SciELO - Scientific Electronic Library Online

vol.58 issue4Baja producción científica de decanos en facultades de medicina y salud de Colombia: ¿una realidad común en Latinoamérica?Underestimation of mortality caused by traffic in Mexico: an analysis at the subnational level author indexsubject indexsearch form
Home Pagealphabetic serial listing  

Services on Demand




Related links

  • Have no similar articlesSimilars in SciELO


Salud Pública de México

Print version ISSN 0036-3634

Salud pública Méx vol.58 n.4 Cuernavaca Jul./Aug. 2016 

Artículos originales

The individual contribution and relative importance of self-management and quality of care on glycaemic control in type 2 diabetes *

Contribución individual e importancia relativa del auto-cuidado y la calidad de la atención en el control glucémico de la diabetes tipo 2

Yolanda V Martínez, PhD1 

Stephen M Campbell, PhD1 

Mark Hann, PhD1 

Peter Bower, PhD1 

1NIHR School for Primary Care Research, University of Manchester. Manchester, UK.



To evaluate the relative importance of selfmanagement (SM) and quality of care (QoC) inpredicting glycaemic control in patients with type 2 diabetes.

Materials and methods:

A longitudinal cohort study was conducted in 204 adults diagnosed with type 2 diabetes. Self-management and quality of care were measured at baseline. HbA1c was measured at baseline and at six-month follow-up.


None of the measures of self-management were significantly associated with HbA1c.Treatment intensification (TI) (a proxy for quality of care) resulted in lower HbA1c at follow-up. Other variables were associated with HbA1c at follow-up: HbA1c at baseline, age, diabetes duration, and combination of oral glucose-lowering medications. An exploratory analysis showed that patients who did not receive treatment intensification but performed more self-management behaviours had lower HbA1c levels at follow-up.


Treatment intensification might be more important for glycaemic control than self-management but the interaction between treatment intensification and self-management needs further research.

Keywords: diabetes mellitus, type 2; hemoglobin A; glycosylated; family practice; self care; quality of health care; Mexico



Evaluar la importancia relativa del autocuidado (AU) y calidad de la atención (CA) para predecir control glucémico en diabetes mellitus tipo 2 (DM2).

Material y métodos:

Estudio longitudinal en 204 adultos con DM2. AU y CA evaluados en la medición basal. HbA1c evaluada en la medición basal y a los seis meses.


A los seis meses, ninguna de las mediciones de AU se asoció significativamente con HbA1c. La intensificación en el tratamiento (IT) (proxy de CA) se asoció con disminución de HbA1c. Otras variables asociadas con HbA1c: HbA1c en medición basal, edad, duración de diabetes, y combinación de anti-hiperglucemiantes. En un análisis exploratorio, los participantes que no recibieron IT pero desempeñaron más conductas de AU tuvieron niveles más bajos de HbA1c.


IT parece ser más importante para el control glucémico que AU, pero la interacción entre IT y conductas de AU se deben estudiar con más profundidad.

Palabras clave: diabetes mellitus tipo 2; hemoglobina A glucosilada; medicina familiar y comunitaria; autocuidado; calidad de la atención de salud; México

Complications and premature mortality can be minimised by interventions to keep glycaemic control under target levels as stipulated in clinical guidelines.1,2 The management of type 2 diabetes is broad and complex, including diet, exercise, education, oral antidiabetic medications, and insulin.3-6 Good diabetes management and good glycaemic control requires that patients both receive high quality clinical care and achieve effective self-management. 7

Previous studies have focused on specific aspects of self-management, such as diet and exercise,8 medication adherence,9,10 or specific aspects of quality of care, such as continuity of care.8 Very few observational studies have measured both self-management and quality of care at the same time,8,9 and most randomised controlled trials focusing on diabetes have not assessed the relative importance of these components.11-13

The objective of this study is to examine the individual contribution and relative importance of selfmanagement and quality of care in the prediction of glycaemic control at six-month follow-up in patients with type 2 diabetes.

Materials and methods

Participants were recruited from the Mexican Institute for Social Security (Instituto Mexicano del Seguro Social IMSS) in the city of Aguascalientes, Mexico (December 2009 to April 2010). Eligible participants were adults (≥40 years old), diagnosed with type 2 diabetes for at least 1 year and under consecutive IMSS care with a monthly prescription of oral glucose-lowering medications but no insulin prescription. This study was based on a model (Figure 1) including multiple measures of both quality of care and self-management related to glycaemic control.

Source: reference 14

Figure 1 Model of predictors of glycaemic control in patients with type 2 diabetes. Aguascalientes, Mexico, December 2009 - April 2010 

Data was collected prospectively. Self-management (SM), quality of care (QoC), clinical and demographic variables were measured at baseline.14 SM measures: Diabetes Knowledge Questionnaire (DKQ-24),15 Medical Prescription Knowledge Questionnaire (MPKQ),16 Diabetes Self-Efficacy (DSE) scale,17 and Summary of Diabetes Self-Care Activities (SDSCA).18 SDSCA was transformed to a total score using four items: healthy eating plan, physical activity, foot care, and taking medications ('good' diabetes self-management: ≥3 behaviours ≥4 days per week). QoC measures: Continuity item from the General Practice Assessment Questionnaire (GPAQ),19 Treatment intensification,9 Patient-Doctor Communication Scale,20 and Patient Satisfaction with Diabetes Care.21 Treatment intensification was evaluated from medical records using the availbale blood glucose measure (fasting blood glucose [FBG] or glycated hemoglobin [HbA1c]) given prioprity to HbA1c. Appropriate treatment intensification was assumed when general practitioners (GP) intensified treatment if necessary (patient had FBG >130 mg/dl [7.2 mmol/l] or HbA1c >7.0%) or didn't intensify treatment if it was not necessary (patient had FBG ≤130 mg/dl [7.2 mmol/l] or HbA1c ≤7.0%).

Age, gender, marital status, level of education, occupation, comorbidities, complications, duration of diabetes in years, and depression22 were collected from interviews. HbA1c was measured in all participants at baseline and follow-up as part of this study. Body mass index (BMI) and current medical prescription (oral glucose-lowering medications) were extracted from medical records.

This study complies with the principles of the Helsinki Declaration. Written informed consent was obtained at recruitment with approval of the Ethics Committee of Research on Human Beings at the University of Manchester (ref 09121) and from the IMSS Local Health Research Committee N° 101 (R-2009-101-12).

Sample size and selection

Assuming a correlation between SM and QoC of 0.1, an intra-cluster correlation of 0.1 [recognising that outcomes of patients at the same practice may not be independent, given that they consult the same GP(s)] and 20% loss to follow-up at six months, a sample of 405 patients would enable a difference as small as 0.2 (e.g. 0.25 vs. 0.05) to be detected between the correlations of HbA1c/SM and HbA1c/QoC with approximately 75% power at the 5% level of significance.23,24

Given the time and financial constraints of the study, it was decided that 80% of patients would be selected using a consecutive sample (all eligible patients attending an appointment would be approached). For the remaining 20% of the sample, a random sample of patients was also taken to evaluate selection bias. It was ensured that at least one patient per GP was selected. Any differences between the two samples were controlled for in the analyses.

Statistical analysis

Baseline data was summarised using descriptive statistics. First, univariate linear regressions were examined to evaluate the individual contribution of SM and QoC to glycaemic control (HbA1c at follow-up). Second, a multivariate model was then fitted to determine their relative importance. The analysis was repeated controlling for HbA1c at baseline, practice, sampling method, demographic and clinical characteristics. Third, HbA1c at baseline was removed from the model because it was strongly associated with HbA1c at follow-up. We assumed that this strong association could have covered weaker but clinically relevant associations. Continuous variables were centred at their mean value in order to minimise multicollinearity in the presence of interactions.25 Given that independent variables can be measured in different units, the estimated coefficients (β) were standardised to identify which independent variables have a greater effect on the dependent variable HbA1c at followup. By comparing the coefficients associated with SM and QoC from univariate and multivariate regressions, it was possible to examine their individual contribution to HbA1c, as well as their relative importance.

HbA1c had a skewed distribution. Therefore, bootstrapping, free from parametric assumptions, was used to derive estimates of error variance for tests of statistical significance, using 10 000 bootstrap samples of data from the original dataset.26 An β-level of 0.05 was selected for significance for all statistical tests.

Secondary analysis

As a secondary, exploratory analysis, we tested whether the effects of SM on HbA1c were different at different levels of QoC. For this analysis, we used the SDSCA and treatment intensification.



Figure 2 shows a CONSORT diagram outlining recruitment. Two hundred and seventy patients agreed to participate in the study, attended the laboratory, and completed interviews. There were 45 exclusions (mainly incomplete data from medical records to evaluate treatment intensification). Twenty two participants were lost at follow-up (10%). The final sample analysed was 204 patients.

Sample size for the analysis was 204 patients. This sample represents: (a) 0.8% of patients with diabetes; b) 51% of the intended sample; (c) 39% of approached patients;(d) 55% of patients who agreed to participate; (e) 76% of patients with laboratory and interview data; (f) and 90% of patients at baseline

Source: Reference 27

Figure 2 CONSORT diagram. Glycaemic control in type 2 diabetes. Aguascalientes, Mexico (December 2009 - April 2010) 

Demographic and clinical characteristics

Mean age was 60.8 years; 62% were female, 75% had a partner, 67% had at best primary school and 72% were not in paid employment (Table I). Median duration of diabetes was eight years, 48% were overweight, 40% were obese, 68% self-reported hypertension, 54% had depression, and 68% were prescribed more than one oral glucose lowering medication. Mean HbA1c at baseline was 7.9% and 8.3% at follow-up (SD 2.0).

Table I Demographic and clinical characteristics at baseline. Glycaemic control in type 2 diabetes. Aguascalientes, Mexico (December 2009 - April 2010) 


Mean score of DKQ-24 was 15.8 (SD 3.6); 25% of patients answered ≥80% of the questions correctly. Thirty one percent of patients were classified as having 'strong' knowledge of their medical prescription. The mean score of DSE was 7.0 (SD 1.7) and 55% of patients scored ≥7 on this scale. Over half of the participants (57%) reported good diabetes self-management behaviours.

Quality of care

GP appropriately intensified treatment in just over half of the sample (58%). Eighty-six percent of patients selfreported being seen by their usual GP 'a lot of the time', 'almost always' or 'always'. One hundred and forty one patients (69%) reported good communication with their GPs (scoring ≥4 in the Patient-Doctor Communication Scale), but only 34 patients (17%) were satisfied with their diabetes care (scoring ≥4 in the Patient Satisfaction with Diabetes Care).

Statistical modelling - multivariate including HbA1c at baseline

The multivariate linear regression model, that included all variables, explained 41% of the variation in HbA1c at follow-up (Table II). A 1% increase in HbA1c at baseline (e.g. from 7% to 8%) was significantly associated with a 0.59% increase in HbA1c at follow-up (95% CI 0.46 to 0.72, P <0.01). The 10-year increase in diabetes duration was significantly associated with 0.3% increase of HbA1c (95% CI 0.02 to 0.7, P <0.05). Combination therapy was still significantly associated with HbA1c at follow-up(0.63%,95%CI0.09to1.17,P<0.05);however, in the univariate analysis, this association was stronger (1.4% increase of HbA1c, 95% CI 0.88 to 2.01 P <0.01).

Table II Linear regressions with HbAIc at follow-up as dependent variable. Glycaemic control in type 2 diabetes. Aguascalientes, Mexico (December 2009 - April 2010) 

* p-value <0.05

‡ p-value <0.01

§ Global p-value <0.01

Statistical modelling - multivariate without HbA1c at baseline

When HbA1c at baseline was removed from the model, appropriate treatment intensification was a significant predictor (Table II). This model explained 14% of thevariation in HbA1c at follow-up. HbA1c levels were significantly lower (0.65% reduction of HbA1c) in patients receiving appropriate treatment intensification (95% CI −1.23 to −0.07, P <0.05).

A significant interaction was found between SM behaviours and treatment intensification and it is shown in Figure 3 (β = 0.21, P <0.05). Controlling for HbA1c at baseline, this interaction suggests that if treatment is not intensified appropriately, effective SM can result in reduced HbA1c. Whereas when treatment is appropriately intensified, SM behaviour appears to make little, if any, difference to HbA1c.

SDSCA poor = 0 to 2 self-management behaviours ≥4 days/week; SDSCA good = 3 or 4 self-management behaviours ≥4 days/week; inappropriate treatment intensification = GP didn't intensify treatment when necessary (patient had fasting blood glucose >130 or HbA1c >7.0) or GP intensified treatment when not necessary (patient had fasting blood glucose ≤130 or HbA1c ≤7.0); appropriate treatment intensification = GP intensified treatment when necessary (patient had fasting blood glucose >130 or HbA1c >7.0) or GP didn't intensify treatment when not necessary (patient had fasting blood glucose ≤130 or HbA1c ≤7.0)

Figure 3 Interaction between self-management behaviours (SDSCA) and treatment intensification in their association with change in HbAIc from baseline to follow-up. Aguascalientes, Mexico (December 2009 - April 2010) 


This study suggests that QoC (treatment intensification) was more important than SM as a predictor of glycaemic control at six months follow-up, at least in the context of Mexican primary care. Treatment intensification was significant only if HbA1c at baseline was not controlled for.

The effect of treatment intensification has been reported previously.9,28-31 Treatment intensification was extracted from medical records in this study. We acknowledge that medical records might have not been fully completed by GPs. This measure is also limited because it deals only with increases in dosage or the addition of more medications but does not take into account the trade-offs between risks and benefits of intensifying medications. The aim of treatment intensification in diabetes is to achieve recommended HbA1c levels, with a target level of <7% in Mexico.4 The American Diabetes Association suggests the same target except in the presence of a history of severe hypoglycaemia, limited life expectancy, advanced diabetes complications, and multiple comorbidities (recommending to achieve 'less-stringent' HbA1c <8%).3 Medical records lack the necessary information to identify patients who are not appropriate candidates (e.g. history of severe hypoglycaemia) or refuse treatment intensification. Previous studies have encountered the same limitation.9,29,32,33

The analyses for this study were undertaken on the assumption that treatment intensification in the context of raised HbA1c was generally an appropriate measure of QoC. However, it is possible that treatment intensification could be associated with reductions in quality of life among patients if it leads to additional burden or anxiety. We did not measure quality of life and were therefore unable to assess any impact on it. Other measures could have confounded the results like changes in body weight and adherence to treatment but these were not measured either.

Treatment intensification had an effect on HbA1c at follow-up when HbA1c at baseline was not included in the analysis but duration of diabetes and combination therapy remained significant. It has been suggested that glycaemic control deteriorates overtime due to pancreatic β-cell dysfunction34,35 and therefore multiple therapies are required.36 This might be why patients with longer duration of diabetes had higher HbA1c levels and were under combination therapy. Twelve of the participants were under combination therapy with three medications and above treatment targets (HbA1c >7.0%). They were potential candidates for insulin treatment.

The findings of this study showed that none of the SM predictors were associated with HbA1c at follow-up. Systematic reviews and meta-analysis of RCTs have reported variable results: significant effects of SM interventions on SM behaviours as well as on glycaemic control, effects on SM but no consequent effect on glycaemic control, and effects of SM interventions on glycaemic control without demonstrating any effects on SM.13,37-39 For our study, it might be possible that the lack of power concealed an association between SM and glycaemic control. Furthermore, social desirability might have affected participants' report of SM behaviours. Participants might have reported behaviours that they did not actually do. The National Health Survey in Mexico found that less than 30% of adults with diabetes reported having a diet or doing exercise40 which is lower than the percentage of SM behaviours reported in our study.

Secondary analyses did suggest an interaction between SM behaviours and treatment intensification. In patients who did not receive treatment intensification when indicated, greater numbers of SM behaviours predicted significantly lower HbA1c at follow-up. This interaction has not been reported previously which suggests a hypothesis to test in future research.

There are limitations in the study. Half of the participants were not included in the analysis because they did not agree to participate, did not attend blood test, or were lost to follow-up. It is possible that nonparticipants might have had different SM and QoC characteristics, but lacking data on non-respondents, we could not analyse their characteristics. Some nonparticipants reported that they did not have time to attend blood tests and interviews, and these patients may be less keen on performing SM and attending medical appointments. There were some differences between consecutive and random samples which were controlled in the multivariate model along other variables. The intended sample size was 405 participants, 270 attended laboratory and interviews but had incomplete data in their medical records; the total analysis sample was 204 patients. This sample size was still sufficient to provide 70% power to detect the required difference in correlations, assuming a correlation of 0.3 between SM and QoC. The average follow-up time was 5.8 months. It is likely that a more noticeable change would be observed with a longer follow-up period.

The approach taken in this longitudinal cohort study has significant strengths. The study measured various aspects of both SM and QoC, a comprehensive approach which has not been reported in the literature up to now. The analysis allowed statistical control for relevant covariates (demographic and clinical), and the follow-up rate of 90% was higher than the recommended acceptable rate of 80% meaning that the risk of bias due to attrition was small.41,42


Treatment intensification was the main predictor of lower HbA1c levels at follow-up, when HbA1c at baseline was not controlled for, suggesting that QoC was more important than SM as a predictor of glycaemic control. Although none of the SM predictors was significantly related to HbA1c, an exploratory interaction showed that patients who did not receive treatment intensification when they needed it and performed more SM behaviours had lower HbA1c levels at follow-up.


Thanks to the Instituto Mexicano del Seguro Social; to the participants; to Renata Isabel Galindo Andrade, who recruited and interviewed patients; to the interviewers and data extractors: América Fabiola Aguiñaga Rincón, Brenda Berenice Valenciano Martínez, Enrique Cardona, Karina Alejandra Pedroza García, José Luis Juárez Ruiz Esparza, Laura Carolina Soto Ham, and María Angélica Alonso Álvarez.


1. UK Prospective Diabetes Study (UKPDS) Group. Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). Lancet 1998;352(9131):854-865. ]

2. UK Prospective Diabetes Study (UKPDS) Group. Intensive bloodglucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 1998;352(9131):837-853. ]

3. American Diabetes Association.(6) Glycemic targets. Diabetes Care 2015;38 Suppl:S33-S40. ]

4. Diagnostico y tratamiento de la diabetes mellitus tipo 2. Mexico: Instituto Mexicano del Seguro Social, 2009. [ Links ]

5. National Collaborating Centre for Chronic Conditions. Type 2 diabetes: national clinical guideline for management in primary and secondary care (update). London: Royal College of Physicians, 2008. [ Links ]

6. Diagnóstico, metas de control ambulatorio y referencia oportuna de prediabetes y diabetes mellitus tipo 2 en el primer nivel de atencion. Mexico: Secretaría de Salud, 2013. [ Links ]

7. Si D, Bailie R, Weeramanthri T. Effectiveness of chronic care modeloriented interventions to improve quality of diabetes care: a systematic review. Prim Health Care Res Dev 2008;9:25-40. ]

8. Parchman ML, Pugh JA, Noel PH, Larme AC. Continuity of care, self-management behaviors, and glucose control in patients with type 2 diabetes. Med Care 2002;40(2):137-144. ]

9. Schmittdiel JA, Uratsu CS, Karter AJ, Heisler M, Subramanian U, Mangione CM, et al. Why don't diabetes patients achieve recommended risk factor targets? Poor adherence versus lack of treatment intensification. J Gen Intern Med 2008;23(5):588-594. ]

10. O'Connor PJ, Asche SE, Crain AL, Rush WA, Whitebird RR, Solberg LI, Sperl-Hillen JM. Is patient readiness to change a predictor of improved glycemic control? Diabetes Care 2004;27(10):2325-2329. ]

11. Bodenheimer T, Wagner EH, Grumbach K. Improving primary carefor patients with chronic illness: the chronic care model, Part 2. JAMA 2002;288(15):1909-1914. ]

12. Renders CM, Valk GD, Griffin S, Wagner EH, Eijk JT, Assendelft WJ. Interventions to improve the management of diabetes mellitus in primary care, outpatient and community settings. Cochrane Database Syst Rev 2001;(1):CD001481. [ Links ]

13. Norris SL, Engelgau MM, Narayan KM. Effectiveness of self-management training in type 2 diabetes: a systematic review of randomized controlled trials. Diabetes Care 2001;24(3):561-587. ]

14. Martinez Y, Campbell SM, Hann M, Bower P. The relationship between quality of care and self-management in patients with type 2 diabetes: a cross-sectional survey in primary care in Mexico. Qual Prim Care. Quality in Primary Care 2014;22(6):262-269. [ Links ]

15. Garcia AA, Villagomez ET, Brown SA, Kouzekanani K, Hanis CL. The Starr County Diabetes Education Study: development of the Spanish-language diabetes knowledge questionnaire. Diabetes Care 2001;24(1):16-21. ]

16. Prado-Aguilar CA, Martinez YV, Segovia-Bernal Y, Reyes-Martinez R, rias-Ulloa R. Performance of two questionnaires to measure treatment adherence in patients with Type-2 diabetes. BMC Public Health 2009;9:38. ]

17. Stanford Patient Education Research Center. Diabetes Self-Efficacy Scale. Stanford Patient Education Research Center 2009 [accessed on 2014 October 10].Available at: research/sediabetes.htmlLinks ]

18. Vincent D, McEwen MM, Pasvogel A. The validity and reliability of a Spanish version of the summary of diabetes self-care activities questionnaire. Nurs Res 2008;57(2):101-106. ]

19. Mead N, Bower P, Roland M. The General Practice Assessment Questionnaire (GPAQ) development and psychometric characteristics. BMC Fam Pract 2008;9:13. ]

20. Velazquez-Abad DE. Validez de criterio concurrente de un cuestionario que valora la comunicación en la relación medico-paciente en Diabetes Mellitu tipo 2 [dissertation]. Aguascalientes (Ags): Universidad Autonoma de Aguascalientes, 2010. [ Links ]

21. Prado-Aguilar CA. Indicadores de la calidad de atencion del medico familiar asociados a la severidad de la retinopatia en diabeticos tipo 2 [thesis]. Guadalajara (Jal): Universidad de Guadalajara-University of Washington, 2007. [ Links ]

22. Jurado S, Villegas ME, Mendez L, Rodriguez F, Loperena V, Varela R. La estandarizacion del Inventario de Depresion de Beck para los residentes de la ciudad de Mexico. Salud Ment (Mex) 1998;21(3):26-31. [ Links ]

23. Faul F, Erdfelder E, Buchner A, Lang AG. Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses. Behav Res Methods 2009;41(4):1149-1160. ]

24. Steiger JH. Tests for comparing elements of a correlation matrix. Psychol Bull 1980;87(2):245-251. ]

25. Aiken LS, West SG. Multiple regression: testing and interpreting interactions. California, USA, Sage Publications, 1991. [ Links ]

26. Efron B, Tibshirani RJ. An introduction to the bootstrap. NewYork: Chapman & Hall, 1993. [ Links ]

27. The CONSORT Flow Diagram [internet]. Ottawa: The Ottawa Hospital Research Institute, 2010 [accessed on 2014 October 10]. Available at: ]

28. Sidorenkov G, Haaijer-Ruskamp FM, Bilo H, Denig P. Review: relation between quality-of-care indicators for diabetes and patient outcomes: a systematic literature review. Med Care Res Rev 2011;68(3):263-289. ]

29. Katon W, Russo J, Lin EH, Heckbert SR, Karter AJ, Williams LH, et al. Diabetes and poor disease control: is comorbid depression associated with poor medication adherence or lack of treatment intensification? Psychosomatic Medicine 2009;71(9):965-972. ]

30. Brown JB, Nichols GA. Slow response to loss of glycemic control in type 2 diabetes mellitus. Am J Manag Care 2003;9(3):213-217. [ Links ]

31. Riddle M, Umpierrez G, DiGenio A, Zhou R, Rosenstock J. Contributions of basal and postprandial hyperglycemia over a wide range of A1C levels before and after treatment intensification in type 2 diabetes. Diabetes Care 2011;34(12):2508-2514. ]

32. FuAZ,QiuY,Davies MJ, Radican L, Engel SS. Treatment intensification in patients with type 2 diabetes who failed metformin monotherapy. Diabetes Obes Metab 2011;13(8):765-769. ]

33. Grant RW, Pabon-Nau L, Ross KM, Youatt EJ, Pandiscio JC, Park ER. Diabetes oral medication initiation and intensification: patient views compared with current treatment guidelines. Diabetes Educ 2011;37(1):78-84. ]

34. Fonseca VA. Defining and characterizing the progression of type 2 diabetes. Diabetes Care 2009;32 Suppl 2:S151-S156. ]

35. Marchetti P, Lupi R, Del GS, Bugliani M, D'Aleo V, Occhipinti M, et al. Goals of treatment for type 2 diabetes: beta-cell preservation for glycemic control. Diabetes Care 2009;32 Suppl 2:S178-S183. ]

36. Turner RC, Cull CA, Frighi V, Holman RR. Glycemic control with diet, sulfonylurea, metformin, or insulin in patients with type 2 diabetes mellitus: progressive requirement for multiple therapies (UKPDS 49). JAMA 1999;281(21):2005-2012. ]

37. Deakin T, McShane CE, Cade JE, Williams RD. Group based training for self-management strategies in people with type 2 diabetes mellitus. Cochrane Database Syst Rev 2005;(2):CD003417. ]

38. Gary TL, Genkinger JM, Guallar E, Peyrot M, Brancati FL. Meta-analysis of randomized educational and behavioral interventions in type 2 diabetes. Diabetes Educ 2003;29(3):488-501. ]

39. Sarkisian CA, Brown AF, Norris KC, Wintz RL, Mangione CM.A systematicreviewofdiabetesself-careinterventionsforolder,AfricanAmerican, or Latino adults. Diabetes Educ 2003;29(3):467-479. ]

40. Jiménez-Corona A, Aguilar-Salinas CA, Rojas-Martínez R, Hernández Ávila M. Diabetes mellitus tipo 2 y frecuencia de acciones para su prevención y control. Salud Publica Mex 2013;55 supl 2:S137-S143. [ Links ]

41. Altman DG. Statistics in medical journals: some recent trends. Stat Med 2000;19(23):3275-3289. ]

42. Kristman V, Manno M, Cote P. Loss to follow-up in cohort studies: how much is too much? Eur J Epidemiol 2004;19(8):751-760. [ Links ]

*This work was supported by The Consejo Nacional de Ciencia '/Tecnología (Conacyt) as part of a PhD scholarship [scholarship number 210903] and the Secretaria de Educación Pública (SEP).YVM was an employee of the Instituto Mexicano del Seguro Social during data collection for this study.

Received: October 12, 2015; Accepted: April 12, 2016

Corresponding author: Yolanda Martinez. University of Manchester,Williamson Building, Oxford Road, Manchester, M13 9PL, United Kingdom. E-mail:

Declaration of conflict of interests. The authors declare that they have no conflict of interests.

Creative Commons License This is an open-access article distributed under the terms of the Creative Commons Attribution License