age:sex interaction1.48.21113.165<.001

    Interpretation of Results by Team Members With Diabetes

    Approximately equal numbers of team members living with type 1 and type 2 diabetes attended the meetings to discuss and interpret results. In these meetings, people living with diabetes raised potential post hoc explanations for the findings, asked technical questions about the analyses, raised potential study limitations, and discussed implications for policy and future research.

    Specifically, with respect to explanations of findings, women living with type 1 diabetes suggested that the potential pattern among female people with type 1 diabetes mimicked their own life courses and may reflect lower HbA1c levels during potential childbearing years and higher HbA1c levels during potential menopausal years. People with diabetes also noted the differing sex-based differences between type 1 diabetes and type 2 diabetes. In both types of diabetes, HbA1c results for people who are female dipped around the age of 30 years and peaked just before the age of 50 years, whereas HbA1c values for people who are male decreased more smoothly with age. However, in the case of type 1 diabetes, curves met and intersected at multiple ages, whereas there were no similar meetings and intersections in the case of type 2 diabetes. Finally, people with diabetes questioned whether differences in the HbA1c levels at different ages might reflect differences in the effort that people are able to put into diabetes management at different stages of their life course, depending on when they were diagnosed.

    With respect to technical questions, people living with diabetes noted that accurate HbA1c measurement was not possible for some individuals [59], queried how hypoglycemia unawareness might influence HbA1c results, and raised the issue that HbA1c is a highly imperfect measure. HbA1c is essentially analogous to average blood glucose, and averages can mask substantial variation. Nonetheless, it remains a standard measure, as other methods of measurement (eg, time-in-range measured by flash or continuous glucose monitoring) are not universally available across Canada.

    People living with diabetes also raised study limitations including the relatively small amount of data available for people with type 1 diabetes in these primary care electronic medical records, the need to use an algorithm to estimate diabetes type because of the lack of specificity about this in the Canadian electronic medical records, the lack of data on education level to better identify the contribution of income and education to glycemic control, and the lack of data on race and ethnicity, which strongly impact outcomes for people with diabetes in Canada and elsewhere. Team members with diabetes also questioned whether HbA1c results might have differed during the 10-year span of the study given the introduction of new technologies in Canada between 2010 and 2019 that allow greater glycemic control.

    With respect to implications for policy and future research, people living with diabetes noted that an HbA1c level of ≤7% appeared to be very difficult to reach and maintain for many people living with diabetes in Canada. They suggested that these results be communicated to health professionals to help set realistic expectations and that people living with diabetes should consider being “insistent” with their health professionals to explore options for treatment and appropriate goals. Although guidelines suggest that HbA1c targets should be set between a health professional and an individual living with diabetes while taking into account all relevant aspects of the individual’s life, in practice, many people with diabetes do not receive this level of individualized care [24].

    Indigenous patient partners expressed interest in data specifically for Indigenous peoples in Canada. The National Diabetes Repository does not currently contain data from health centers specifically serving Indigenous communities, although there may be data from urban Indigenous people within the data set that cannot be identified separately from the larger data set. Our goal with this project was to establish a means for people living with diabetes to determine research questions and drive epidemiological cohort studies. Although we knew our data source would not allow us to answer research questions specific to Indigenous peoples, we specifically included Indigenous patient partners, researchers, and non-Indigenous researchers who work with Indigenous communities in the project to help us collectively ensure that our approaches would not harm potential future Indigenous-led projects conducted under relevant ethical frameworks such as the First Nations’ Principles of Ownership, Control, Access, and Possession [60].


    Principal Findings

    In this study co-designed with people living with diabetes, we aimed to explore differences in HbA1c levels between people with type 1 or type 2 diabetes of different sexes at different ages and with different levels of socioeconomic status, using a large database of primary care electronic medical records for people with diabetes in Canada. We report 6 main observations from our study.

    First, the differences in statistical significance between the smaller sample of people with type 1 diabetes and the much larger sample of people with type 2 diabetes are reflective of broader patterns in research. These patterns have policy implications that can affect the lives of people living with more or less common conditions, including more or less common types of diabetes. Larger sample sizes allow identification of smaller associations or effects [61]. In the case of diabetes, this means that it is easier to identify statistically significant effects in the much larger populations of people with type 2 diabetes than in the smaller populations of people with type 1 diabetes or the even smaller populations of people with other types. This can have negative policy impacts on people with diabetes, for different reasons. For people with type 2 diabetes, as health research enters the era of big data and personalized medicine, large data sets analyzed by research teams with little clinical, epidemiological, or personal expertise may allow identification of associations or effects that may not be clinically or personally meaningful. For people with type 1 diabetes, who represent an estimated 5% to 10% of cases of diabetes, minority status within the larger disease community has led to policy issues such as type 2 diabetes being identified as a risk factor for severe COVID-19 outcomes, whereas type 1 diabetes, which demonstrated higher odds ratios or hazard ratios for severe COVID-19 outcomes in multiple studies but had wider CIs owing to smaller populations [62-64], was identified only as a “potential risk factor” [65]. As with other less common conditions, it is important that policy decisions that affect people with diabetes account for differences in type of diabetes and avoid applying the same statistical standards to differently sized populations without accounting for the influence of sample size.

    Second, age is an important consideration for HbA1c targets for people with both types of diabetes. Similar to studies in other countries [7-10,66,67], our study demonstrated overall higher HbA1c values among adolescents with diabetes compared with people with diabetes in other age groups. This may reflect the ways in which adolescents differ from people in other age groups due to biology (eg, puberty, menarche), diabetes management (eg, time since diagnosis to develop useful habits and patterns, lack of full control over management options due to family preferences and finances), and life stage (eg, externally imposed structures of school, work, family; internally-directed focus on social development). Although providing high-quality health care to children and adolescents with diabetes has long been an area of focus in Canada [68], our results suggested that adolescents may still need more support both within and outside of clinical encounters.

    Third, there is a tendency toward different patterns across the life course between people of different sexes with type 1 diabetes. Our relatively small sample of people with type 1 diabetes in this data set did not allow us to draw definitive conclusions. However, women living with type 1 diabetes in our team noted that the potential pattern we observed mimicked their own HbA1c patterns during childbearing, perimenopausal, and menopausal years. Hermann et al [69] similarly observed significantly higher HbA1c among female people with type 1 diabetes compared with male people with type 1 diabetes before the age of 30 years and after the age of 50 years. People with type 1 diabetes who plan to bear children may be particularly motivated to maintain a lower HbA1c level during their childbearing years because of more stringent recommendations regarding glycemic control during pregnancy [70,71] and societal, medical, and self-directed expectations regarding how pregnant people, especially those at increased risk, should prioritize the health of their offspring [72-75].

    Following childbearing, the hormonal shifts of perimenopause and menopause combined with common life stressors of middle age and gendered parenting roles may explain the somewhat higher HbA1c values among many middle-aged female people. As noted by the lead patient partner in this project (DG), the evidence available about menopause and type 1 diabetes is scarce. There are a small number of studies addressing age of menopause among people with type 1 diabetes [76-79] and associated health risks [80]. However, there is little evidence about how to manage one’s diabetes and other health concerns post menopause [81]. This evidence gap negatively impacts the lives of people with type 1 diabetes who progress through menopause.

    Fourth, the direction of the overall sex-based differences we observed among people with type 2 diabetes differed from some previous studies in other countries. In our study using data from people in Canada, male people with type 2 diabetes had overall higher HbA1c values, indicating a potentially higher risk of diabetes-related complications compared with female people. Other studies using data from people in Portugal [13], Brazil and Venezuela [15], Korea [19], and Spain [82] reported the opposite, with overall higher HbA1c values among those who are female compared with those who are male. Studies in the United States [22] and the Netherlands [23] reported no sex-based differences in HbA1c values, and a study in Sweden reported higher HbA1c values in male people compared with female people [9]. All of these other studies either focused on type 2 diabetes or did not distinguish between types of diabetes, meaning that the data were necessarily drawn from people with type 2 diabetes, who constitute 95% of people living with diabetes globally according to the World Health Organization [1]. The lack of agreement among different studies regarding sex-based differences might occur because overall differences may be a product of both biology and gender equality as seen via social roles. The World Economic Forum’s Global Gender Gap Report offers data in support of the suggestion that gender roles may explain the different results regarding sex-based differences across countries. As measured by this index, the countries in which male people have lower HbA1c values than female people (Spain, Korea, Portugal, Brazil, Venezuela) have lower mean gender equality (mean 0.723, SD 0.043) than the countries in which there was no difference (the United States and the Netherlands: mean 0.763, SD 0.001), which in turn had lower mean gender equality than countries in which male people had higher HbA1c values (Canada, Sweden: mean 0.798, SD 0.036) [83]. In other words, in countries with better overall gender equality, female people may have better diabetes-relevant health outcomes relative to male people, while in countries with worse gender equality, female people with type 2 diabetes may have worse diabetes-relevant health outcomes relative to male people.

    Fifth, people with type 2 diabetes living in less affluent areas in Canada had higher HbA1c levels than those living in more affluent areas. This is unfortunately unsurprising, as type 2 diabetes is a progressive disease and is more prevalent in Canada among people with lower incomes than among those with higher incomes [84-86]. There may also be a similar pattern among people with type 1 diabetes that is not detectable in our relatively small sample. People with type 1 diabetes with higher incomes have less variability in their HbA1c results [12]. Lower income has been shown to be associated with higher HbA1c among children with type 1 diabetes in Canada [87], and complications are more prevalent among people living with type 1 diabetes in Canada with lower incomes [88]. Owing to the uneven coverage of diabetes medications and devices across Canada, unequal access to high-quality health care, and large differences in levels of food security, people with both types of diabetes who live on lower incomes may face additional challenges in diabetes management compared with those with higher incomes [89-92]. Without efforts to address these inequities, such patterns may worsen in the coming years with the advent of new technologies and medications.

    Finally, a substantial proportion of people with diabetes in Canada have not yet demonstrated guideline-recommended HbA1c values. This shows the difficulty in reaching and maintaining this goal [93]. Similarly, Aronson and colleagues [94] showed, within a larger sample of 3600 adults living with type 1 diabetes and receiving care from endocrinologists in Canada, that less than a quarter of people demonstrated HbA1c values ≤7%. Health professionals and policy makers should be aware of this gap to better support people living with diabetes in Canada. As noted by the people on our team who live with diabetes, health professionals’ acknowledgment of the difficulty of attaining this target would help them feel less like they are failing and more like they are part of a large group living with a difficult condition and potentially struggling to achieve targets set by academic researchers and health professionals. Policy makers can improve this situation by better funding health care, education, medications (eg, insulin and other medications), supplies (eg, test strips, flash or continuous glucose monitors, and closed-loop artificial pancreas systems) [95], food security initiatives (eg, access to affordable healthy foods) [96], healthy environment initiatives (eg, walking trails, bicycle paths, and community gardens) [97], broader antipoverty initiatives [98], and research aimed at supporting people living with diabetes in Canada to achieve self-directed health goals [99].

    This study had 3 main limitations. First, although the data set was large and of overall high quality, the lack of relevant data (eg, electronic medical records in Canada do not include relevant data such as ethnicity), the use of proxy variables (eg, socioeconomic status as an unvarying quintile derived from the most recent postal codes), and the necessity of using an algorithm to predict the type of diabetes may have limited our results. We cannot be certain that all people predicted by the algorithm to have type 1 diabetes were correctly identified, and even then, the small number of people identified by the algorithm as having type 1 diabetes meant that our findings with respect to type 1 diabetes were less conclusive than those for type 2 diabetes. Other rarer types of diabetes such as latent autoimmune diabetes in adults were not represented at all. Second, we did not include comorbid medical conditions in this preliminary study, as these were not part of the research question identified by people living with diabetes. Although people who live with other conditions in addition to diabetes may have higher or lower HbA1c values than those who live only with diabetes; for this preliminary analysis, we sought only to determine broad patterns in this large national data set. Third, our analyses had some threats to external validity (ie, generalizability). All medical records from the National Diabetes Repository were collected from the primary care records of 5 provinces in Canada. Approximately 15% of people in Canada lack access to a primary health care provider, and lack of access is not distributed evenly [100,101], meaning that our study may have had some selection bias.

    This study had 3 main strengths. First, the entire study, from the initial idea and development of the research question to the interpretation of results and drafting of this manuscript, was conducted in full partnership with people living with the condition studied. This allowed us to identify a research question that is relevant to people living with diabetes in Canada, enrich our interpretation of results, and avoid framing our results in ways that are heedless of the humanity of people whose medical data were analyzed. Second, this study offered insight into glycemic control for people living with diabetes across Canada, included key variables of sex and age, and accounted for the potential influence of socioeconomic status. It is important to avoid a one-size-fits-all approach when discussing diabetes management. Identifying patterns according to common variables is a step toward more individualized care. Third, data from the Canadian National Diabetes Repository represent high-quality big data across multiple provinces, allowing national-level analyses that were previously difficult to perform in Canada.

    Conclusions

    This study demonstrated the value and potential of patient-led research and of a national data repository for diabetes. People who live with a condition should have the power to set health research agendas so that research serves their needs. Responding to a research question developed by people living with diabetes, we mapped nearly a million HbA1c results over 10 years from people with type 1 and type 2 diabetes in Canada and showed how HbA1c results may differ by age, sex, and socioeconomic status. These factors are important to consider when setting HbA1c targets and when studying the relationship between HbA1c levels and complications of diabetes. Further research and support are needed to help people manage diabetes across life stages, with notable challenges during adolescence for people of both sexes and during menopause for people of female sex. As noted by members of our research team living with diabetes, people living with diabetes may not consistently receive adequate support from their health care team, family, employer, and regulatory and funding systems that determine the availability of medications and technologies. Health professionals should be aware of the difficulty in maintaining an HbA1c value below the guideline-recommended targets without access to additional support, medications, and technologies. Policy makers should set policies that enable people with diabetes in Canada to live healthy lives.

    Acknowledgments

    The authors gratefully acknowledge the contributions of all team members who contributed greatly to this research project as a whole but did not accept the invitation to participate in this particular paper as coauthors, Conrad Pow and Tao Chen for their assistance with the Canadian National Diabetes Repository, and Anne-Sophie Julien for her advice on statistical analyses, and for conducting data analysis and interpretation. This study was funded by the Canadian Institutes of Health Research grants 148426 and 169416. The Canadian Institutes of Health Research had no role in determining the study design, the plans for data collection or analysis, the decision to publish, or the preparation of this manuscript. HOW was funded by a Tier 2 Canada Research Chair in Human-Centered Digital Health.

    The members of the group author are as follows: Lahssen Abbassi, Javed Alloo, André Amyot, Elaine Brière, Stéphane Brière, Sasha Delorme, Lanie Deslauriers, Sophie Desroches, Maman Joyce Dogba, Noah M. Ivers, Tania Leclerc, Alex M. McComber, Danièle Remy, Larry Spence, Nadia Tabiou, Frank Tang, Joshua Tepper, Marie-Claude Tremblay, David Wells, and Catherine Yu.

    Data Availability

    National Diabetes Repository data are available for analysis [102].

    Authors' Contributions

    SMM and HOW contributed to the study design. SMM, MG, and HOW contributed to data collection. SMM and HOW conducted data analysis and interpretation. SMM and HOW drafted the first version of the manuscript. SMM, DG, RN, MG, OD, SCD, DB, JMC, SD, RF, MG, JMG, AN, MR, TW, DJW, AD, and HOW critically revised the manuscript and approved the final version for publication. SMM had full access to all the data in the study. SMM and HOW had the final responsibility for the decision to submit for publication.

    Conflicts of Interest

    None declared.

    ‎
    Multimedia Appendix 1

    Pairwise contrasts and sensitivity analyses.

    PDF File (Adobe PDF File), 960 KB
    1. Diabetes. World Health Organization.   URL: https://www.who.int/health-topics/diabetes [accessed 2020-06-09]
    2. Xu G, Liu B, Sun Y, Du Y, Snetselaar LG, Hu FB, et al. Prevalence of diagnosed type 1 and type 2 diabetes among US adults in 2016 and 2017: population based study. BMJ 2018 Sep 04;362:k1497 [FREE Full text] [CrossRef] [Medline]
    3. Carstensen B, Rønn PF, Jørgensen ME. Prevalence, incidence and mortality of type 1 and type 2 diabetes in Denmark 1996-2016. BMJ Open Diabetes Res Care 2020 May;8(1):e001071 [FREE Full text] [CrossRef] [Medline]
    4. Welsh KJ, Kirkman MS, Sacks DB. Role of glycated proteins in the diagnosis and management of diabetes: research gaps and future directions. Diabetes Care 2016 Aug;39(8):1299-1306 [FREE Full text] [CrossRef] [Medline]
    5. Silverman RA, Thakker U, Ellman T, Wong I, Smith K, Ito K, et al. Hemoglobin A1c as a screen for previously undiagnosed prediabetes and diabetes in an acute-care setting. Diabetes Care 2011 Sep;34(9):1908-1912 [FREE Full text] [CrossRef] [Medline]
    6. Burson R, Moran KJ. Individualizing targets for glycemic control. Home Healthc Now 2018;36(3):190-191. [CrossRef] [Medline]
    7. Pinhas-Hamiel O, Hamiel U, Boyko V, Graph-Barel C, Reichman B, Lerner-Geva L. Trajectories of HbA1c levels in children and youth with type 1 diabetes. PLoS One 2014;9(10):e109109 [FREE Full text] [CrossRef] [Medline]
    8. Clements MA, Foster NC, Maahs DM, Schatz DA, Olson BA, Tsalikian E, T1D Exchange Clinic Network. Hemoglobin A1c (HbA1c) changes over time among adolescent and young adult participants in the T1D exchange clinic registry. Pediatr Diabetes 2016 Aug;17(5):327-336. [CrossRef] [Medline]
    9. Julin B, Willers C, Leksell J, Lindgren P, Looström Muth K, Svensson A, et al. Association between sociodemographic determinants and health outcomes in individuals with type 2 diabetes in Sweden. Diabetes Metab Res Rev 2018 May;34(4):e2984. [CrossRef] [Medline]
    10. Hessler DM, Fisher L, Mullan JT, Glasgow RE, Masharani U. Patient age: a neglected factor when considering disease management in adults with type 2 diabetes. Patient Educ Couns 2011 Nov;85(2):154-159 [FREE Full text] [CrossRef] [Medline]
    11. Whyte MB, Hinton W, McGovern A, van Vlymen J, Ferreira F, Calderara S, et al. Disparities in glycaemic control, monitoring, and treatment of type 2 diabetes in England: a retrospective cohort analysis. PLoS Med 2019 Oct;16(10):e1002942 [FREE Full text] [CrossRef] [Medline]
    12. Auzanneau M, Lanzinger S, Bohn B, Kroschwald P, Kuhnle-Krahl U, Holterhus PM, et al. Area deprivation and regional disparities in treatment and outcome quality of 29,284 pediatric patients with type 1 diabetes in germany: a cross-sectional multicenter DPV analysis. Diabetes Care 2018 Dec;41(12):2517-2525. [CrossRef] [Medline]
    13. Góis C, Duarte TA, Paulino S, Raposo JF, do Carmo I, Barbosa A. Depressive symptoms are associated with poor glycemic control among women with type 2 diabetes mellitus. BMC Res Notes 2018 Jan 16;11(1):38 [FREE Full text] [CrossRef] [Medline]
    14. Willers C, Iderberg H, Axelsen M, Dahlström T, Julin B, Leksell J, et al. Sociodemographic determinants and health outcome variation in individuals with type 1 diabetes mellitus: a register-based study. PLoS One 2018;13(6):e0199170 [FREE Full text] [CrossRef] [Medline]
    15. G Duarte F, da Silva Moreira S, Almeida MD, de Souza Teles CA, Andrade CS, Reingold AL, et al. Sex differences and correlates of poor glycaemic control in type 2 diabetes: a cross-sectional study in Brazil and Venezuela. BMJ Open 2019 Mar 05;9(3):e023401 [FREE Full text] [CrossRef] [Medline]
    16. Larkin ME, Backlund J, Cleary P, Bayless M, Schaefer B, Canady J, Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Research Group. Disparity in management of diabetes and coronary heart disease risk factors by sex in DCCT/EDIC. Diabet Med 2010 Apr;27(4):451-458. [CrossRef] [Medline]
    17. Mair C, Wulaningsih W, Jeyam A, McGurnaghan S, Blackbourn L, Kennon B, Scottish Diabetes Research Network (SDRN) Epidemiology Group. Glycaemic control trends in people with type 1 diabetes in Scotland 2004-2016. Diabetologia 2019 Aug;62(8):1375-1384 [FREE Full text] [CrossRef] [Medline]
    18. Kautzky-Willer A, Kosi L, Lin J, Mihaljevic R. Gender-based differences in glycaemic control and hypoglycaemia prevalence in patients with type 2 diabetes: results from patient-level pooled data of six randomized controlled trials. Diabetes Obes Metab 2015 Jun;17(6):533-540 [FREE Full text] [CrossRef] [Medline]
    19. Choe S, Kim JY, Ro YS, Cho S. Women are less likely than men to achieve optimal glycemic control after 1 year of treatment: a multi-level analysis of a Korean primary care cohort. PLoS One 2018;13(5):e0196719 [FREE Full text] [CrossRef] [Medline]
    20. Clemens KK, Woodward M, Neal B, Zinman B. Sex disparities in cardiovascular outcome trials of populations with diabetes: a systematic review and meta-analysis. Diabetes Care 2020 May;43(5):1157-1163. [CrossRef] [Medline]
    21. Shah VN, Wu M, Polsky S, Snell-Bergeon JK, Sherr JL, Cengiz E, for T1D Exchange Clinic Registry. Gender differences in diabetes self-care in adults with type 1 diabetes: findings from the T1D Exchange clinic registry. J Diabetes Complications 2018 Oct;32(10):961-965. [CrossRef] [Medline]
    22. Misra R, Lager J. Ethnic and gender differences in psychosocial factors, glycemic control, and quality of life among adult type 2 diabetic patients. J Diabetes Complications 2009;23(1):54-64. [CrossRef] [Medline]
    23. de Jong M, Vos RC, de Ritter R, van der Kallen CJ, Sep SJ, Woodward M, et al. Sex differences in cardiovascular risk management for people with diabetes in primary care: a cross-sectional study. BJGP Open 2019 Jul;3(2):bjgpopen19X101645 [FREE Full text] [CrossRef] [Medline]
    24. Coons MJ, Greiver M, Aliarzadeh B, Meaney C, Moineddin R, Williamson T, et al. Is glycemia control in Canadians with diabetes individualized? A cross-sectional observational study. BMJ Open Diabetes Res Care 2017;5(1):e000316 [FREE Full text] [CrossRef] [Medline]
    25. Strategy for patient-oriented research. Canadian Institutes of Health Research.   URL: https://cihr-irsc.gc.ca/e/41204.html [accessed 2022-12-19]
    26. Witteman HO, Chipenda Dansokho S, Colquhoun H, Fagerlin A, Giguere AM, Glouberman S, et al. Twelve lessons learned for effective research partnerships between patients, caregivers, clinicians, academic researchers, and other stakeholders. J Gen Intern Med 2018 Apr;33(4):558-562 [FREE Full text] [CrossRef] [Medline]
    27. Dogba MJ, Dipankui MT, Chipenda Dansokho S, Légaré F, Witteman HO. Diabetes-related complications: which research topics matter to diverse patients and caregivers? Health Expect 2018 Apr;21(2):549-559 [FREE Full text] [CrossRef] [Medline]
    28. Wayne N, Perez DF, Kaplan DM, Ritvo P. Health coaching reduces HbA1c in type 2 diabetic patients from a lower-socioeconomic status community: a randomized controlled trial. J Med Internet Res 2015 Oct 05;17(10):e224 [FREE Full text] [CrossRef] [Medline]
    29. National diabetes repository. Diabetes Action Canada - SPOR network.   URL: https://repository.diabetesaction.ca/ [accessed 2021-02-12]
    30. Willison DJ, Trowbridge J, Greiver M, Keshavjee K, Mumford D, Sullivan F. Participatory governance over research in an academic research network: the case of Diabetes Action Canada. BMJ Open 2019 Apr 20;9(4):e026828 [FREE Full text] [CrossRef] [Medline]
    31. Queenan JA, Williamson T, Khan S, Drummond N, Garies S, Morkem R, et al. Representativeness of patients and providers in the Canadian Primary Care Sentinel Surveillance Network: a cross-sectional study. CMAJ Open 2016;4(1):E28-E32 [FREE Full text] [CrossRef] [Medline]
    32. Weisman A, Tu K, Young J, Kumar M, Austin PC, Jaakkimainen L, et al. Validation of a type 1 diabetes algorithm using electronic medical records and administrative healthcare data to study the population incidence and prevalence of type 1 diabetes in Ontario, Canada. BMJ Open Diabetes Res Care 2020 Jun;8(1):e001224 [FREE Full text] [CrossRef] [Medline]
    33. Measuring health inequalities: a toolkit - Area-level equity stratifiers using PCCF and PCCF+. Canadian Institute for Health Information. 2018.   URL: https:/​/www.​cihi.ca/​sites/​default/​files/​document/​cphi-toolkit-area-level-measurement-pccf-2018-en-web.​pdf [accessed 2022-11-10]
    34. Table 1: Rural-urban distribution*. Repository.   URL: https://repository.diabetesaction.ca/wp-content/uploads/2020/11/ndr_ses.html [accessed 2021-07-01]
    35. Pekár S, Brabec M. Generalized estimating equations: a pragmatic and flexible approach to the marginal GLM modelling of correlated data in the behavioural sciences. Ethology 2017 Dec 11;124(2):86-93. [CrossRef]
    36. Curran-Everett D. Explorations in statistics: the log transformation. Adv Physiol Educ 2018 Jun 01;42(2):343-347 [FREE Full text] [CrossRef] [Medline]
    37. Schweiger B, Klingensmith GJ, Snell-Bergeon JK. Menarchal timing in type 1 diabetes through the last 4 decades. Diabetes Care 2010 Dec;33(12):2521-2523 [FREE Full text] [CrossRef] [Medline]
    38. Raha O, Sarkar B, Godi S, GhoshRoy A, Pasumarthy V, Chowdhury S, JDRF-India, et al. Menarcheal age of type 1 diabetic Bengali Indian females. Gynecol Endocrinol 2013 Nov;29(11):963-966. [CrossRef] [Medline]
    39. Mukherjee MS, Coppenrath VA, Dallinga BA. Pharmacologic management of types 1 and 2 diabetes mellitus and their complications in women of childbearing age. Pharmacotherapy 2015 Feb;35(2):158-174. [CrossRef] [Medline]
    40. Rasmussen B, Nankervis A, Skouteris H, McNamara C, Nagle C, Steele C, et al. Factors associated with breastfeeding to 3 months postpartum among women with type 1 and type 2 diabetes mellitus: an exploratory study. Women Birth 2020 May;33(3):e274-e279. [CrossRef] [Medline]
    41. Gaudio M, Dozio N, Feher M, Scavini M, Caretto A, Joy M, et al. Trends in factors affecting pregnancy outcomes among women with type 1 or type 2 diabetes of childbearing age (2004-2017). Front Endocrinol (Lausanne) 2020;11:596633 [FREE Full text] [CrossRef] [Medline]
    42. Caruso S, Cianci A, Cianci S, Monaco C, Fava V, Cavallari V. Ultrastructural study of clitoral cavernous tissue and clitoral blood flow from type 1 diabetic premenopausal women on phosphodiesterase-5 inhibitor. J Sex Med 2019 Mar;16(3):375-382. [CrossRef] [Medline]
    43. Purnamasari D, Puspitasari MD, Setiyohadi B, Nugroho P, Isbagio H. Low bone turnover in premenopausal women with type 2 diabetes mellitus as an early process of diabetes-associated bone alterations: a cross-sectional study. BMC Endocr Disord 2017 Nov 29;17(1):72 [FREE Full text] [CrossRef] [Medline]
    44. Sjöberg L, Pitkäniemi J, Harjutsalo V, Haapala L, Tiitinen A, Tuomilehto J, et al. Menopause in women with type 1 diabetes. Menopause 2011 Feb;18(2):158-163. [CrossRef] [Medline]
    45. Khalil N, Sutton-Tyrrell K, Strotmeyer ES, Greendale GA, Vuga M, Selzer F, et al. Menopausal bone changes and incident fractures in diabetic women: a cohort study. Osteoporos Int 2011 May;22(5):1367-1376 [FREE Full text] [CrossRef] [Medline]
    46. Scott AR, Dhindsa P, Forsyth J, Mansell P, Kliofem Study Collaborative Group. Effect of hormone replacement therapy on cardiovascular risk factors in postmenopausal women with diabetes. Diabetes Obes Metab 2004 Jan;6(1):16-22. [CrossRef] [Medline]
    47. Wickham H. tidyverse: Easily Install and Load the 'Tidyverse'. CRAN.   URL: https://CRAN.R-project.org/package=tidyverse [accessed 2021-07-08]
    48. Wickham H, François R, Henry L, Müller K, RStudio. dplyr: a grammar of data manipulation. CRAN. 2022 Sep 1.   URL: https://CRAN.R-project.org/package=dplyr [accessed 2022-12-19]
    49. Wickham H. The split-apply-combine strategy for data analysis. J Stat Soft 2011;40(1). [CrossRef]
    50. Robinson D, Hayes A. Convert statistical analysis objects into tidy tibbles. Broom.   URL: https://broom.tidymodels.org/reference/broom.html#author [accessed 2022-12-19]
    51. data.table: Extension of 'data.frame'. CRAN. 2022 Oct 17.   URL: https://cran.r-project.org/web/packages/data.table/index.html [accessed 2022-12-19]
    52. Barton K. Multi-Model Inference. CRAN. 2022 Aug 31.   URL: https://cran.r-project.org/web/packages/MuMIn/MuMIn.pdf [accessed 2022-12-19]
    53. Fox J, Weisberg S. An R Companion to Applied Regression. Thousand Oaks, California: SAGE Publications; 2010.
    54. Halekoh U, Højsgaard S, Yan J. The R Package geepack for generalized estimating equations. J Stat Soft 2005;15(2):1-11. [CrossRef]
    55. Yan J, Fine J. Estimating equations for association structures. Stat Med 2004 Mar 30;23(6):859-74; discussion 875. [CrossRef] [Medline]
    56. Yan J. Geepack: yet another package for generalized estimating equations. R-news 2002;2(3):12-14.
    57. Estimated marginal means, aka least-squares means. CRAN. 2022 Oct 28.   URL: https://cran.r-project.org/web/packages/emmeans/emmeans.pdf [accessed 2022-12-19]
    58. Ndjaboue R, Chipenda Dansokho S, Boudreault B, Tremblay M, Dogba MJ, Price R, et al. Patients' perspectives on how to improve diabetes care and self-management: qualitative study. BMJ Open 2020 Apr 29;10(4):e032762 [FREE Full text] [CrossRef] [Medline]
    59. Radin MS. Pitfalls in hemoglobin A1c measurement: when results may be misleading. J Gen Intern Med 2014 Feb;29(2):388-394 [FREE Full text] [CrossRef] [Medline]
    60. The First Nations Principles of OCAP®. OCAP.   URL: https://fnigc.ca/ocap-training/ [accessed 2021-06-18]
    61. Groft S, de la Paz MP. Rare diseases - avoiding misperceptions and establishing realities: the need for reliable epidemiological data. Adv Exp Med Biol 2010;686:3-14. [CrossRef] [Medline]
    62. Gregory JM, Slaughter JC, Duffus SH, Smith TJ, LeStourgeon LM, Jaser SS, et al. COVID-19 severity is tripled in the diabetes community: a prospective analysis of the pandemic's impact in type 1 and type 2 diabetes. Diabetes Care 2021 Feb;44(2):526-532 [FREE Full text] [CrossRef] [Medline]
    63. Barron E, Bakhai C, Kar P, Weaver A, Bradley D, Ismail H, et al. Associations of type 1 and type 2 diabetes with COVID-19-related mortality in England: a whole-population study. Lancet Diabetes Endocrinol 2020 Oct;8(10):813-822 [FREE Full text] [CrossRef] [Medline]
    64. McGurnaghan SJ, Weir A, Bishop J, Kennedy S, Blackbourn LA, McAllister DA, Public Health Scotland COVID-19 Health Protection Study Group, Scottish Diabetes Research Network Epidemiology Group. Risks of and risk factors for COVID-19 disease in people with diabetes: a cohort study of the total population of Scotland. Lancet Diabetes Endocrinol 2021 Feb;9(2):82-93 [FREE Full text] [CrossRef] [Medline]
    65. Cooney E. People with type 1 diabetes have a higher risk of dying from Covid-19. Why are they lower on CDC’s vaccine priority list? STAT. 2021 Jan 11.   URL: https:/​/www.​statnews.com/​2021/​01/​11/​for-people-with-type-1-diabetes-cdc-guidelines-for-covid-19-vaccine-priority-are-puzzling/​comment-page-1/​ [accessed 2021-07-21]
    66. Foster NC, Beck RW, Miller KM, Clements MA, Rickels MR, DiMeglio LA, et al. State of type 1 diabetes management and outcomes from the T1D exchange in 2016-2018. Diabetes Technol Ther 2019 Feb;21(2):66-72 [FREE Full text] [CrossRef] [Medline]
    67. Beck RW, Tamborlane WV, Bergenstal RM, Miller KM, DuBose SN, Hall CA, T1D Exchange Clinic Network. The T1D Exchange clinic registry. J Clin Endocrinol Metab 2012 Dec;97(12):4383-4389. [CrossRef] [Medline]
    68. Canadian Diabetes Association Clinical Practice Guidelines Expert Committee, Cheng AY. Canadian Diabetes Association 2013 clinical practice guidelines for the prevention and management of diabetes in Canada. Introduction. Can J Diabetes 2013 Apr;37 Suppl 1:S1-S3. [CrossRef] [Medline]
    69. Hermann JM, Miller KM, Hofer SE, Clements MA, Karges W, Foster NC, T1D Exchange Clinic Network and the DPV initiative. The Transatlantic HbA gap: differences in glycaemic control across the lifespan between people included in the US T1D Exchange Registry and those included in the German/Austrian DPV registry. Diabet Med 2020 May;37(5):848-855. [CrossRef] [Medline]
    70. Feig DS, Berger H, Donovan L, Godbout A, Kader T, Keely E, et al. Diabetes and pregnancy. Diabetes Canada. 2021.   URL: https://guidelines.diabetes.ca/cpg/chapter36 [accessed 2021-07-16]
    71. Murphy HR. Continuous glucose monitoring targets in type 1 diabetes pregnancy: every 5% time in range matters. Diabetologia 2019 Jul;62(7):1123-1128 [FREE Full text] [CrossRef] [Medline]
    72. Myers S, Grasmick H. The social rights and responsibilities of pregnant women: an application of parsons's sick role model. J Applied Behavioral Sci 2016 Jul 26;26(2):157-172. [CrossRef]
    73. Berg M, Honkasalo ML. Pregnancy and diabetes--a hermeneutic phenomenological study of women's experiences. J Psychosom Obstet Gynaecol 2000 Mar;21(1):39-48. [CrossRef] [Medline]
    74. Berg M, Sparud-Lundin C. Experiences of professional support during pregnancy and childbirth - a qualitative study of women with type 1 diabetes. BMC Pregnancy Childbirth 2009 Jul 03;9:27 [FREE Full text] [CrossRef] [Medline]
    75. McGrath M, Chrisler JC. A lot of hard work, but doable: pregnancy experiences of women with type-1 diabetes. Health Care Women Int 2017 Jun;38(6):571-592. [CrossRef] [Medline]
    76. Yarde F, van der Schouw YT, de Valk HW, Franx A, Eijkemans MJ, Spiering W, OVADIA study group. Age at menopause in women with type 1 diabetes mellitus: the OVADIA study. Hum Reprod 2015 Feb;30(2):441-446. [CrossRef] [Medline]
    77. Yi Y, El Khoudary SR, Buchanich J, Miller R, Rubinstein D, Matthews K, et al. Women with Type 1 diabetes (T1D) experience a shorter reproductive period compared with nondiabetic women: the Pittsburgh Epidemiology of Diabetes Complications (EDC) study and the Study of Women's Health Across the Nation (SWAN). Menopause 2021 Mar 01;28(6):634-641 [FREE Full text] [CrossRef] [Medline]
    78. Yi Y, El Khoudary SR, Buchanich JM, Miller RG, Rubinstein D, Orchard TJ, et al. Association of age at diabetes complication diagnosis with age at natural menopause in women with type 1 diabetes: the Pittsburgh Epidemiology of Diabetes Complications (EDC) Study. J Diabetes Complications 2021 Mar;35(3):107832 [FREE Full text] [CrossRef] [Medline]
    79. Yi Y, El Khoudary SR, Buchanich JM, Miller RG, Rubinstein D, Orchard TJ, et al. Predictors of the age at which natural menopause occurs in women with type 1 diabetes: the Pittsburgh Epidemiology of Diabetes Complications (EDC) study. Menopause 2021 Apr 05;28(7):735-740 [FREE Full text] [CrossRef] [Medline]
    80. Keshawarz A, Pyle L, Alman A, Sassano C, Westfeldt E, Sippl R, et al. Type 1 diabetes accelerates progression of coronary artery calcium over the menopausal transition: the CACTI study. Diabetes Care 2019 Dec;42(12):2315-2321 [FREE Full text] [CrossRef] [Medline]
    81. Mackay L, Kilbride L, Adamson KA, Chisholm J. Hormone replacement therapy for women with type 1 diabetes mellitus. Cochrane Database Syst Rev 2013 Jun 06(6):CD008613. [CrossRef] [Medline]
    82. Galbete A, Cambra K, Forga L, Baquedano FJ, Aizpuru F, Lecea O, et al. Achievement of cardiovascular risk factor targets according to sex and previous history of cardiovascular disease in type 2 diabetes: a population-based study. J Diabetes Complications 2019 Dec;33(12):107445. [CrossRef] [Medline]
    83. Global Gender Gap Report 2021. Geneva Switzerland: World Economic Forum; Mar 2021.
    84. Dinca-Panaitescu M, Dinca-Panaitescu S, Raphael D, Bryant T, Pilkington B, Daiski I. The dynamics of the relationship between diabetes incidence and low income: longitudinal results from Canada's National Population Health Survey. Maturitas 2012 Jul;72(3):229-235. [CrossRef] [Medline]
    85. Lysy Z, Booth GL, Shah BR, Austin PC, Luo J, Lipscombe LL. The impact of income on the incidence of diabetes: a population-based study. Diabetes Res Clin Pract 2013 Mar;99(3):372-379. [CrossRef] [Medline]
    86. Bird Y, Lemstra M, Rogers M, Moraros J. The relationship between socioeconomic status/income and prevalence of diabetes and associated conditions: a cross-sectional population-based study in Saskatchewan, Canada. Int J Equity Health 2015 Oct 12;14:93 [FREE Full text] [CrossRef] [Medline]
    87. Zuijdwijk CS, Cuerden M, Mahmud FH. Social determinants of health on glycemic control in pediatric type 1 diabetes. J Pediatr 2013 Apr;162(4):730-735. [CrossRef] [Medline]
    88. Butalia S, Patel AB, Johnson JA, Ghali WA, Rabi DM. Geographic clustering of acute complications and sociodemographic factors in adults with type 1 diabetes. Can J Diabetes 2017 Apr;41(2):132-137. [CrossRef] [Medline]
    89. Beryl Pilkington F, Daiski I, Bryant T, Dinca-panaitescu M, Dinca-panaitescu S, Raphael D. The experience of living with diabetes for low-income Canadians. Can J Diabetes 2010;34(2):119-126. [CrossRef]
    90. Visekruna S, McGillis Hall L, Parry M, Spalding K. Intersecting health policy and the social determinants of health in pediatric type 1 diabetes management and care. J Pediatr Nurs 2017;37:62-69. [CrossRef] [Medline]
    91. Campbell R, Larsen M, DiGiandomenico A, Davidson M, Booth G, Hwang S, et al. The challenges of managing diabetes while homeless: a qualitative study using photovoice methodology. CMAJ 2021 Jul 11;193(27):E1034-E1041. [CrossRef]
    92. Hershey JA, Morone J, Lipman TH, Hawkes CP. Social determinants of health, goals and outcomes in high-risk children with type 1 diabetes. Can J Diabetes 2021 Jul;45(5):444-50.e1. [CrossRef] [Medline]
    93. Ivers NM, Jiang M, Alloo J, Singer A, Ngui D, Casey CG, et al. Diabetes Canada 2018 clinical practice guidelines: key messages for family physicians caring for patients living with type 2 diabetes. Can Fam Physician 2019 Jan;65(1):14-24 [FREE Full text] [Medline]
    94. Aronson R, Brown RE, Abitbol A, Goldenberg R, Yared Z, Ajala B, et al. The Canadian LMC diabetes registry: a profile of the demographics, management, and outcomes of individuals with type 1 diabetes. Diabetes Technol Ther 2021 Jan;23(1):31-40. [CrossRef] [Medline]
    95. Haque WZ, Demidowich AP, Sidhaye A, Golden SH, Zilbermint M. The financial impact of an inpatient diabetes management service. Curr Diab Rep 2021 Jan 15;21(2):5 [FREE Full text] [CrossRef] [Medline]
    96. Ippolito MM, Lyles CR, Prendergast K, Marshall MB, Waxman E, Seligman HK. Food insecurity and diabetes self-management among food pantry clients. Public Health Nutr 2017 Jan;20(1):183-189 [FREE Full text] [CrossRef] [Medline]
    97. Sidawi B, Alhariri MT, Albaker WI. Creating a healthy built environment for diabetic patients: the case study of the eastern province of the Kingdom of Saudi Arabia. Glob J Health Sci 2014 Apr 14;6(4):136-147 [FREE Full text] [CrossRef] [Medline]
    98. Hsu C, Lee C, Wahlqvist ML, Huang H, Chang H, Chen L, et al. Poverty increases type 2 diabetes incidence and inequality of care despite universal health coverage. Diabetes Care 2012 Nov;35(11):2286-2292 [FREE Full text] [CrossRef] [Medline]
    99. What we fund. Diabetes Canada.   URL: https://www.diabetes.ca/research/what-we-fund [accessed 2021-07-27]
    100. Glauser W. Primary care system outdated and inconvenient for many millennials. CMAJ 2018 Dec 03;190(48):E1430-E1431 [FREE Full text] [CrossRef] [Medline]
    101. Primary health care providers, 2019. Statistics Canada. 2020 Oct 22.   URL: https://www150.statcan.gc.ca/n1/pub/82-625-x/2020001/article/00004-eng.htm [accessed 2022-12-19]
    102. Request Access. Diabetes Action Canada.   URL: https://repository.diabetesaction.ca/request-access/ [accessed 2022-12-22]

    ‎
    HbA1c: hemoglobin A1c

    Edited by K Mizokami-Stout; submitted 10.01.22; peer-reviewed by K McBrien, T Risling, A Syrowatka; comments to author 23.03.22; revised version received 14.10.22; accepted 15.10.22; published 27.04.23

    Copyright

    ©Seyedmostafa Mousavi, Dana Tannenbaum Greenberg, Ruth Ndjaboué, Michelle Greiver, Olivia Drescher, Selma Chipenda Dansokho, Denis Boutin, Jean-Marc Chouinard, Sylvie Dostie, Robert Fenton, Marley Greenberg, Jonathan McGavock, Adhiyat Najam, Monia Rekik, Tom Weisz, Donald J Willison, Audrey Durand, Holly O Witteman, Diabetes Action Canada Research Questions Prioritization Study. Originally published in JMIR Diabetes (https://diabetes.jmir.org), 27.04.2023.

    This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Diabetes, is properly cited. The complete bibliographic information, a link to the original publication on https://diabetes.jmir.org/, as well as this copyright and license information must be included.