Benefits of self-monitoring blood glucose in the management of new-onset Type 2 diabetes mellitus: The St Carlos Study, a prospective randomized clinic-based interventional study with parallel groups


Alfonso L. Calle-Pascual, Department of Endocrinology and Nutrition, 1aS, Hospital, Clínico San Carlos, c/Prof. Martin Lagos s/n, E-28040 Madrid, Spain.
Tel: +34 91 3303281
Fax: +34 91 3303117


Background:  Intensive treatment of patients with Type 2 diabetes mellitus (T2DM) from the moment of diagnosis facilitates β-cell recovery. Self-monitoring of blood glucose (SMBG)-based educational and pharmacological intervention may be better than conventional HbA1c algorithms in the treatment of newly diagnosed T2DM.

Methods:  Newly diagnosed T2DM patients were randomized to either an SMBG-based intervention or an HbA1c-based control group (n = 99 and 62, respectively) and were followed for 1 year.

Results:  Higher rates of diabetes regression (HbA1c < 6% on metformin alone) and remission (HbA1c between 6.0% and 6.4%) were achieved in the intervention compared with the control group (39% vs 5% (< 0.001) and 37% vs 30% (< 0.01), respectively). Furthermore, significantly greater reductions in median HbA1c (6.6% to 6.1%; < 0.05) and body mass index (29.6–27.9 kg/m2; < 0.001) were seen in the intervention over the 1 year of therapy. The percentage of patients achieving a lifestyle score >12 was significantly greater in the SMBG compared with the control group (38.4% vs 9.7% respectively; < 0.001). An inverse correlation was observed between SMBG and HbA1c levels (< 0.04).

Conclusions:  The results indicate that SMBG-based structured educational and pharmacological programs empower patients to achieve nutritional and physical activity goals, and encourage physicians and patients to use SMBG to optimize therapy. We believe that the concept of intensive treatment of T2DM patients should be modified; instead of referring to the type of treatment (insulin use), the term should reflect the intensity with which we work to reach glucose objectives.


Intensive therapy of patients with Type 2 diabetes mellitus (T2DM) facilitates β-cell recovery.1,2 Furthermore, both the Steno-2 and UKPDS (UK Prospective Diabetes Study) studies have reported that patients who attained near-normal glycemic levels from the moment T2DM was detected presented lower long-term cardiovascular mortality than patients with worse initial control, possibly due to metabolic memory.3–5

Different therapeutic algorithms can be used to manage T2DM,6,7 but none includes the self-monitoring of capillary blood glucose (SMBG). Currently, SMBG is not considered obligatory for patients not on insulin,8–13 perhaps because the glucose values obtained are not necessarily used to modify diet, exercise regimens, or pharmacological therapy. Information from the SMBG can be used by both the patient and the diabetes care team to improve glycemic control by altering treatment regimens.9,14 Furthermore, wide oscillations in levels of glycemia per se are an independent risk factor for diabetic complications15,16 and cannot be detected without monitoring.

We postulated that the SMBG should form an integral part of the treatment of newly diagnosed T2DM patients, enabling the patient to adapt his/her lifestyle more effectively to obtain better glycemic control (educational tool). Furthermore, in combination with simple algorithms that modify the doses of glucose-lowering medication, SMBG can prevent acute complications, such as hypoglycemia, as well as alerting the patient when specialist help and support are needed (therapeutic tool). The use of SMBG as both an educational and therapeutic tool empowers patients, permitting them to take on a more active role in disease control and to learn how to make lifestyle changes to control SMBG values.

In the present study, we report data obtained after 1 year of follow-up of new-onset T2DM patients included in an SMBG-based teaching and treatment program.


The present study was designed as a prospective randomized clinic-based interventional study with parallel groups in which all diabetic patients who attended the Endocrinology outpatients’ clinic between January 2006 and December 2007and who meet the inclusion criteria were invited to participate in the study. The inclusion criteria were: (i) newly diagnosed T2DM after two fasting glucose plasma values >125 mg/dL; (ii) age 18–80 years; (iii) <6 months from the first fasting plasma glucose value >126 mg/dL; and (iv) the absence of ketones in two-first morning urine samples. Patients were excluded from the study if they had had any fasting glucose levels >125 mg/dL in previous 12 months, if they had HbA1c levels >8% at diagnosis (because these patients are more like to start on insulin within a short period of time), or if they were unable to perform SMBG. In addition, patients were excluded from the study if they had a life-threatening disease. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Hospital San Carlos Ethics Committee.

Newly diagnosed T2DM patients who were eligible for inclusion in the study were randomly (2:1) assigned to one of two groups: (i) an intervention group that received lifestyle intervention and used the SMBG as an educational tool to adhere to lifestyle changes, as well as a therapeutic tool to apply step-by-step pharmacological treatment; or (ii) a control group who received standard treatment based on HbA1c values without SMBG. All patients were treated with 850 mg metformin (half a tablet at breakfast, nothing at lunch and another half tablet at dinner; ½–0–½). Adjustments to the randomization were made for age, body mass index (BMI), and HbA1c values. The structure of the clinical trial is shown in Fig. 1.

Figure 1.

 Structure of the clinical trial. SMBG, self-monitoring of blood glucose.

Initially, 250 eligible newly diagnosed T2DM patients attending our outpatient clinic between January 2006 and December 2007 were recruited to the study. Thirty-four patients were excluded from the study and 21 patients declined to participate. The remaining 195 patients were randomized to the SMBG group (n = 130) and to the HbA1c group (n = 65). In addition, a supervised exercise program was offered to half the patients in the SMBG group (we expected a 1:1 allocation in these subgroups) but, surprisingly, only 29 patients agreed to participate. Given the small number of SMBG patients in the exercise program and the possible influence of physical activity on the three endpoints evaluated, the patients in this subgroup were excluded from subsequent analysis. Five patients, two from the SMBG group and three from the HbA1c group, failed to complete the first year of the study. Herein, we report data obtained at the 1-year follow-up of 99 patients in the SMBG group [45 men, 54 women; median age 62 years (range 55–70 years)] and 62 patients in the HbA1c group [29 men, 33 women; median age 67 years (range 58–72 years)].


Lifestyle interventions were similar for all patients and were developed after a 2-h session for each patient individually and were reinforced at each follow-up visit. A questionnaire was developed to evaluate adherence to recommended lifestyle changes (see Appendix I). Each of the 18 items on the questionnaire were assigned a score of 1, 0 or −1, with a score of 1 indicating that the beneficial recommendation was regularly performed, −1 indicating that the beneficial recommendation had not been adopted or that patients were persisting with an unhealthy habit, and 0 indicating intermediate consumptions or exercise frequency between healthy and unhealthy adults. Different composite variables from the questionnaire were assessed, including a Physical Activity score (items 1–3), a Nutrition score (items 4–15), a Low Glycemic Index score (items 4, 5, 9, and 10), an Unsaturated Fat score (items 6–8), and an overall Lifestyle score (all items). This questionnaire is based on American Diabetes Association (ADA) evidence-based nutrition recommendations17 adapted to the Spanish population following the Diabetes Nutrition and Complications Trial (DNCT), as reported previously18–20 and validated.21 The aim was to achieve a lifestyle score >12 and/or an increase in lifestyle score >7.

SMBG group

Patients attended an additional 1-h session to learn how to perform SMBG and how to collect data. Patients’ know-how and methodology were reviewed at each visit and confounding factors that could have impacted on glycemic values were evaluated. Finger sites were tested and the accuracy and timing of the recorded SMBG values were checked. We recommended six-point profiles every 3 days, before and 2 h after breakfast, lunch, and dinner, as well as after any change in pharmacological therapy.22 Therapy based on SMBG values is shown in Fig. 2a. The trigger to initiate or change therapy was set as patients failing to achieve fasting and preprandial capillary glucose values between 70 and 110 mg/dL and/or 2-h postprandial capillary glucose values between 70 and 145 mg/dL in 60% of SMBG determinations (i.e. three of five). If fasting SMBG values were outside the target levels, metformin was titrated if tolerated. If this was insufficient, then pioglitazone was added. If glucose levels were still not within target values, basal insulin was added. If postprandial SMBG values were high, treatment with glinide, a dipeptidyl peptidase (DPP)-4 inhibitor, or sulphonlyurea was considered. If this step was insufficient, bolus insulin was initiated. Glucose objectives were considered to be mean glycemia <125 mg/dL and HbA1c < 6%. After stabilization, defined as when five complete SMBG profiles were on target in two successive visits, patients were recommended to construct at least one profile every 2 weeks if they were on metformin or metformin plus pioglitazone or at least one profile per week if they were receiving some form of treatment other than metformin and/or pioglitazone. Patients were followed-up every 2 weeks during the first 3 months to evaluate five SMBG profiles, and then every 3 months.

Figure 2.

 Algorithm for the initiation and adjustment of therapy. SMBG, self-monitoring of blood glucose; FBG, fasting blood glucose; PPBG, postprandial blood glucose; DPP-4, dipeptidyl peptidase 4.

HbA1c group

After recommendations regarding lifestyle changes, all diabetic patients were started on metformin. Treatment regimens were changed on the basis of HbA1c levels determined every 3–6 months. The target was HbA1c < 6.5%. SMBG was started when the diabetes care team considered it appropriate and always with insulin treatment. The algorithm used for this group is shown in Fig. 2. Patients were followed-up for between 3 and 6 months.

To evaluate how much disease control was interfering with patients’ lifestyles, three different scales (i.e. workplace activity, family time, and social activities), with scores ranging from 0 to 100. If disease control and therapy did not impact on a patient’s workplace activity, the work score was 100; however, if a patient could not continue working, the score was 0. If disease control and therapy did not change the amount and quality of the time patients spent with their family, the family score was 100; however, if the patient could not spend any time with his/her family, the score was 0. If disease control and therapy did not interfere with a patient’s social life [including leisure activities, such as going out to the movies, to dinner, to concerts, and on dates, holidays, trips, bar hopping (tapas) etc.], the leisure score was 100; however, if the patient no longer had a social life, the score was 0. The global satisfaction scale was determined as the sum of scores for all three scales. All scores were determined twice: once when patients were first enrolled in the study and then again 12 months later. The two scores were compared.

Severe hypoglycemic episodes, requiring assistance from a third person, were recorded.

Statistical analysis

With a final number of 65 patients in each group, the study had 80% power at 5% significance (two-sided) to detect a clinically significant difference (20%) in the primary outcomes between the SMBG and HbA1c groups. The primary outcome was to estimate the remission and regression rate of T2DM. Regression was considered yearly when patients achieved an HbA1c of <6% on metformin treatment. Remission was considered yearly when patients achieved an HbA1c between 6.0% and 6.4%. Secondary outcomes were to determine changes in HbA1c, fasting insulin, homeostasis model assessment of insulin resistance (HOMA-IR), total cholesterol [high-density lipoprotein (HDL) and low-density lipoprotein (LDL)], triglycerides, apolipoprotein B, body weight, waist circumference, blood pressure, and adherence to the suggested lifestyle changes.

Parametric, one-way analysis of variance and non-parametric Mann–Whitney and Kruskall–Wallis tests were used, as appropriate, to determine whether there were any significant differences between two or more independent groups.

Data are presented as the median or mean, with the range or 95% confidence intervals given in parentheses.


Patient characteristics at the time of study entry were similar between the two groups, with the exception of higher LDL–cholesterol levels in the SMBG compared with the HbA1c group [120 (101–138) vs 105 (77–125) mg/dL, respectively; < 0.05]. In the SMBG group, 96 of 99 patients (97%) performed a median of 251 capillary measurements (range 148–300) compared with 21 of 62 patients (33.9%) in the control group who performed a median of 68 SMBG measurements (range 26–144) during follow-up. The median number of provider visits in the control and SMBG groups was 3.9 (3.1–5.4) and 8.6 (4.9–10.1), respectively.

After 1 year of follow-up, median HbA1c levels and median BMI were significantly reduced in patients in the intervention group (from 6.6% (5.8–7%) to 6.1% (5.8–6.5%) and from 29.6 (26.2–32.8) to 27.9 (25.6–31.1) kg/m2, respectively; < 0.05 and < 0.01, respectively), but not in the control group (Table 1).

Table 1.   Changes in clinical and laboratory data from baseline to 1 year follow-up
 Control group (n = 62)SMBG group (n = 99)
Baseline1 yearBaseline1 year
  1. Data are expressed as the median, with the first–third quartiles given in parentheses. *< 0.05, **< 0.01, ***< 0.001 compared with baseline; < 0.05, ††< 0.01, †††< 0.001 compared with the control group.

  2. SMBG, self-monitoring of blood glucose; BMI, Body mass index; FP, fasting plasma; HOMA-IR, homeostasis model assessment of insulin resistance; SBP, systolic blood pressure; DBP, diastolic blood pressure; HDL-C, high-density lipoprotein–cholesterol; LDL-C, low-density lipoprotein–cholesterol.

Body weight (kg)76 (67–89) 76.5 (64–91.7) 80.5 (69–87) 76 (68.8–85.7)*
BMI (kg/m2) 28.5 (25.9–30.7) 28.5 (25.9–29.8) 29.6 (26.2–32.8) 27.9 (25.6–31.1)**††
Waist circumference (cm)
 Men103 (99–108)102 (99–109)107 (101–113)103 (97–110)**
 Women94 (88–109)90 (84–108)**98 (90–104) 93 (83–99)***
FP insulin (μIU/mL)9 (6.9–14.5) 8.2 (4.8–12.7)10 (5.7–16.3)  9.3 (5.9–13.9)
HOMA-IR  3.2 (2.5–7.2) 2.5 (1.1–4.7)**4 (2.3–5.7)  2.7 (1.6–4.2)**
SBP (mmHg)145 (132–163)138 (132–159)**141 (133–155)137 (125–150)**
DBP (mmHg)86 (75.8–88)80 (71.5–84.8)*83 (76–90) 79 (73–86)*
Total cholesterol (mg/dL)194 (153–211)164 (147–205)***196 (172–219)178 (157–200)***
HDL-C (mg/dL)53 (45–69)54 (42–66)52 (44–65) 53 (45–65)
LDL-C (mg/dL)105 (77–125) 81.5 (70–112)**120 (101–138)† 97 (81–118)***
Triglycerides (mg/dL)126 (93–166)128 (96–148)114 (85–144)95 (82–134)*
Apolipoprotein B (mg/dL)88 (77.8–109.5)83 (74.5–98)99 (84–115) 85 (73–98)**
Albumin:creatinine ratio8 (6–19)9 (5–21)6 (3–11)6 (4–15)
HbA1c (%)  6.6 (6.4–7.1)  6.6 (6.2–7.3)  6.6 (5.8–7)  6.1 (5.8–6.5)*††

The median Lifestyle score increased significantly from –1 (–2, 4) to 11 (8, 14) in the intervention group (< 0.01) and from –2 (–8, 2) to 5 (1, 8.5) in the HbA1c group (< 0.01). The difference between the two groups was significant (< 0.001). Significant differences between groups were also found for the Physical Activity score and scores for the consumption of vegetables, nuts, high-fat fish, high-fiber cereals, legumes, low-fat milk, and juices (Table 2). Furthermore, after follow-up, a higher percentage of patients in the intervention group than in the control group had increased their level of physical activity (76% vs 27%, respectively; < 0.001) and consumption of vegetables (72% vs 42%, respectively; < 0.004), nuts (68% vs 26%, respectively; < 0.0001), and high-fiber cereals (64% vs 21%, respectively; < 0.0001). Similarly, higher rates of diabetes regression and remission were found for the SMBG group (39% and 37%, respectively) than the control group (5% and 30%, respectively; < 0.001 and < 0.01; respectively; Table 3). An inverse correlation was observed between SMBG and HbA1c levels (< 0.04), whereas a positive correlation was found between SMBG and the Nutrition score (< 0.02) and Physical Activity score (< 0.07).

Table 2.   Changes in lifestyle score from baseline to 1 year follow-up
 Control group (n = 62)SMBG group (n = 99)
Baseline1 yearBaseline1 year
  1. Data are expressed as the median, with the first–third quartiles given in parentheses. *< 0.05, **< 0.01, ***< 0.001 compared with baseline; < 0.05, ††< 0.01, †††< 0.001 compared with the control group.

  2. SMBG, self-monitoring of blood glucose.

Walking time each day−1 (−1, 0)1 (0, 1)*0 (−1, 1)1 (1, 1)*††
Climbing stairs−1 (−1, −1)−1 (−1, 0)−1 (−1, 0)0 (−1, 1)*†††
At least 30 min exercise of more than moderate intensity−1 (−1, −1)−1 (−1, −1)−1 (−1, −1)1 (1, 1)***†††
Physical activity score−3 (−3, −2)−1 (−2, 0)*−2 (−3, −1)0 (−1, 2)***†††
Dietary factors
 Vegetable intake0 (0, 0)1 (0, 1)**0 (−1, 0)1 (1, 1)**†††
 Number of pieces of fruit (no juice)0 (0, 1)1 (1, 1)*1 (0, 1)1 (1, 1)
 Nuts−1 (−1, −1)−1 (−1, 0)−1 (−1, 0)1 (0, 1)***†††
 Olive oil1 (1, 1)1 (1, 1)1 (1, 1)1 (1, 1)
 High-fat fish or Iberico ham−1 (−1, 0)0 (−1, 1)*0 (−1, 0)1 (0, 1)**†††
 High-fiber bread and cereals−1 (−1, −1)−1 (−1, 0.5)−1 (−1, −1)1 (0, 1)**†††
 Legumes0 (−1, 0)0 (0, 1)*0 (−1, 0)0 (0, 1)*
 Low-fat milk and cheese−1 (−1, −1)1 (−1, 1)**−1 (−1:1)1 (1, 1)***††
 Red meat 0.5 (0, 1)1 (1, 1) 0.5 (0, 1)1 (1, 1)
 Sauces (no mayonnaise)1 (1, 1)1 (1, 1)1 (1, 1)1 (1, 1)
 Cookies−1 (−1, 0.3)1 (0, 1)**−0.5 (−1, 1)1 (1, 1)**
 Juices and sweets drinks−1 (−1, 0)1 (0, 1)**−1 (−1, 1)1 (1, 1)*
 Coffee0 (0, 0)0 (0, 0)0 (0, 0)1 (1, 1)
 Alcoholic beverages0 (0, 0)0 (0, 0.5)0 (0, 1)0 (0, 1)
 Water1 (1, 1)1 (0.5, 1)1 (1, 1)1 (1, 1)
Nutrition score−1 (−5, 0.25)5 (1, 8)**0 (−2, 3)9 (6, 11)***†††
Lifestyle score−2 (−8, 2)5 (1, 8.5)**−1 (−2, 4)11 (8, 14)**†††
Table 3.   Changes in lifestyle patterns and targets of diabetes control after 1 year in the control and self-monitoring of blood glucose groups
 Control group (n = 62)SMBG group (n = 99)P value
  1. Data show the number of patients, with percentages given in parentheses.

  2. SMBG, self-monitoring of blood glucose.

Increased climbing stairs13 (21)54 (55)0.003
Increased physical exercise4 (6)41 (41)0.002
Increased physical activity score17 (27)76 (77)0.001
Increased consumption of vegetables26 (42)71 (72)0.004
Increased consumption of nuts16 (26)68 (69)0.0001
Increased consumption of high-fiber cereals13 (21)65 (66)0.0001
Lifestyle score > 126 (10)38 (38)0.001
Increased Nutrition score >720 (32)62 (63)0.001
Increased consumption of low glycemic index carbohydrates24 (39)61 (62)0.006
Increased consumption of unsaturated fat33 (53)79 (80)0.002
HbA1c < 6% (regression)3 (5)39 (39)0.001
HbA1c 6.0%–6.4% (remission)19 (31)37 (37)0.01
HbA1c < 7%46 (74)91 (92)0.04
HbA1c > 7%16 (26)8 (8)0.01
Weight loss (>0.5 kg)23 (37)63 (64)0.03

After 1 year of follow-up, a total of 101 T2DM patients remained on metformin alone, 64 (65%) from the intervention group and 37 of 62 patients (59.7%) from the control group. Four T2DM patients (4%) from the intervention group were also on glinides, one (1%) was on sulphonylureas, seven (7%) were on pioglitazone and 23 (23%) were on insulin. In the control group, nine T2DM patients (14%) were on glinides, 13 of 62 patients (21%) were on sulphonylureas, and three (5%) were on insulin. The pharmacological changes were earlier (< 0.002) and more frequent (< 0.001) in the intervention group.

No severe hypoglycemic episodes requiring third-party or medical assistance were reported in either group.

Scores obtained using visual scales to evaluate the impact of SMBG on patients’ lives increased significantly within each group during the study period, with the final scores of all scales used significantly higher for patients in the SMBG compared with the control group (< 0.001). Specifically, the work score increased from a median (range) of 36 (28–48) to 90 (90–95) in the intervention group compared with an increase from 38 (25.5–55) to 65 (55–74.5) in the control group (both < 0.001). The family score also improved, from a median of 26 (21–30) to 92 (90–95) in the intervention group compared with an increase from 36 (27–50) to 67 (58–75.5) in the control group (both < 0.001). The leisure score in the intervention group increased from a median of 29 (27–35) to 90 (90–95) compared with and increase from 41 (28–55) to 63 (55–69) in the control group (both < 0.001). The global satisfaction scale improved in both groups from baseline, but the increase was significantly greater in the SMBG group (< 0.001). The global satisfaction score increased from 30 (28–37) to 90 (90–97) in the SMBG group and from 33 (27.5–48) to 59 (49.5–67) in the HbA1c group (both < 0.001 compared with baseline).


The results of the present study indicate that using SMBG in a step-by-step treatment program effectively improves metabolic control in newly diagnosed T2DM patients. The use of SMBG in a structured educational program results in greater adherence to nutritional recommendations and improves patient satisfaction without increasing the risk of severe hypoglycemia. Furthermore, it permits health personnel to detect the need for a change in therapy (i.e. the use of insulin) when the desired level of glucose control is not achieved, thus explaining why more patients in the SMBG group ended up on insulin than in the control group.

The program as described in the present study considered treatment changes when 60% of capillary glucose values were not within the target range (i.e. three of five) over a 2-week period. SMBG is a useful tool for the selection of the most adequate medication (targeting fasting versus postprandial blood glucose control) and doses for a given patient (therapeutic tool). Furthermore, SMBG indicates optimal dietary and exercise changes for each patient, helping diabetic patients chose food type, quantity, timing, and preparation to best ensure optimal glucose levels. Although HbA1c levels indicate the risk of vascular disease,6,7,9 they do not provide real-time information regarding hyper- or hypoglycemia and thus cannot be used to adapt therapy to oscillations in glucose levels.14–16 Data obtained with the SMBG reflect the true state of glucose control, thereby enabling short-term modifications of therapy. This improvement in glucose control prevents increases in HbA1C% levels.23–25

The reason why the SMBG was so effective in the present study is probably because it enabled timely and appropriate treatment decisions to be made on the basis of the values obtained. This is supported by the fact that treatment decisions were made on a greater number of visits for patients in the SMBG group, as well as the fact that significantly fewer patients in the HbA1c group were on insulin. Patients in the HbA1c group may not have been seen often enough to enable the collection of sufficient data to show that their treatment was no longer effective. Only when insulin was prescribed to patients in the HbA1c group was the SMBG used. The reason why SMBG was not effective in lowering HbA1c levels in patients in the HbA1c group who were not on insulin is that in these patients SMBG data were not used to make clinical decisions, only to educate patients. According to the recent International Diabetes Federation (IDF) consensus statement,24 the present study emphasized the importance of treatment decisions being made on the basis of SMBG results. We believe that ours is the first study to demonstrate that basing treatment on SMBG results is very effective.

The diabetes regression rate in patients in the SMBG group was almost 50%. The patients included in the present study had moderate hyperglycemia at the time of diagnosis and disease duration <1 year; therefore, the regression of diabetes (i.e. HbA1c <6% on metformin alone) may be expected after modifications of lifestyle alone26,27 or in combination with metformin,27 as has been reported in T2DM prevention trials.26,27

The improvement in glycemic control in the SMBG group was unexpectedly associated with weight loss, but this relationship was not evident in patients in the HbA1c group. This difference can probably be explained by the selection of lower calorie foods by patients in the SMBG group (e.g. vegetables, food with a high fiber content and a low glycemic index), increased in physical activity, and the progressive application of pharmacological treatment, thereby reducing the risk of hypoglycemia.28 Thus, according to our results, SMBG reinforces lifestyle changes that lead to weight loss.

A major issue in the cost–benefit analysis of using the SMBG in T2DM patients is its impact on quality of life. Some recent studies have reported no improvement in glucose control in these patients, yet observed higher anxiety levels and more cases of depression in those using SMBG.10–13 The satisfaction scale used in the present study has some limitations because of the subjectivity of any visual analogue scale. However, any tool, such as SMBG, directed towards making the diabetic patient more self-sufficient in the context of using glucose levels to modify diet, exercise, and medication could improve quality of life by both improving well being and increasing patient empowerment. Finding high glucose levels and not doing anything about it can be discouraging; finding high glucose levels, knowing what to do, and then observing an improvement can be encouraging. Patients in the former category are more dependent on their families and health personnel; those in the latter category become progressively more self-reliant. Our data show that SMBG induces an increment in patient satisfaction, probably because the patients know how to interpret the data obtained and what to do, including with any unexpected glycemic values.

The lifestyle in Spain is extremely variable.29 This is one of the reasons we recommended the use of SMBG throughout the study to reinforce adherence to treatment. The evaluation proposed at the beginning and after any therapeutic change is at least one profile consisting of six points each day every 3 days. We then recommend patients construct at least one profile every 2 weeks. Our results clearly show that SMBG induces an improvement in the adherence to nutritional and physical recommendations and decreases HbA1c levels and body weight without increasing the number of severe hypoglycemic events. Most diabetes associations consider diabetes education at the time of diagnosis a priority.6,7,24,25,30 Structured and continuous education that involves patients in making decisions about their treatment regimens and permits them to learn how to modify doses is specifically recommended as a means to achieve target HbA1c levels. We consider SMBG to be a basic, essential part of T2DM education and self-management. However, the use of SMBG remains suboptimal by T2DM patients.31,32 Perhaps the concept of intensive treatment of the T2DM patient should be modified, from bolo-basal insulin treatment to intensive self-monitoring of blood glucose several times a day, several days a week, together with appropriate lifestyle modifications or pharmacological treatment. What would define “intensive” would not be the specific therapy, but the intensity with which we work to reach glucose objectives. The “intensive” control of glucose levels by T2DM patients would allow them to make the necessary changes in eating habits, physical activity, and medication doses to optimize their control, thereby improving their overall health and sense of well being. Our findings support this hypothesis and, according to the recent IDF consensus statement,33 our study emphasized the importance of treatment decisions being made on the basis of data obtained by the SMBG. We think this is the first study that demonstrates the effectiveness of basing treatment on SMBG results and, consequently, we suggest that SMBG-based programs should be extended to primary care settings that routinely attend to diabetic patients.


Clinical trial number ISRCTN81672669 available at (accessed April 2009).


This work was supported by grants from the Ministerio de Sanidad from Spain (Fondos de Cohesion 2008) and the Fundación de Estudios Endocrinometabólicos. LdV was supported by a grant from the Ministerio de Sanidad of Spain.


Table AppendixI.   Lifestyle questionnaire
  1. *If diabetic patients do not drink any alcoholic beverages, drinking is not recommended; however, the patients do drink, an intake of 1–4 servings/day (between 10 and 40 g alcohol) is recommended.

  2. Water intake reflects the degree to which water is the usual beverage at meal time and with snacks, as opposed to juice, soft drinks, or low-sugar beverages, but not as a substitute for coffee, tea, or alcoholic beverages.

Physical activity
  1. Walking daily (>5 days/week)>1 hAt least 30 min<30 min
  2. Climbing stairs (no. floors/day, >5 days a week)>164–16<4
  3. At least 30 min of more than moderate intensity>3 days/week2 or 3 days/week<2 days/week
Servings per week
  4. Vegetables>126–12<6
  5. Fruits (pieces)>126–12<6
  6. Nuts>31–3<1
  7. Olive oilDaily>3 days<3 days
  8. High-fat fish or Iberico ham>31–3<1
  9. Bread and cereals (high fiber content)>63–6<3
 10. Legumes>21–2<1
 11. Low-fat milk and cheeses>63–6<3
 12. Red meat<33–6>6
 13. Sauces (except mayonnaise)<22–4>4
 14. Juices and sugar-sweetened beverages<22–4>4
 15. Cookies<22–4>4
 16. Coffee>3/day<3 
 17. Alcoholic beverages (no. servings/day)*1–40 or >4 and <6>6
 18. WaterExclusivelyIn addition to other beveragesNever