Community-Based Culturally Preferred Physical Activity Intervention Targeting Populations at High Risk for Type 2 Diabetes: Results and Implications

ARTICLE IN PRESS Can J Diabetes xxx (2016) 1–9

Contents lists available at ScienceDirect

Canadian Journal of Diabetes journal homepage: w w w. c a n a d i a n j o u r n a l o f d i a b e t e s . c o m

Original Research

Community-Based Culturally Preferred Physical Activity Intervention Targeting Populations at High Risk for Type 2 Diabetes: Results and Implications Chip P. Rowan PhD a, Michael C. Riddell PhD a,b, Norman Gledhill PhD a, Veronica K. Jamnik PhD a,* a b

School of Kinesiology and Health Science, York University, Toronto, Ontario, Canada LMC Diabetes and Endocrinology & Manna Research, Toronto, Ontario, Canada

a r t i c l e i n f o

a b s t r a c t

Article history: Received 18 February 2016 Received in revised form 9 May 2016 Accepted 12 May 2016

Objectives: In Canada, an ageing population, obesity rates and high risk among certain ethnocultural populations are driving diabetes prevalence. Given the burden associated with type 2 diabetes and its link to modifiable risk factors, this study aimed to implement culturally preferred physical activities at the community level, targeting individuals at high risk for type 2 diabetes. Glycated hemoglobin (A1C) levels were used to detect potential improvements in glycemic control. Methods: Participants were screened for diabetes risk using a questionnaire and capillary point-of-care A1C blood testing. Participants were offered community-based physical activity classes 2 to 3 times per week for 6 months. A subset of participants (n=84) provided additional measurements. Results: In total, 718 subjects were reached during recruitment. Substantial participant dropout took place, and 487 participants were exposed to the intervention. Among those who participated in the physical activity and provided follow up, mean A1C levels were reduced by 0.17 (p=0.002) after 3 months (n=84) and by 0.06 (p=0.35; n=49) after 6 months. The homeostatic model assessment (HOMA-beta) showed a significant improvement of 23.6% after 3 months (n=20; p=0.03) and 45.2% after 6 months (n=12; p=0.02). Resting systolic blood pressure and diastolic blood pressure plus combined hand-grip strength improved after 6 months (n=12). Conclusions: Implementation of this community-based, culturally preferred physical activity program presented several challenges and was associated with significant participant dropout. After considering participant dropout, the relatively small group who participated and provided follow-up measures showed improvements various physiologic measures. Despite efforts to enhance accessibility, it appears that several barriers to physical activity participation remain and need to be explored to enhance the success of future programs. © 2016 Canadian Diabetes Association.

Keywords: community ethnicity exercise glycated hemoglobin (A1C) prediabetes

r é s u m é Mots clés : communauté de l’origine ethnique exercice en fonction hémoglobine glyquée (A1C) prédiabète

Objectifs : Au Canada, le vieillissement de la population, les taux et le risque élevé d’obésité au sein de certaines populations ethnoculturelles entraînent la prévalence du diabète. Étant donné le fardeau associé au diabète de type 2 et son lien avec les facteurs de risque modifiables, la présente étude avait pour objectif la mise en place d’activités physiques culturellement privilégiées à l’échelle communautaire qui visent les individus exposés à un risque élevé de diabète de type 2. Les concentrations d’hémoglobine glyquée (A1c) ont été utilisées pour déceler les améliorations potentielles de la régulation de la glycémie. Méthodes : Les participants ont été soumis à un dépistage du risque de diabète au moyen d’un questionnaire et d’un prélèvement sanguin par voie capillaire hors laboratoire de l’A1c. Les participants ont été invités à suivre des séances communautaires d’activités physiques 2 à 3 fois par semaine durant 6 mois. Un sous-ensemble de participants (n=84) a fourni des mesures supplémentaires.

* Address for correspondence. Veronica Jamnik, PhD, School of Kinesiology and Health Science, York University, 358 Bethune College, 4700 Keele Street, Toronto, Ontario M3J 1P3, Canada. E-mail address: [email protected] 1499-2671 © 2016 Canadian Diabetes Association. http://dx.doi.org/10.1016/j.jcjd.2016.05.011

ARTICLE IN PRESS 2

C.P. Rowan et al. / Can J Diabetes xxx (2016) 1–9

Résultats : Au total, 718 sujets ont été choisis durant le recrutement. Après qu’un nombre important de participants eurent abandonné, 487 participants ont participé à l’intervention. Parmi ceux qui ont pris part à l’activité physique et fourni des mesures de suivi, les concentrations moyennes d’A1c ont montré une réduction de 0,17 (p=0,002) après 3 mois (n=84) et de 0,06 (p=0,35; n=49) après 6 mois. Le modèle d’évaluation de l’homéostasie (HOMA-bêta) a montré une amélioration significative de 23,6 % après 3 mois (n=20; p=0,03) et de 45,2 % après 6 mois (n=12; p=0,02). La pression artérielle systolique et diastolique au repos et la force de préhension combinées se sont améliorées après 6 mois (n=12). Conclusions : La mise en place de ce programme communautaire d’activités physiques culturellement privilégiées comportait de nombreux défis et était associée à un abandon significatif des participants. Ceux qui ont participé et fourni des mesures de suivi ont montré des améliorations dans les diverses mesures physiologiques. En dépit des efforts pour améliorer l’accessibilité, il semble que plusieurs obstacles à la participation aux activités physiques demeurent et doivent être examinés pour accroître le succès des programmes futurs. © 2016 Canadian Diabetes Association.

Introduction As type 2 diabetes continues to impact funding of healthcare systems dramatically around the world, more community-based resources are needed to help limit the development of the disease. A shift in focus toward prevention is imperative, considering that management of diabetes and its complications are projected to cost approximately $17 billion in Canada by the year 2020 (1). Without alterations in the current healthcare paradigm, key stakeholders (government, employers and insurers) will all struggle to manage the overwhelming financial burden of diabetes and its associated complications. A shift of responsibility, allowing physicians to share the treatment and prevention burdens with other qualified professionals, such as certified exercise physiologists and community physical activity leaders, would broaden the reach of community-based diabetes-prevention initiatives. A targeted approach focusing on those at highest risk for developing diabetes should translate into reducing the impact of this disease on our healthcare system. There has been a great deal of research into the effects of physical activity (PA) on the treatment/ prevention of type 2 diabetes (2,3) and prediabetes (4). In Canada, there is a highly diverse population consisting of several ethnicities that are known to be at elevated risk for type 2 diabetes. It has been well documented that individuals of South Asian, Chinese, African-Caribbean and Aboriginal descent are at 3 to 5 times higher risk for developing type 2 diabetes, which tends to happen at earlier ages and in a seemingly healthier body compositions (5–8). It is important that prevention strategies in Canada address these high-risk ethnicities and identify ethnospecific recommendations for the assessment of diabetes risk and for effective PA-focused intervention. When targeting high-risk ethnicities in the community, volunteer support plus “buy in” and engagement by community members can be key contributors to successful intervention (9,10). This study was designed to identify persons at risk for type 2 diabetes from specific high-risk ethnicities and enrol them in a PA program that was community based and culturally preferred. The primary goal of the study was to successfully implement a communitybased PA program specific to individuals at high risk for type 2 diabetes by using a culturally preferred approach. The success of the program was determined primarily through participant adherence and measured changes in levels of glycated hemoglobin (A1C) over a 6-month period. Additional measures of physical and physiologic fitness and health were also examined in a subset of participants to evaluate potential concurrent health benefits relevant to prediabetes and type 2 diabetes as well as to several other chronic diseases. Methods Study design This was a nonrandomized longitudinal effectiveness study designed to target persons at high risk for type 2 diabetes, as

determined by the Prediabetes Detection and Physical Activity Intervention Delivery (PRE-PAID) risk questionnaire (11,12) and pointof-care capillary blood screening (A1C) for a period of 6 months during which these persons participated in various forms of communitybased PA. This investigation received ethics approval from the York University Human Participants Review subcommittee, and prior to screening and baseline data collection, all participants provided written, informed consent. Recruitment efforts and screening opportunities targeting persons at high risk for type 2 diabetes were concentrated in specific communities chosen on the basis of their cultural demographics as well as their known prevalences of diabetes. The selection process utilized the City of Toronto neighbourhood profiles that were developed in conjunction with Statistics Canada (13) as well as the Institute for Clinical Evaluative Studies Atlas of Diabetes (14). Communities with high reported rates of type 2 diabetes and self-identified visible minorities were selected for participant recruitment. In each community, participants from the identified, high-risk ethnicities (Chinese, South Asian, Aboriginal and African Caribbean) were targeted, but all who expressed interest were able to participate, regardless of ethnicity. This process took place in various public locations, such as community health centres, religious centres and shopping malls and relied heavily upon partnerships with community organizations to provide access to space, volunteer support and a recognizable trusted relationship with community members. A schematic of participant flow through the components of the study is provided in Figure 1, and select participant demographics are illustrated in Figure 2. Participant inclusion/exclusion criteria Participants were initially screened using the PRE-PAID risk questionnaire and point-of-care capillary A1C levels based on fingerstick evaluations. Despite the fact that the project targeted those at high risk for type 2 diabetes, all participants who underwent screening were invited to attend the community-based PA intervention, regardless of diabetes status, in an effort to promote participation and social support. When classifying participants’ A1C levels, the Canadian Diabetes Association (CDA) diagnostic criteria were used for normal glycemic control (<6.0%), prediabetes (6.0% to 6.4%) and diabetes (≥6.5%) (15). Participants were excluded if they were under the age of 18 or if they possessed a condition or functional limitation that would preclude their engagement in moderate-intensity PA. Such conditions included chronic heart failure, severe musculoskeletal injury or joint or mobility pain. All participants completed the Physical Activity Readiness Questionnaire for Everyone (PAR-Q+) to identify any potential risks for PA participation prior to enrolment (16). If risk factors were identified, subsequent questions are provided on the PAR-Q+ (www.eparmedx.com) and, in many cases, they determine that the individual is cleared for unrestricted progressive light- to moderate-intensity PA participation. The PAR-Q+ was administered and interpreted by a qualified exercise professional

ARTICLE IN PRESS C.P. Rowan et al. / Can J Diabetes xxx (2016) 1–9

3

Total Recruited n=760 Recruitment

Intervention

Participants who attended 1 PA session or more but provided no A1C follow up n=403 Participants who attended 1 PA session or more and provided A1C at 3 months but not 6 months n=35

Participants who attended 1 PA session or more and provided A1C at 6 months but not 3 months n=10

Total with Bio-Rad A1C results n=718

Participants who declined blood testing n=39 Participants who failed to attend a single PA session n=231

Total who participated in at least 1 communitybased PA Session n=487 (exposed to intervention)

Total who provided followup A1C sample after 3 months n=84

Total who provided followup A1C sample after 6 months n=49

Total who provided followup A1C sample at both 3 and 6 months n=39

Total who elected to provide secondary outcome measures in addition to A1C at baseline. n=84

Total who provided 3-month follow-up secondary blood measures n=21

Participants who provided baseline secondary blood measures but no 3-month followup measures n=63

Total who provided 6-month follow-up secondary blood measures n=14

Participants who provided baseline secondary blood measures but no 6-month followup measures n=70

Total who provided 6-month follow-up fitness measures n=21

Participants who provided baseline secondary fitness measures but no 6-month followup measures n=63

Figure 1. Flow diagram of participants, including participation in the PA intervention, provision of follow-up A1C samples plus participants who provided secondary outcomes at baseline and upon follow up. AC, African-Caribbean; SA, South Asian; SSM, Sault Ste. Marie, Ontario, Canada.

(QEP) in all cases. English proficiency was encouraged but not required, as crucial project materials were translated into Chinese, Punjabi and Hindi to enhance clarity for participants. Blood measures The primary blood outcome measure was A1C level, and it was measured using the Bio-Rad in2it point-of-care device (Bio-Rad Laboratories, Hercules, California, USA), which performs the analysis using boronate affinity chromatography. Capillary blood samples were collected via fingerstick, using sterile techniques. A 75 gram oral glucose tolerance test (OGTT) was also performed on a subset of participants who elected to provide the secondary outcome measures. For this test, participants arrived at the laboratory in a fasted state (no food or drink for a minimum of 8 hours) and provided baseline intravenous blood samples (10 mL). Upon sample collection, participants were provided with a 75 gram dose of glucose (Trutol; ThermoFisher Scientific, Waltham, Massachusetts, USA) to be consumed within a 5-minute period. Blood samples (5 mL) were then drawn every 30 minutes from an indwelling venous catheter so that a single venipuncture was required for the entire test to alleviate potential burdens and discomfort for the participants. Blood samples were analyzed (LifeLabs, Thunder Bay, Ontario, Canada) for the following blood measures; fasting glucose, fasting insulin, high-density lipoprotein-cholesterol (HDL-C), lowdensity lipoprotein-cholesterol (LDL-C), total cholesterol, triglycerides,

HDL to LDL as well as glucose and insulin values for each of the time points (30, 60, 90 and 120 minutes). From the results of the OGTT, the calculation of the homeostatic model assessment (HOMA-beta), HOMA-insulin resistance (IR) was performed using standard equations (17). These measures provide indirect information about the participants’ beta cell functions and insulin resistance status. Physical and physiologic fitness The term physical fitness refers to anthropometric and strength measures, whereas physiologic fitness refers to blood pressure, resting heart rate (HR) and aerobic fitness. These characteristics were assessed in the subset of participants who agreed to come to the laboratory for additional testing. The assessment included measures of body composition, resting blood pressure and resting HR using the BpTRU device (BpTRU Medical Devices, Coquitlam, British Columbia, Canada), aerobic fitness and various strength measures. Height was measured using a stadiometer and body mass plus percent body fat were measured using a digital scale (Tanita Corporation of America, Arlington, Illinois, USA). Waist circumference was measured by a QEP following the National Institutes of Health guidelines (18). Upper-body strength was assessed using a hand-grip dynamometer (Takei T.K.K. 5401, Niigata, Japan), and lower-body power was measured via vertical jump (Vertec, JumpUSA, Sunnyvale, California, USA) and Probotics Just Jump Mat (Probotics, Huntsville, Alabama, USA). A customized, incremental

ARTICLE IN PRESS 4

C.P. Rowan et al. / Can J Diabetes xxx (2016) 1–9

Sex: Baseline Male

A1C: Baseline

Female

Normal

Prediabetes 19%

27%

73%

Health- and lifestyle-related questionnaires Diabetes

During the time spent between OGTT samples, participants completed questionnaires to help describe their lifestyle habits and psychosocial characteristics. The package included a study-participation questionnaire, in which participants identified their places of birth, times of entering Canada, languages spoken, religions and levels of religious devotion, levels of education, marital status and number of children, menopausal status and whether or not they had family doctors. The International Physical Activity Questionnaire (20) was also administered to ascertain participants’ levels of job-related PA participation, transportation PA, housework or maintenance PA, recreational, sport and leisure-time PA and time spent sitting. An acculturation scale (21,22) assessed the adjustment level to Canadian living if participants had not been born in Canada. The EuroQol Five Dimensions questionnaire (EQ-5D) (23) was included to assess perceived levels of mobility, self-care, activity, pain, anxiety and overall health status. A risk-factor knowledge and lifestyle habits questionnaire gauged participants’ perceptions and awareness of recommended healthy lifestyle habits, including both diet and PA participation. A perceived stress scale (24) was completed and, finally, the 12-Item Short Form Health Survey (SF-12) was administered to assess participants’ health-related quality of life (25).

26%

55%

A1C: 3 Months

Sex: 3 Months Male

Normal

Female 20%

Diabetes

31%

56%

80%

Sex: 6 Months Male

Prediabetes 13%

A1C: 6 Months

Female

Normal

Prediabetes 8%

29%

71%

Diabetes

23%

69%

Ethnicity: Baseline Chinese

African-Caribbean

South Asian 7% 8%

5% 15%

Aboriginal

Community-based physical activity intervention Caucasian

Other

25% 40%

Ethnicity: 3 Months Chinese

African-Caribbean 6%

8%

South Asian 4%

Aboriginal

Caucasian

Other

Caucasian

Other

1%

8%

73%

Ethnicity: 6 Months Chinese

African-Caribbean

South Asian Aboriginal 4% 2% 0% 4% 0% 90%

Figure 2. Summary of demographics (% of total number) of participants at baseline (n=718), at 3 months (n=84) and at 6 months (n=49).

aerobic fitness treadmill walking protocol was used and VO2 was measured by direct analysis of expired gas using open-circuit spirometry (O 2 and CO 2 sensors; AEI Technologies, Pittsburgh, Pennsylvania, USA). The protocol was adapted from that of Ebbeling et al (19), in which the participant walked for 4 minutes at 0 grade to determine a walking speed that was safe and comfortable while inducing an HR that was 50% to 70% age-predicted HRmax. Participants then walked for an additional 4 minutes at a 5% grade at the same walking speed. After the second 4-minute workload, participants completed subsequent 2-minute intervals at the same walking speed while adding a 2% grade at each interval until the test was terminated. Aerobic fitness test termination was determined by volitional fatigue. During the aerobic fitness test, HR and VO2 peak were recorded, along with time on treadmill and maximum speed and grade achieved. The fitness test was repeated only at 6 months, not at 3 months, in order to minimize participants’ burdens.

At conveniently situated locations, such as community centres or churches in each of the target communities, culturally preferred PA classes were provided and were instructed by culturally matched fitness instructors 2 to 3 times per week in each of the selected communities; sessions lasted 60 to 90 minutes. In total, 4 primary communities held PA sessions, and the locations within each of these communities remained fairly constant, with a few exceptions. In the event that a location had to be changed, participants were given ample notice, and the updated locations were still in close proximity to the initial screening/recruitment session locations. The average total number of sessions offered in each community was 56 over the 6-month time frame. A combination of weekday evenings and weekend days were selected so as to promote convenience and attendance. The days and times were selected based on participants’ feedback during the recruitment process. Public transit tokens and child-minding services were provided at the majority of the sessions in an attempt to eliminate child supervision as a potential barrier to attendance. The primary goal of the PA classes was for participants to attain 150 minutes per week of supervised, moderate-intensity PA. This goal is in accordance with recommendations set forth by Canadian and international governing bodies, including the CDA and the World Health Organization (15,26). Participants were also encouraged to increase their PA during other days of the week and to document it in a PA journal that was provided. The selection of culturally preferred PA was identified through informal focus-group interviews with participants prior to their engagement in the program. From this, the following PA programs were selected: Socasize and African-Dance for the AfricanCaribbean population, Bollyfit for the South Asian population and line dancing and tai chi for the Chinese population. These groupbased classes all involve whole-body, rhythmic and aerobic movements and were supplemented by body-weight-based resistance exercises to enhance the stimulus of health benefits. The classes were set to culturally themed music and were provided by culturally matched instructors. The Aboriginal group indicated preferences for more traditional PA regimens involving walking and traditional calisthenics. In all cases, the activity sessions were open to the entire community and were provided free of charge. Throughout the PA intervention, the sessions were monitored by QEPs to ensure safety and appropriate PA intensity. The QEPs involved in this project

ARTICLE IN PRESS C.P. Rowan et al. / Can J Diabetes xxx (2016) 1–9

possessed advanced fitness certification and undergraduate education in kinesiology as well as specific background knowledge pertaining to chronic-disease management and prevention. Both the QEPs and the culturally matched PA instructors received specific training to ascertain intensity levels during the group-based PA classes. Statistical analyses Participants’ demographics and various baseline outcome measures were analyzed using descriptive statistics. Baseline questionnaire data were summarized using descriptive statistics and frequencies. Prior to analysis, all participants were pooled together, regardless of their community affiliation or ethnicity. Temporal changes in A1C levels and additional blood tests were analyzed using paired t tests, which were performed for all comparisons between 2 time points. Given that this study should be considered an effectiveness trial as opposed to an efficacy trial, all subjects who provided follow-up measures, regardless of their adherence to the intervention, were included. Analyses described in this investigation were performed using a 2-sided 5% level of significance. All analyses were performed using IBM SPSS v. 22 (Chicago, Illinois, USA). Results A total of 760 participants were reached directly by the PREPAID study personnel, and 718 elected to provide baseline A1C measurements. None of the participants identified during the recruitment process were excluded from participation in the PA intervention classes because of any physical or health limitations. In all cases, if an interested participant identified a potential risk factor for PA participation on the PAR-Q+, the questionnaire, interpreted by a QEP, still determined that the accumulation of 150 minutes of moderateintensity PA was recommended. None of the participants required additional screening prior to PA participation. The results of the prediabetes screening portion of the study have been previously published (11). A total of 487 participants were exposed to the PA intervention and took part in a minimum of 1 PA class over the entire study duration, and 84 individuals elected to provide all of or some of the secondary outcome measures at baseline. Figure 1 illustrates that a total of 84 participants provided follow-up A1C measurements at 3 months, 49 provided A1C levels at 6 months, and 39 provided A1C levels at all 3 time points. Figure 2 illustrates the sex, ethnicity and A1C-level results from participants at baseline, 3 months and 6 months. Figure 1 also shows that 21 participants provided a portion or all of the follow-up secondary measures and the significant dropout rate observed during the study. Given the large scope of the intervention, the use of multiple communitybased locations and the limited access to the participants, capturing comprehensive data for all participants who dropped out was not accomplished. Participant adherence Of the 718 participants who provided baseline A1C measurements during recruitment, 487 (66% of the initial recruitment)

5

attended 1 or more PA intervention sessions and were thus exposed to the intervention. Only 19% of those who attended 1 or more PA sessions provided follow-up A1C measurements at either 3 or 6 months. Across the primary communities offering PA sessions, the average number of sessions offered over the entire 6 months was 56. The mean participation rate for all participants was 19%. This is equivalent to an average of 10.3 sessions over a period of 6 months. A total of 144 participants were able to attend 10 sessions or more. The Chinese cohort displayed the greatest adherence to the program, with an average adherence rate of 58%. Participants from this cohort were also the only group to provide 3- and 6-month follow-up data for the secondary outcomes assessed at baseline. No adverse events were reported during the PA sessions or laboratory testing days. Blood measures The mean A1C level for the entire subject pool (n=718) at baseline was 6.1%±0.87 (mean±SD). These data include both the A1C results from the entire pool of participants as well as the findings from those who provided secondary outcomes assessed during the OGTT. The results of the analysis of temporal changes in A1C levels are shown in Table 1. There were a total of 84 valid cases for analysis at 3 months and 49 valid cases at 6 months. Only 39 participants provided both 3- and 6-month follow-up measures. The results show that there was a significant reduction in A1C levels of 0.16% from baseline to 3 months’ follow up (p=0.003). No significant differences were observed between baseline and 6 months or between 3- and 6-month follow up, probably because of considerable participant dropout. Fasting glucose concentrations fell by 0.33 mmol/L between baseline and 3-month follow up (p=0.02). HOMA-beta, a marker of pancreatic beta cell function, increased by 23.6% between baseline and 3 months (p=0.03) and by 45.2% between baseline and 6 months (p=.02). HOMA-IR, a marker of insulin resistance, showed a significant reduction of −7.7 (p<0.001) between baseline and 3 months and −8.5 (p=0.003) between baseline and 6 months. No significant differences were observed in the blood lipid profiles. These results are reported in Table 2. Physical and physiologic fitness The body-composition profile of this group was characterized by a mean waist circumference of 91.7 cm, a mean body mass index of 26.2 kg·m2–1 and a mean body fat level of 32.3%. In addition to this, the mean VO2 peak (n=71) was 26.8 mL·kg−1·min−1, which indicates a poor level of aerobic fitness (27). Fitness results (n=12) represent an entirely Chinese population who were, by far, the most adherent to the PA intervention. The results indicate that after 6 months, there was a significant reduction in resting systolic blood pressure of 11 mm Hg (p=.005) and a reduction in diastolic blood pressure of 14 mm Hg (p=.04). In addition to the improvement in blood pressure levels, combined grip strength increased 9.32 kg (p<.001), relative VO2 peak increased 5.15 mL·kg−1·min−1(p=.003) and peak treadmill speed increased by 0.29 mph (p<.001). The temporal results from the fitness assessment are summarized in Table 3.

Table 1 Results of paired t tests showing temporal changes in A1C levels at 3 months (n=84) and at 6 months (n=49) measured using the Bio-Rad in2it devicea 95% confidence interval Variable

Timepoint

n

Mean

A1C (%)

Baseline 3 months Baseline 6 months

84 84 49 49

6.1 6.0 6.1 6.0

* Significance; p<0.05.

Mean difference from baseline

p value

−0.17

0.002*

−0.06

0.35

Lower boundary

Upper boundary

0.06

0.27

−0.71

0.19

ARTICLE IN PRESS 6

C.P. Rowan et al. / Can J Diabetes xxx (2016) 1–9

Table 2 Results of paired t tests showing temporal changes in secondary blood measures at 3 months (n=21) and at 6 months (n=12) measured during the OGTT 95% confidence interval Variable

Time point

n

Mean

Total cholesterol (mmol/L)

Baseline 3 months Baseline 6 months Baseline 3 months Baseline 6 months Baseline 3 months Baseline 6 months Baseline 3 months Baseline 6 months Baseline 3 months Baseline 6 months Baseline 3 months Baseline 6 months Baseline 3 months Baseline 6 months Baseline 3 Months Baseline 6 months Baseline 3 months Baseline 6 months Baseline 3 months Baseline 6 months Baseline 3 months Baseline 6 months

19 19 12 12 19 19 12 12 19 19 12 12 19 19 12 12 19 19 12 12 19 19 12 12 20 20 12 12 20 20 12 12 18 18 10 10 20 20 13 13 21 21 12 12

5.1 5.1 5.3 5.3 3.0 3.1 3.1 3.1 1.6 1.5 1.6 1.6 3.4 3.5 3.5 3.5 1.2 1.2 1.3 1.4 5.4 5.1 5.4 5.2 35.3 35.0 38.8 48.7 51.8 75.4 49.2 94.4 9.1 1.4 10.6 2.1 7.1 7.5 6.7 6.6 441.8 584.1 492.2 808.5

LDL cholesterol (mmol/L)

HDL cholesterol (mmol/L)

TC:HDL

Triglycerides (mmol/L)

Fasting glucose (mmol/L)

Fasting insulin (pmol/L)

HOMA-beta

HOMA-IR

2-hour glucose (mmol/L)

2-hour insulin (pmol/L)

Mean difference from baseline

p value

Lower boundary

Upper boundary

−0.04

0.77

−0.31

0.24

−0.03

0.89

−0.53

0.46

−0.08

0.40

−0.26

0.11

0.01

0.96

−0.34

0.36

0.04

0.24

−0.03

0.10

−0.01

0.80

−0.09

0.07

−0.09

0.40

−0.31

0.13

−0.06

0.67

−0.35

0.23

0.00

1.00

−0.28

0.28

−0.07

0.62

−0.38

0.23

0.33

0.02*

0.07

0.59

0.28

0.21

−0.18

0.73

0.30

0.91

−5.15

5.75

29.89

0.001*

14.66

45.11

−23.60

0.03*

−44.30

−2.89

−45.21

0.02*

−80.11

−10.30

7.72

<0.001*

4.79

10.65

8.49

0.003*

3.77

13.22

−0.37

0.24

−0.99

0.26

0.07

0.54

−0.17

0.31

−142.29

0.01*

−248.64

−35.93

−316.33

0.05

−631.73

−0.94

HDL, high-density lipoprotein; HOMA-IR, homeostatic model assessment-insulin resistance; LDL, low-density lipoprotein; OGTT, oral glucose tolerance test; TC, total cholesterol. * Significance; p<0.05.

Health- and lifestyle-related questionnaires Among those who elected to provide secondary outcomes (n=83), 58% (n=48) completed the questionnaire package. These data indicate that in the pool of participants who completed the questionnaires, 100% were born outside of Canada, yet the average time spent living in Canada was 20.4 years. Also, 46.9% reported that English was their second language. According to the International Physical Activity Questionnaire, respondents were generally inactive, as evidenced by a mean time spent sitting of 5.8 hours per weekday and 6.1 hours per weekend day as well as fewer than 2 days per week of participation in leisure-time PA. The EQ-5D results indicated that participants self-reported no physical limitations in performing self-care and typical activities, and 31% of participants reported moderate anxiety or depression. The EQ-5D visual analogue scale results also showed a mean score of 73.8/100 for selfreported overall health state. The results of the acculturation scale revealed that the respondents showed moderate agreement with almost all of the items on the questionnaire, which reflects a belief that both the practices of their heritage culture and those of “typical” Canadian culture are important and are followed. The risk-factor knowledge and lifestyle habits questionnaire revealed that 80% of

respondents perceived their knowledge of PA to be average or above, and 73.3% reported that PA was “very important” to them, whereas only 27.7% reported that were aware of the CDA’s PA guidelines. With respect to diet, 88.9% rated their diets as average or better, but only 17.4% reported having heard of Canada’s Food Guide. Results of the SF-12 questionnaire show that participants had a mean physical health composite score of 47.1/100 and a mean mental health composite score of 50.2/100.

Discussion The primary goal of this study was to attempt to implement a community-based PA program specific to individuals at high risk for type 2 diabetes using a culturally preferred approach. The study also aimed to demonstrate clinically significant reductions in A1C levels and improvements in other health and fitness parameters through regular participation in the culturally preferred PA intervention classes. Overall, we found modest improvement in nearly all cardiometabolic measures with this intervention, including A1C levels, but also found high participant dropout, even with a number of strategies in place to engage the community.

ARTICLE IN PRESS C.P. Rowan et al. / Can J Diabetes xxx (2016) 1–9

7

Table 3 Results of paired t tests showing temporal changes in secondary fitness measured during the follow-up fitness assessments at 6 months (n=12) 95% confidence interval Variable Weight (kg) Waist circumference (cm) % Body fat Resting HR (bpm) Resting SBP (mm Hg) Resting DBP (mm Hg) Combined grip (kg) Vertical jump height (cm) Jump mat height (inches) VO2 peak (mL·kg−1·min−1) Peak treadmill speed (mph) Peak treadmill grade (%) Peak HR (bpm)

Baseline 6 months Baseline 6 months Baseline 6 months Baseline 6 months Baseline 6 months Baseline 6 months Baseline 6 months Baseline 6 months Baseline 6 months Baseline 6 months Baseline 6 months Baseline 6 months Baseline 6 months

n

Mean

12 12 12 12 12 12 12 11 12 12 12 12 11 12 10 12 11 12 11 12 11 12 11 12 11 12

55.8 56.2 85.1 84.8 27.2 27.5 70.1 71.8 125.3 114.0 91.9 77.4 37.0 46.3 20.2 21.2 11.2 11.5 27.2 32.4 2.8 3.1 14.1 14.4 161.5 169.0

Mean difference from baseline

p value

Lower boundary

Upper boundary

−0.40

0.22

−1.07

0.27

0.26

0.84

−2.56

3.08

−0.31

0.54

−1.37

0.75

−0.73

0.77

−6.01

4.55

11.25

0.01*

3.58

18.92

14.50

0.03*

2.13

26.87

−9.22

<0.001*

−12.88

−5.56

−1.02

0.43

−3.82

1.79

−0.45

0.30

−1.36

0.47

−4.78

0.01*

−8.21

−1.35

−0.30

<0.001*

−0.43

−0.18

0.00

1.00

−1.30

1.30

−7.00

0.12

−16.06

2.06

DBP, diastolic blood pressure; HR, heart rate; SBP, systolic blood pressure. Note: All subjects providing follow-up fitness data were Chinese. * Significance; p<0.05.

Previous publications have noted that the community approach taken for this study can effectively identify persons with prediabetes (11), yet the ability to engage individuals in regular PA programming is clearly the major challenge. Given that the mean adherence rate for all participants was only ~20%, there is a great deal of room for improvement when it comes to retention of participants and adherence to PA sessions. Despite efforts to utilize community partners with established presences in the community, locations in close proximity to the participants’ homes and provision of culturally matched study personnel and PA instructors, adherence rates remained low. Without these measures to make the program accessible, culturally preferred and free, adherence would have likely been even lower. Interestingly, the subpopulation of Chinese participants had the highest adherence rate (~60%), while the other ethnocultural groups had much lower participation rates (<15%). Some of the barriers experienced by participants did not pertain to cultural backgrounds, so it is likely that many of these barriers could apply to all populations. The temporal changes described in the results provide limited evidence that improvements in glycemic control were achieved. The observed changes in A1C levels, although statistically significant, reflect only a small portion of the initial sample and were clinically modest compared to those reported in a meta-analysis of clinical randomized control trials in which reductions of 0.8% have been reported among persons with type 2 diabetes (28). Smaller reductions in A1C levels were expected in this study, given the lower baseline A1C levels. In general, those with a higher baseline A1C levels tended to have the greater opportunity for improved glycemic control with exercise in the population with type 2 diabetes (2). The clinically modest but statistically significant reductions in A1C levels from baseline to 3 months (for those who participated in the PA and pre- and postevaluations) in our study clearly indicated a group shift from a prediabetes classification toward a more normal state (15). Evidence for this potential improvement in glycemic control is supported by significant improvements in HOMA-beta, improving by 23.6%

(p=0.03) by 3 months and by 45.2% (p=0.02) by 6 months, indicating a significant improvement in beta-cell function. Once again, these results should be interpreted with caution because they reflect a small portion (~17% or less, depending on the variable and timepoint) of the sample, those who provided follow-up measures after baseline screening and participation in at least 1 PA session in their community. Positive changes in several of the secondary outcomes were also observed with our intervention among those who participated in the PA and provided pre- and postevaluations. Most notably, a reduction in both systolic and diastolic blood pressure of −11 mm Hg and −14 mm Hg, respectively, were observed. These reductions correspond to a change in classification from prehypertensive to normal, based on the American Heart Association guidelines (29,30) and, consequently, relate to expected reductions in cardiovascular mortality (29,30). It is important, however, to interpret these results with some scrutiny, given that they reflect only a small percentage of the participants who provided follow-up measurements (14% of those providing secondary outcomes, 2.5% of all participants who attended at least 1 PA session). The observed reductions in blood pressure are also significantly higher than those typically reported from PA interventions (28). Improvements in grip strength and VO2 peak are indicative of significant improvements in upper body strength and aerobic fitness, which are also known to be closely linked to reductions in cardiovascular mortality, improved functional capacity and improvements in quality of life (31–33). Hand-grip strength has been shown to be a predictor of disability and functional limitation later in life (34), whereas simply being among the lowest 20% of the population in terms of aerobic fitness is associated with a relative risk of 2.1 for all-cause mortality and 2.79 for cardiovascular mortality, respectively (33). Similar to the interpretation of the blood pressure results, these reflect only a small percentage of the participants who provided follow-up measurements (14% of those providing secondary outcomes and 2.5% of all participants who attended at least 1 PA session).

ARTICLE IN PRESS 8

C.P. Rowan et al. / Can J Diabetes xxx (2016) 1–9

Although the intensity of exercise for the entire group of participants was not measured directly, QEPs and properly trained instructors were present in all classes. This helped to maximize the likelihood of participants’ performing their activities at safe and effective intensity levels that would promote health benefits while simultaneously minimizing risks for adverse events. With respect to the baseline results from the various questionnaires, in this sample of persons with prediabetes, the score on the EQ-5D of 73.8/100 was lower than that of the general population of the United States (79.2/ 100) as well as that of a representative sample of American adults 40 to 60 years of age (78.5/100) (35). The observed EQ-5D score was, however, higher than a representative cohort of American individuals who had been diagnosed with diabetes (68.5/100) (35). The results of the Perceived Stress Scale-10 indicated that the mean score was indicative of stress levels that were above average compared to normative data (36). The results from the SF-12 showed that the mean values for both physical and mental health composite scores fall slightly below average compared to normative data from individuals aged 45 to 64 years (37). These results describe this study population as being below average in terms of overall health, perceived stress and PA participation. The primary limitation of this study is the lack of patients’ adherence to the program and the lack of follow-up data provided by the participants. This was a major challenge faced by the research team because the majority of sessions were held away from the laboratory, and there were no guarantees that participants who were eligible for their follow-up testing sessions would be in attendance during the data collection days, which were held during the regularly scheduled PA sessions. The lack of follow-up for the laboratorybased secondary outcomes is perhaps a reflection of the substantial time commitment and comprehensive and invasive nature of the study protocol. Another limitation of this study was the scarcity of male participants, which limits the ability to make conclusions regarding a more representative sample. Other potential limitations of the sample included the fact that subjects all self-selected to participate and that there was no randomization or control group, which would have allowed comparisons between participants who received no intervention and those who enrolled in the communitybased PA classes. Without a control group, implications and conclusions about temporal changes need to be interpreted cautiously because they only apply to the small number of participants who provided follow-up measurements and who were not necessarily representative of the sample at baseline. Continuing work in this field could pursue a similar study design using a randomized, controlled approach. It would require a great deal of resources and time to evaluate this type of program adequately. More appropriate, perhaps, would be to reimplement similar programs on a smaller scale, focusing specifically on 1 community or high-risk population. This would allow for more control and for additional resources to be allocated to the identification and alleviation of barriers to PA participation. Additional ways of enhancing adherence such as prearranged check-ins with QEPs could also promote participant accountability and facilitate continued participation in PA sessions. There should also be long-term follow-up at the community level, using diagnosis of type 2 diabetes as an outcome to detect the effectiveness of similar programs and to generate potential cost-savings models associated with reductions in diabetes progression. During these investigations, the role of the QEPs and other culturally-matched personnel should also be evaluated. Given their unique skill set and education regarding exercise prescription and lifestyle counselling, it is reasonable to propose that QEPs should be given a larger platform to facilitate programs at the community level, with the overall goal of chronic-disease prevention and management. In this study, despite the efforts of the investigators and volunteers, barriers to PA persisted, as evidenced by the large drop in participant numbers at each juncture

of the study design. An examination of long-term adherence to the program and identification of barriers that are culturally specific should also take place, and that may help to improve participant adherence in future programs.

Conclusions Implementation of community-based and culturally preferred PA interventions targeting persons with or at risk for type 2 diabetes is an important and challenging endeavour. There are many barriers to PA participation that, despite efforts to alleviate them, persisted and led to poor adherence to the PA intervention. Among those who provided follow-up measurements, participation in the PA intervention was associated with modest improvements in glycemic control observed using A1C, HOMA-beta and HOMA-IR levels. Other protective health benefits included improvements in strength and cardiovascular fitness, both of which are relevant to persons at risk for type 2 diabetes as well as several other chronic diseases. Although poor participant adherence was a limitation in this study, these results are in line with the overall project goals and should inform the design of future interventions aiming to prevent type 2 diabetes at the community level. Although only a small number of participants where highly adherent, it is at least somewhat encouraging that a large number were willing to try at least 1 PA session. This is an important first step toward habitual PA participation and provides evidence that community-based screening and awareness of diabetes risk are motivating factors. Maintaining participation, however, over long periods of time continues to be a challenge. Ongoing identification of barriers to PA, including those that are enjoyable, preferred and not necessarily culturally specific, need to be considered so as to enhance habitual PA participation and broaden the catchment of people who may benefit from these observed benefits, which may lead to improved quality of life and alleviation of the substantial financial burden faced by the healthcare system in Canada and worldwide.

Acknowledgments The authors thank PRE-PAID participants, volunteers and community partners. Funding for this initiative was provided by the Ontario Ministry of Health Promotion and Sport as well as the Ontario Trillium Foundation (9995128).

Author Contributions CPR is the primary author and was involved in study design, participant recruitment, data collection and analysis as well as manuscript preparation and revision; MCR was involved with study design, participant recruitment and critical revision of the manuscript; NG was involved with study design, participant recruitment and critical revision of the manuscript; VKJ was involved with study design, participant recruitment, data collection and critical revision of the manuscript.

References 1. Canadian Diabetes Association. An economic tsunami: The cost of diabetes in Canada. CDA, 2009. 2. Church T, Blair S, Cocreham S, et al. Effects of aerobic and resistance training on hemoglobin A1c levels in patients with type 2 diabetes. JAMA 2010;304:2253– 62.

ARTICLE IN PRESS C.P. Rowan et al. / Can J Diabetes xxx (2016) 1–9

3. Sigal R, Kenny G, Boulé N, et al. Effects of aerobic training, resistance training, or both on glycemic control in type 2 diabetes. Ann Intern Med 2007;147:357– 69. 4. Burr JF, Rowan CP, Jamnik VK, Riddell MC. The role of physical activity in type 2 diabetes prevention: Physiological and practical perspectives. Phys Sportsmed 2010;38:72–82. 5. Oldroyd J, Banerjee M, Heald A, Cruickshank K. Diabetes and ethnic minorities. Postgrad Med J 2005;81:486–90. 6. Chowdhury TA, Lasker SS. Complications and cardiovascular risk factors in South Asians and Europeans with early-onset type 2 diabetes. QJM 2002;95:241–6. 7. Anand SS, Yusuf S, Vuksan V, et al. Differences in risk factors, atherosclerosis and cardiovascular disease between ethnic groups in Canada: The study of health assessment and risk in ethnic groups (SHARE). Indian Heart J 2000;52:S35– 43. 8. Lorenzo C, Serrano-Ríos M, Martínez-Larrad MT, et al. Geographic variations of the International Diabetes Federation and the National Cholesterol Education Program-Adult Treatment Panel III definitions of the metabolic syndrome in nondiabetic subjects. Diabetes Care 2006;29:685–91. 9. Jones CA, Nanji A, Mawani S, et al. Feasibility of community-based screening for cardiovascular disease risk in an ethnic community: The South Asian Cardiovascular Health Assessment and Management Program (SA-CHAMP). BMC Public Health 2013;13:1–9. 10. Davachi S, Flynn MA, Edwards AL. A health region/community partnership for type 2 diabetes risk factor screening in Indo-Asian communities. Can J Diabetes 2005;29:87–94. 11. Rowan CP, Miadovnik LA, Riddell MC, et al. Identifying persons at risk for developing type 2 diabetes in a concentrated population of high-risk ethnicities in Canada using a risk assessment questionnaire and point-of-care capillary blood HbA1c measurement. BMC Public Health 2014;14:1–9. 12. Rowan CP, Riddell MC, Jamnik VK. The Prediabetes Detection and Physical Activity Intervention Delivery (PRE-PAID) program. Can J Diabetes 2013;37:415– 19. 13. Toronto Public Health, Statistics Canada. City of Toronto neighbourhood profiles 2008. http://www.toronto.ca/demographics/profiles_map_and_index.htm. Accessed May 20, 2009. 14. Booth GL, Creatore MI, Gozdyra P, et al. Patterns of diabetes prevalence, complications and risk factors: Neighborhood Environmental Resources for Healthy Living: A Focus Diabetes Toronto ICES atlas. Toronto, Ontario: Institute for Clinical Evaluative Studies, 2007. 15. Canadian Diabetes Association Clinical Practice Guidelines Expert Committee. 2013 Clinical Practice Guidelines. Can J Diabetes 2013;37:S1–212. 16. Warburton D, Jamnik V, Bredin S, Gledhill N. The Physical Activity Readiness Questionnaire for Everyone (PAR-Q+) and electronic Physical Activity Readiness Medical Examination (ePARmed-X+). Health Fit J Canada 2011;4:3–17. 17. Wallace T, Levy J, Matthews D. Use and abuse of HOMA modeling. Diabetes Care 2004;27:1487–95. 18. National Institutes of Health. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: The evidence report. Obes Res 1998;6:S51–210.

9

19. Ebbeling C, Ward A, Puleo E, et al. Development of a single-stage submaximal treadmill walking test. Med Sci Sports Exerc 1991;23:966–73. 20. Craig CL, Marshall A, Sjostrom M, et al. International Physical Activity Questionnaire: 12-country reliability and validity. Med Sci Sports Exerc 2003;35:1381– 95. 21. Ryder AG, Alden LE, Paulhus DL. Is acculturation unidimensional or bidimensional? A head-to-head comparison in the prediction of personality, selfidentity, and adjustment. J Pers Soc Psychol 2000;79:49–65. 22. Cuellar I, Arnold B, Maldonado R. Acculturation Rating Scale for Mexican Americans-II: A revision of the original ARSMA scale. Hisp J Behav Sci 1995; 17:275–304. 23. Group TE. EuroQol *: A new facility for the measurement of health-related quality of life. Health Policy 1990;16:199–206. 24. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav 1983;24:385–96. 25. Ware J, Keller S, Kosinski M. SF-12: How to score the SF-12 physical and mental health summary scales, 1998. 26. World Health Organization. Global recommendations on physical activity for health. Geneva, Switzerland; WHO, 2010. 27. Heyward VH, Gibson A. Advanced fitness assessment and exercise prescription. 7th edn. Champaign, IL: Human Kinetics, 2014. 28. Snowling N, Hopkins W. Effects of different modes of exercise training on glucose control and risk factors for complications in type 2 diabetic patients. Diabetes Care 2006;29:2518–27. 29. Pickering TG, Hall JE, Appel LJ, et al. Recommendations for blood pressure measurement in humans and experimental animals, Part 1: Blood pressure measurement in humans. Hypertension 2005;45:142–61. 30. Chobanian AV, Bakris GL, Black HR, et al. Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension 2003;42:1206–52. 31. Taylor LM, Spence JC, Raine K, et al. Physical activity and health-related quality of life in individuals with prediabetes. Diabetes Res Clin Pract 2010;90:15–21. 32. Leenders M, Verdijk LB, van der Hoeven L, et al. Patients with type 2 diabetes show a greater decline in muscle mass, muscle strength, and functional capacity with aging. J Am Med Dir Assoc 2013;14:585–92. 33. Blair SN, Kampert JB, Iii HWK, et al. Influences of cardiorespiratory fitness and other precursors on cardiovascular disease and all-cause mortality in men and women. JAMA 1996;276:205–10. 34. Rantanen T, Guralnik J, Foley D, et al. Midlife hand grip strength as a predictor of old age disability. JAMA 1999;281:558–60. 35. Lubetkin EI, Jia H, Franks P, Gold MR. Relationship among sociodemographic factors, clinical conditions, and health-related quality of life: Examining the EQ-5D in the U.S. general population. Qual Life Res 2005;14:2187–96. 36. Cohen S, Janicki-Deverts D. Who’s stressed? Distributions of psychological stress in the United States in probability samples from 1983, 2006 and 2009. J Appl Soc Psychol 2012;42:1320–34. 37. Gandek B, Ware JE, Aaronson NK, et al. Cross-validation of item selection and scoring for the SF-12 Health Survey in nine countries: Results from the IQOLA Project. J Clin Epidemiol 1998;51:1171–8.