Académique Documents
Professionnel Documents
Culture Documents
Original article
Received 17 May 2010; received in revised form 9 August 2010; accepted 12 August 2010
Available online 14 October 2010
Abstract
Aim. – This study evaluated the profiles of patients with type 2 diabetes (T2DM) to identify sets of opinions and attitudes towards the disease
that might influence self-care behaviours.
Methods. – Altogether, 1,092 patients with T2DM, aged 45 or older from a large representative French cohort, completed a self-questionnaire
exploring their knowledge and perceptions of diabetes, its impact on various aspects of daily life and self-management practices. Canonical and
cluster analyses were used to identify sets of homogeneous ‘profiles’ of patients linking attitudes and opinions to specific disease-related behaviours
(such as changes in lifestyle, drug compliance, treatment satisfaction, impact on everyday life and weight gain).
Results. – Demographics of the T2DM study population were previously reported along with the main results (60% male; mean age: 66 years;
mean age at diagnosis: 55 years; mean BMI: 29 kg/m2 ). Five distinct patient types emerged from the typological approach: ‘committed’ (25%);
‘carefree’ (23%); ‘bitter’ (19%); ‘disheartened’ (19%); and ‘overwhelmed’ (15%). Each patient type defined a set of attitudes and beliefs towards
T2DM that influenced disease-related behaviours, leading to different degrees of diabetes self-management.
Conclusion. – The DIABASIS survey provides important information for diabetes care by identifying distinct patients’ profiles that express
different degrees of difficulty in implementing self-management. For this reason, patients in each category require different kinds of customized
support from their physician to induce behavioural changes that may be key in improving their metabolic control.
© 2010 Elsevier Masson SAS. All rights reserved.
Résumé
Comment les opinions et les attitudes des patients vis-à-vis de leur diabète de type 2 influencent leur implication dans sa prise en charge ? Les
leçons de Diabasis.
Objectif. – Décrire des profils de patients diabétiques de type 2 (DT2) afin d’identifier leurs opinions et attitudes vis-à-vis du diabète susceptibles
d’influencer leur engagement dans la prise en charge de la maladie.
Méthodes. – Mille quatre-vingt-douze patients DT2 âgés de 45 ans ou plus, issus d’un large échantillon représentatif français, ont répondu à un
autoquestionnaire explorant leurs connaissances et leurs perceptions vis-à-vis du diabète, son impact sur différents aspects de leur vie quotidienne
et leur niveau d’implication dans sa prise en charge. Des analyses multivariées explicatives de type canonique ont été utilisées pour identifier des
profils homogènes de patients liant certaines attitudes et opinions sur leur diabète à des comportements spécifiques (modification du mode de vie,
observance thérapeutique, satisfaction du traitement, retentissement sur la vie quotidienne et prise de poids).
1262-3636/$ – see front matter © 2010 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.diabet.2010.08.004
H. Mosnier-Pudar et al. / Diabetes & Metabolism 36 (2010) 476–483 477
Résultats. – Les caractéristiques démographiques de la population ont été rapportées avec les résultats princeps de l’étude. L’approche typologique
a permis d’identifier cinq types distincts de profils patients : les « patients actifs et impliqués » (25 %), les « insouciants » (23 %), les « aigris » (19 %),
les « découragés » (19 %) et les « accablés » (15 %). Chaque profil est défini par ses croyances et/ou attitudes vis-à-vis du DT2 qui conditionnent
les comportements et possibilités d’implication dans la prise en charge de la maladie.
Conclusions. – Diabasis fournit des informations importantes pour la prise en charge du DT2. L’étude permet d’identifier des profils types de
patients pour qui la vision de la maladie et les difficultés de prise en charge sont très différentes. Chaque profil requiert une approche personnalisée,
afin de faciliter des changements comportementaux adaptés pouvant concourir à l’amélioration du contrôle métabolique.
© 2010 Elsevier Masson SAS. Tous droits réservés.
Fig. 1. The five patients’ profiles positioned along the two axes generated by canonical analysis to identify homogeneous sets of attitudes correlating to specific
behaviours in type 2 diabetes.
The patients’ main demographic and clinical characteristics for and is more committed to following lifestyle advice. These
each of the five types are presented in Table 2. The percentages patients care about what they eat: they have stopped snack-
refer to the responses (to the self-questionnaire) by the patients ing; reduced the overall quantity of food they eat, especially
in one specific profile vs. those of the whole DIABASIS survey sugary and fatty foods; and organize their own food-shopping,
population (where P < 0.05). social life and restaurant invitations. Over half have started
physical activities (53% vs. 30% overall), and the major-
3.2.1. Committed patients (25%) ity declares having no difficulty in exercising (60% vs. 38%
This group of patients makes self-management a priority; overall).
they take action. They believe that their previous dietary habits Their physicians are viewed as partners in their treatment,
and lack of exercise contributed to the onset of their diabetes. and they also describe their close family members and friends as
They express guilty feelings and tend to link the disease to supportive. They declare being more compliant with drugs (90%
their previous way of life with self-blame (37% vs. 22% over- vs. 83% overall), and report fewer adverse drug effects (48% vs.
all). These patients used to enjoy indulging in good food and 61% overall) and less diabetes-related distress in general.
are former bons vivants, appreciating the good things in life, As a result, these patients do not perceive the disease as
and believe they bear the primary responsibility for their dis- a burden in their everyday life, and cope well with personal
ease. (sentimental and sexual life), family, professional and social
This group has the best knowledge of the disease (63% adjustments. They actively manage their disease and are more
vs. 57% overall knew what the HbA1c was used for), compliant than average.
Table 2
Main demographic and disease characteristics of patients according to the five profiles.
Committed (25%) Carefree (23%) Bitter (19%) Disheartened (19%) Overwhelmed (15%)
3.2.2. Carefree patients (23%) These patients see diabetes self-management as an unde-
These individuals are not deeply concerned about having dia- served burden that intrudes on their daily life. Even when they
betes. At disease onset, only a few expressed anxiety (16% recognize the risk of complications, they still have major diffi-
vs. 43% overall), while 36% (vs. 20% overall) were indiffer- culties in changing their lifestyles and adhering to treatment.
ent. They see themselves as being in good health in general (in
87% of cases vs. 74% overall), and only 37% (vs. 51% overall) 3.2.4. Disheartened patients (19%)
consider the disease serious. There are slightly more women than men in this group (47%
There are slightly more men than women in this group (67% vs. 40% overall) and they have little control over the daily routine
vs. 60% overall). They do not believe they have to modify their of living with diabetes. Having low motivation, they also have
lifestyle because they take their medications. This means that, more difficulties in controlling their diet (constantly swinging
while drug compliance is good (90%), self-care is minimal in between periods of restriction and overindulgence). Some are
this group: 94% have not changed their eating habits (vs. 50% obsessed by diet (18% vs. 12% overall), but are easily discour-
overall) and 89% have not changed their exercise habits at all. On aged; they also frequently complain of weight gain (45% vs.
the other hand, they show higher-than-average drug compliance 23% overall) and are more likely to be obese (42% vs. 37%
(90% vs. 83% overall), and report fewer adverse events (47% overall).
vs. 61% overall). Because of their difficulties in implementing long-lasting,
They do not involve their family in their disease manage- sustained changes in lifestyle, they simply give up. This
ment and believe that diabetes does not have an impact on might explain why they declare a low impact of the dis-
their family or social daily life. Indeed, diabetes is not a bur- ease on their daily habits, socializing and, especially, their
den for these patients and does not disturb their day-to-day professional lives (94% vs. 44% overall) despite complain-
(personal, family, professional or sentimental) life. In gen- ing more frequently of drug-related adverse events (69% vs.
eral, they minimize their situation; the disease is taken care 61% overall). Another reason why this group does not feel
of by the drug and does not further interfere with their lives. that diabetes overly impacts their lives could be the limited
However, it is possible that they are not truly more compliant self-care practices that they implement. The obstacles faced
than average, but only that their denial of non-compliance is by these patients, such as overweight and low motivation, are
greater. key considerations that need to be taken into account in their
management.
ily helps them to modify their eating habits, and encourages their drug compliance is better than the other patient types. This
their compliance with treatment and appointments with health- is not surprising as they are convinced that the drug regimen is
care professionals. More of these patients visit a diabetologist the answer to their disease; it also leads to a lack of commitment
(39% vs. 29% overall). to lifestyle changes. It ascribes to a different illness model than
This group of patients has a high level of dissatisfaction, the other groups, who have a more holistic, psychosocial model
experiences considerable difficulty in managing their diabetes [13].
and needs more active support to cope with their burden. The carefree group also differs from the other groups in
family support. As diabetes does not impact their daily life,
4. Discussion and conclusion these patients do not involve their close family in the man-
agement of the disease and, consequently, do not need family
The present analysis of the DIABASIS study provides insight support. This is in contrast to the bitter and overwhelmed groups,
into T2DM patient self-management by identifying five distinct for whom family support is more prominent. For these patient
patient types (described above). Each type of patient constitutes types, the characteristics of the patients’ family setting should
a set of opinions and attitudes towards the disease that influences be taken into account to ensure the effectiveness of interven-
their disease-related behaviours, leading to different degrees of tions, as these characteristics could significantly influence the
commitment to diabetes self-management. These findings could patient’s self-care behaviours [14]. The literature indicates that
assist the HCP in understanding what motivates or inhibits a the individuals in a patient’s support network can also influence
given patient type, and could be the key to helping HCPs adapt the behavioural response to illness [15]. It appears that patients
their approach to break the vicious circle of lack of control that and their ‘significant others’ generally share the same percep-
leads to discouraging self-management. tion of the disease, although the partners’ influence on patients’
The most apparent patients’ differences are in their clini- behaviour is minimal and typically limited to supporting changes
cal characteristics – namely, body weight and medical treatment. in lifestyle and, particularly, physical activity. This could explain
Disheartened and overwhelmed patients are more overweight, why, in our surveyed population, greater family support was not
and bitter and overwhelmed patients are more frequently treated clearly associated with a higher commitment to diet and exercise
with insulin. In addition, overwhelmed patients appear to have guidelines.
feelings of depression more frequently and are overall less sat- Changes in self-management behaviours (dietary changes,
isfied with treatment. Previous studies have already associated regular exercise) impact both the patient and his social milieu
severity of depression with poorer adherence to dietary guide- [15], so it is not surprising that patients who perceive diabetes
lines and medication regimens [6,7], and higher BMI scores as a burden on their social life have greater difficulties in imple-
[8]. This could explain the difficulties faced by this group with menting new behaviours. The overwhelmed group, in particular,
self-management. One aspect of the depression these patients finds that their professional life is especially affected by the dis-
experience has been linked to side-effects of diabetes medica- ease, and these patients are likely to find it particularly difficult to
tions [7]. The main side-effect with significant implications for maintain lifestyle changes when professional demands increase,
diabetes management is hypoglycaemia, reported significantly as has been found in other studies [16]. It is also highly likely
more often in the overwhelmed group (61%). Hypoglycaemia that the fear of hypoglycaemia experienced by the overwhelmed
and fear of hypoglycaemia have both been shown to negatively group extends beyond the patient and has an impact on their
impact diabetes management, glucose control and subsequent family, friends and work colleagues.
health outcomes [9,10]. Diabetic patients who fear hypogly- Patients who find it difficult to adhere to treatment and who
caemic episodes may go to great lengths to maintain blood experience numerous difficulties in implementing a healthy
glucose at levels in a higher-than-normal range just to avoid lifestyle turn to the healthcare system for support, implying
such events [10,11]. greater services use. The role of the HCP is to encourage
Differences in the degree to which patients believe they can and motivate patients to take their disease management seri-
alter the disease course through playing an active role in their ously. This means that the relationship that develops between
own care is probably key and, indeed, committed patients tend the patient and HCP, and the level of satisfaction with the ser-
to be less distressed over their disease and more able to maintain vices received, have an influence on the patient’s perception
lifestyle changes. In contrast, bitter patients, who did not accept of the disease and the role he/she plays in its management
responsibility for their disease – believing that family history [17–19]. Committed patients tend to be satisfied with their rela-
and genetics are the only causes of diabetes – are not as good tionship with their physicians, and see them as partners rather
at implementing lifestyle changes and are unlikely to take an than decision-makers. Ultimately, however, treatment decisions
active role in their disease management. The belief that ascribes are up to the patient, who therefore needs to receive meaning-
variable importance to lifestyle vs. genetic factors in disease cau- ful information in a way that allows him to make his own mind
sation and the patients’ own commitment to diabetes care have up. This approach leads to better adherence to lifestyle changes
already been acknowledged by others [3,10,12]. The carefree than when instructions are given in a controlling or authori-
group differs from the others in that their approach is mainly tarian manner. This idea of supporting patients’ autonomy was
biomedical – they believe that disease management is simply also reported by Williams et al. [20], who found better gly-
about taking a drug. They are relatively unconcerned about their caemic control in patients who perceived their HCPs in this
disease and make no effort to change their behaviours, whereas light.
482 H. Mosnier-Pudar et al. / Diabetes & Metabolism 36 (2010) 476–483
While the present study did not explore the influence of this reason, it would be useful to repeat it over time to see if
patients’ attitudes on glycaemic control, others have found a and how self-management in these patient categories changes.
correlation [2,3,21]. In Sweden, 259 participants were recruited Finally, symptoms and co-morbidities, which can also influence
from T2DM patients attending an educational programme deliv- patients’ self-management commitments, were not investigated
ered by pharmacists [16]. A questionnaire designed to explore [2,3,23] and, in particular, there were no data on complications,
the patients’ attitudes towards self-management was adminis- which may also have a significant impact on patients’ attitudes.
tered at the end of the program. HbA1c was measured in these In addition, the interpretation that individuals apply to their own
patients on four occasions throughout the programme, and three physical changes and the influence that will have on their self-
categories or self-management profiles were identified: ‘disease management practices need to be further explored. Finally, a
managers’ (55%); ‘compliant’ (31%); and ‘disheartened’ (10%). limitation inherent to such typological analyses is that they are
Although these patients did not differ initially in metabolic con- essentially retrospective in nature and so, when the HCP first
trol, by the end of the programme, the disease managers had sees a patient, there is no way to know into which category the
achieved significantly better and more sustained metabolic con- particular patient belongs.
trol than the compliant who, in turn, did better than the smallest Nevertheless, this large-scale survey provides further insights
group, the disheartened, who failed to achieve any reduction in into attitudes and beliefs towards T2DM, and identifies five
HbA1c [16]. Another study in which 44 T2DM patients were patient profiles that express different levels of difficulty in imple-
categorized according to their glycaemic control (from poor, menting self-management practices. Each profile is based on a
with HbA1c > 8.0%, to excellent, with HbA1c < 6.5%) found number of factors – many of which are subjective – that need
similar results [3]. The patients also participated in a one-on- to be taken into account for consistent and complete diabetes
one semistructured interview, which investigated their personal care. Patients in each category require different kinds of support
backgrounds and self-management philosophies. The final five from their HCP to promote effective self-management and the
categories included the ‘committed’ and ‘tentative’, who had behavioural changes that could be key in improving metabolic
good control; and the ‘hopeful’, ‘hassled’ and ‘overwhelmed’, control.
who had poor control [3]. These two reports confirm the correla-
tion between HbA1c levels and personal attitudes, regardless of Conflict of interest statement
whether the investigators’ starting point was based on patients’
profiles or degree of glycaemic control. One US study, conducted All of the authors, except for C. Touboul of TNS Health-
with a nationwide sample of Veterans Affairs (VA) patients with care, were members of the DIABASIS Scientific Committee.
diabetes, found that patients who were more actively engaged in Professor S. Halimi is also on the advisory panel for Novartis
their diabetes self-care had significantly lower HbA1c levels and and has received speaker’s and consultant’s fees from Novar-
also used the recommended diabetes services to a greater extent tis as well as other pharmaceutical companies (Abbott, Amgen,
[22]. Our present study appears to be in line with these findings in Astra-Zeneca, Bayer, Boehringer Ingelheim, Eli Lilly, Glaxo-
that the categories with greater difficulties in accepting and man- SmithKline, LifeScan, Merck Sharp & Dohme-Chibret, Novo
aging diabetes, such as the bitter group, tended to have poorer Nordisk, Pfizer, Roche Diagnostics, Roche Pharma, Sankyo,
metabolic control. However, due to the cross-sectional design Sanofi Aventis, Servier, Takeda and Therval). Professor P.-
of all these studies, a cause-and-effect relationship between J. Guillausseau is on the advisory panel for Novartis, and
self-management and glycaemic control cannot be firmly estab- has received speaker’s and consultant’s fees from Novartis
lished. as well as other pharmaceutical companies (Eli Lilly, Glax-
The DIABASIS survey offers important information for dia- oSmithKline, Novo Nordisk, Pfizer, Roche, Sanofi Aventis,
betes care, as a deeper understanding of patients’ perceptions of Servier, Takeda and Therval). Dr M.L. Virally is a consul-
the disease is key to optimizing customized care. Distinct types tant for Novartis. Dr E. Eschwège has received fees from
of behaviours that influence patients’ ability to make optimal Astra-Zeneca, GlaxoSmithKline, Merck-Lipha, Novartis, Novo
use of available treatments and services are discernible. How- Nordisk, Roche Pharma, Sanofi Aventis, Servier and Takeda.
ever, some methodological limitations of DIABASIS need to be Dr Helen Mosnier-Pudar has received speaker’s and consul-
acknowledged. The survey was based not on samples randomly tant’s fees from Novartis as well as other pharmaceutical
selected from the whole population, but on the permanent cohort companies (Eli Lilly, Novo Nordisk, Pfizer, Roche, Sanofi
of a polling agency (frequently used in epidemiological studies). Aventis, Takeda and LifeScan), and Dr Ghislaine Hochberg
This has to be weighed against its main advantage of feasibility: has received speaker’s fees from Novartis, among other
the low cost and simplicity of the quota method facilitates the companies. Dr Sylvie Dejager is an employee of Novartis
study of large samples with regular updates. Another limitation is Pharma.
the fact that adaptation to a chronic disease is a dynamic process,
and patients’ roles vis-a-vis their disease management fluctuate Acknowledgements
considerably over time [21]. In people whose attitudes towards
diabetes are variable, two factors – weight and age – appear The authors gratefully acknowledge the participating patients
to influence changes in their views and, in turn, their shift into and their families, the operational support of TNS Healthcare
another category. The DIABASIS survey was cross-sectional France, and the helpful discussions with and editorial comments
and captured patients’ perceptions at a single point in time. For of Felicity Neilson.
H. Mosnier-Pudar et al. / Diabetes & Metabolism 36 (2010) 476–483 483
Funding: The work was supported by Novartis Pharmaceuticals [11] Currie CJ, Morgan CL, Poole CD, Sharplin P, Lammert M, McE-
Corporation. wan P. Multivariate models of health-related utility and the fear of
hypoglycaemia in people with diabetes. Curr Med Res Opin 2006;22:
1523–34.
References [12] Parry O, Peel E, Douglas M, Lawton J. Issues of cause and control in patient
accounts of Type 2 diabetes. Health Educ Res 2006;21:97–107.
[1] Mosnier-Pudar H, Hochberg G, Eschwege E, Virally ML, Halimi S, Guil- [13] Glasgow RE, Hampson SE, Strycker LA, Ruggiero L. Personal-
lausseau PJ, et al. How do patients with type 2 diabetes perceive their model beliefs and social-environmental barriers related to diabetes
disease? Insights from the French DIABASIS survey. Diabetes Metab self-management. Diabetes Care 1997;20:556–61.
2009;35:220–7. [14] Fisher L, Chesla CA, Skaff MM, Gilliss C, Mullan JT, Bartz RJ, et al.
[2] Sarkadi A, Veg A, Rosenqvist U. The influence of participant’s self- The family and disease management in Hispanic and European-American
perceived role on metabolic outcomes in a diabetes group education patients with type 2 diabetes. Diabetes Care 2000;23:267–72.
program. Patient Educ Couns 2005;58:137–45. [15] Searle A, Norman P, Thompson R, Vedhara K. Illness representations
[3] Savoca MR, Miller CK, Quandt SA. Profiles of people with type 2 diabetes among patients with type 2 diabetes and their partners: relation-
mellitus: the extremes of glycemic control. Soc Sci Med 2004;58:2655–66. ships with self-management behaviors. J Psychosom Res 2007;63:
[4] Fisher L, Chesla CA, Skaff MA, Gilliss C, Kanter RA, Lutz CP, et al. Dis- 175–84.
ease management status: a typology of Latino and Euro-American patients [16] Veg A, Rosenqvist U, Sarkadi A. Self-management profiles and metabolic
with type 2 diabetes. Behav Med 2000;26:53–66. outcomes in type 2 diabetes. J Adv Nurs 2006;56:44–54.
[5] Kesse E, Boutron-Ruault MC, Clavel-Chapelon F. Regional dietary habits [17] Lawton J, Peel E, Parry O, Araoz G, Douglas M. Lay perceptions of type
of French women born between 1925 and 1950. Eur J Nutr 2005;44: 2 diabetes in Scotland: bringing health services back in. Soc Sci Med
285–92. 2005;60:1423–35.
[6] Ciechanowski PS, Katon WJ, Russo JE. Depression and diabetes: impact of [18] Lawton J, Parry O, Peel E, Douglas M. Diabetes service provision: a qual-
depressive symptoms on adherence, function, and costs. Arch Intern Med itative study of newly diagnosed Type 2 diabetes patients’ experiences and
2000;160:3278–85. views. Diabet Med 2005;22:1246–51.
[7] Chao J, Nau DP, Aikens JE, Taylor SD. The mediating role of health [19] Cooper HC, Booth K, Gill G. Patients’ perspectives on diabetes health care
beliefs in the relationship between depressive symptoms and medica- education. Health Educ Res 2003;18:191–206.
tion adherence in persons with diabetes. Res Social Adm Pharm 2005;1: [20] Williams GC, Freedman ZR, Deci EL. Supporting autonomy to motivate
508–25. patients with diabetes for glucose control. Diabetes Care 1998;21:1644–51.
[8] Sacco WP, Wells KJ, Vaughan CA, Friedman A, Perez S, Matthew R. [21] Veg A, Rosenqvist U, Sarkadi A. Variation of patients’ views on Type 2
Depression in adults with type 2 diabetes: the role of adherence, body mass diabetes management over time. Diabet Med 2007;24:408–14.
index, and self-efficacy. Health Psychol 2005;24:630–4. [22] Heisler M, Smith DM, Hayward RA, Krein SL, Kerr EA. Diabetes Care.
[9] Leiter LA. From hyperglycemia to the risk of cardiovascular disease. Rev How well do patients’ assessments of their diabetes self-management cor-
Cardiovasc Med 2006; 7 (Suppl. 2):S3–9. relate with actual glycemic control and receipt of recommended diabetes
[10] Wild D, von Maltzahn R, Brohan E, Christensen T, Clauson P, Gonder- services? 2003;26:738–43.
Frederick L. A critical review of the literature on fear of hypoglycemia [23] Lai WA, Chie WC, Lew-Ting CY. How diabetic patients’ ideas of illness
in diabetes: Implications for diabetes management and patient education. course affect non-adherent behaviour: a qualitative study. Br J Gen Pract
Patient Educ Couns 2007;68:10–5. 2007;57:296–302.