Usefulness of Home Blood Pressure Measurement in the Morning in Type 2 Diabetic Patients

We studied 170 type 2 diabetic patients who visited our clinic regularly. The diagnosis of type 2 diabetes was based on the World Health Organization (WHO) criteria (6). Of 170 patients, 153 (90%) were treated with oral hypoglycemic drugs and/or insulin regimens for diabetes, whereas 80 (47%) were treated with antihypertensive drugs at the beginning of the study.

CBP was measured once during each clinic visit. HBP was measured once each morning, in the sitting position within 10 min after awakening, for 1 month. For CBP measurements, when patients with diabetes had systolic blood pressure (SBP) ≥130 mmHg and/or diastolic blood pressure (DBP) ≥85 mmHg, we classified these patients as having clinic hypertension (CH) by the criteria of the World Health Organization (WHO) and International Society of Hypertension guidelines (7). When the mean HBP levels in the morning were ≥130 mmHg SBP and/or ≥85 mmHg DBP, we classified these patients as having morning hypertension (MH). When these values were <130 mmHg SBP and <85 mmHg DBP, we classified these patients as having clinic normotension (CN) or morning normotension (MN), respectively.

The microvascular complications detected in this study were nephropathy and retinopathy. Occurrence of nephropathy was evaluated at the beginning of the study, whereas occurrence of retinopathy was evaluated at least once within 6 months before the study. The macrovascular complications defined were coronary heart disease (CHD) and cerebrovascular diseases (CVDs). Prevalence of these events was confirmed by medical history at the beginning of the study.

Glycemic control was evaluated by HbA_1c values. Other variables, including serum concentrations of lipid and creatinine, were also measured.

All subjects were divided into two groups: with CH or MH and without CH or MH.

Finally, we examined whether CBP and HBP is more predictive of these events.

CBP was measured by nurses at the clinic in the daytime in the left arm after a 5-min rest in a sitting position using an automatic device based on the cuff-oscillometric method (FT-200; Parama-Tech, Fukuoka, Japan). HBP was measured at home in the morning within 10 min after awakening, by the patient or a family member, in the left arm in a sitting position. Semiautomatic devices based on the cuff-oscillometric principle that generate a digital display of both SBP and DBP were used. All devices met the criteria set by the Association for the Advancement of Medical Instrumentation. A standard arm cuff was used to measure both CBP and HBP.

Venous samples were collected during each clinic visit and were analyzed for HbA_1c levels and concentrations of total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, and creatinine. HbA_1c levels were measured by high-performance liquid chromatography (normal range 4.5–5.7%).

Albumin concentration in random spot urine was measured by the latex agglutination photometric immunoassay method. Microalbuminuria and clinical albuminuria were defined as urinary albumin excretion ≥30 and ≥300 μg/mg creatinine, respectively (8).

All values are presented as means ± SD. Mean values were compared using unpaired Student’s t test. To compare the prevalence of microvascular and macrovascular complications in the two groups with and without hypertension, Yates’ continuity corrected χ² test with two-tailed P value was performed and odds ratios were calculated; if prevalence of the events was 0, 0.5 was added to all values before calculating the odds ratio and 95% CIs were provided. Multiple logistic analysis was used to determine the contribution of the variables to the events. Correlation between HBP and CBP levels was calculated. In addition, receiver operating characteristic (ROC) curves for HBP and CBP with various end points were used to examine whether HBP and CBP measurements behave differently in allowing ascertainment of the true risk or whether the 130/85 mmHg cutpoints are better for HBP than for CBP. This analysis was performed using the GraphPad Prism software (version 3.02; GraphPad Software, San Diego, CA) and the Statistical Package for the Biosciences (SPBS; Winesteem Institute of Community Medicine, Tokyo, Japan). P < 0.05 was considered statistically significant.

The characteristics of all subjects are shown in Table 1. The prevalences of nephropathy, retinopathy, CVD, and CHD in all subjects were 43, 26, 18, and 11%, respectively (Table 2).

Comparing the characteristics of patients with and without CH, the following trends were noted. The prevalence of CH was four times higher than CN. BMI in CH patients was slightly higher than in CN patients. In contrast, serum creatinine concentration and urinary excretion of albumin in CH patients were significantly lower (P < 0.01) than in CN patients. No significant differences in other variables were noted between the two groups. A total of 48% of CH patients were being treated with antihypertensive drugs, compared with 44% of CN patients (Table 2).

When we compared the prevalence of diabetic complications in the two groups, there were no significant differences in the prevalence of nephropathy, retinopathy, CHD, and CVD between the two groups. However, the prevalence of clinical albuminuria in CH patients was lower than in CN patients (Table 2).

The CH patients were further divided into two groups: with and without MH (Table 1). BMI in MH patients was slightly higher than in MN patients. SBP, DBP, and urinary excretion of albumin in MH patients were significantly higher (P < 0.001) than in MN patients. There were no significant differences in other variables between the two groups. Nephropathy was observed in 69% of MH patients, whereas there was no nephropathy in MN patients. The prevalences of retinopathy and CVD in MH patients were also significantly higher than in MN patients (Table 2). The prevalence of treatment with antihypertensive drugs was higher in MH than in MN (Table 2).

The CN patients were also divided into two groups: with and without MH (Table 1). The means of age, BMI, SBP, and DBP in MH patients were significantly higher (P < 0.01) than in MN patients. Serum creatinine concentration and urinary excretion of albumin were also higher in MH than in MN (Table 2). No significant differences in other variables were shown between the two groups. However, the prevalence of nephropathy in MH patients was high (91%), whereas no nephropathy was observed in MN patients. The prevalences of retinopathy, CHD, and CVD in MH patients were also higher than in MN patients. More MH patients than MN patients were being treated with antihypertensive drugs (Table 2).

Comparing the characteristics of the two patient groups with and without MH, the following trends were noted (Table 3). The means of age, sex, HbA_1c levels, and lipid concentrations were not different between the two groups. However, SBP and DBP, based on HBP in the morning, in MH patients were significantly higher (P < 0.001) than in MN patients. Serum creatinine concentration and urinary albumin excretion were also higher in MH patients (Table 3). The prevalences of treatment with antihypertensive and antidiabetic drugs were 3 and 1.5 times higher, respectively, in MH patients compared with MN patients (Table 4).

The prevalence of nephropathy in MH patients was 75%, whereas no nephropathy was noted in MN patients. The prevalence of retinopathy in MH patients was twice that found in MN patients, although there was no difference in the prevalences of nonproliferative and proliferative retinopathies between the two groups (Table 4). The prevalences of CHD and CVD in MH patients were four and six times higher, respectively, than in MN patients.

Specifically, the prevalence of nephropathy in all subjects was highly associated (P < 0.001) with systolic MH, but not with age, sex, HbA_1c, serum lipid concentrations without LDL cholesterol, and use of antidiabetic drugs by multiple logistic analysis. However, prevalence of nephropathy was associated with BMI, LDL cholesterol concentration, serum creatinine concentration, and use of antihypertensive drugs and was negatively associated with clinic DBP (Table 5).

The relationships between SBP and DBP in morning and clinic measurements were described by regression equations of morning SBP = 0.23 clinic SBP + 111 and morning DBP = 0.26 clinic DBP + 58, respectively. The correlations were poor (morning versus clinic SBP, r = 0.05, P = 0.004; morning versus clinic DBP, r = 0.09, P = 0.0001). In comparison of ROC curves for HBP and CBP with the events, the areas under ROC curves of morning SBP (0.86 ± 0.39) and morning DBP (0.70 ± 0.52) were significantly higher (P < 0.001 and P = 0.035, respectively) than those of clinic SBP (0.52 ± 0.60) and clinic DBP (0.57 ± 0.59) in nephropathy, indicating that HBP has a higher predictive value than CBP. In contrast, there were no statistical differences between them in other events. In nephropathy (Fig. 1), sensitivities of the 130-mmHg threshold in morning and clinic SBP were 1.00 (1.00–1.00 95% CI) and 0.18 (0.10–0.29), respectively, whereas those of the 85-mmHg threshold in morning and clinic DBP were 0.49 (0.37–0.61) and 0.43 (0.31–0.55), respectively. Specificities of the 130-mmHg threshold in morning and clinic SBP were 0.68 (0.58–0.77) and 0.85 (0.76–0.91), respectively, whereas those of the 85-mmHg threshold in morning and clinic DBP were 0.75 (0.65–0.83) and 0.73 (0.64–0.82), respectively.

In type 2 diabetic patients who were regularly treated with diet and exercise or medications for hyperglycemia and hypertension, we found that one half of CH patients had MN, whereas two thirds of CN patients had MH. The prevalences of nephropathy, retinopathy, CHD, and CVD in patients with MH were significantly higher than in patients without MH, even though they had CN. In contrast, the prevalence of these vascular disturbances was significantly lower in patients without MH than in patients with MH, even though they had CH. Specifically, nephropathy, including clinical albuminuria, was observed in patients with systolic MH but not in any patients without MH. The difference is not related to age, sex, BMI, HbA_1c, serum lipid concentrations, or use of antidiabetic and antihypertensive drugs. The finding in the present cross-sectional study indicates that micro- and macrovascular complications of type 2 diabetic patients may be strongly related to HBP in the morning rather than CBP.

The reason for the underlying relation of high HBP in the morning rather than high CBP to the vascular complications is not clearly determined by this study. However, several possibilities are postulated. First, type 2 diabetic patients have high prevalence of increased CBP but normal HBP in the morning (white coat hypertension) (9,10). White coat hypertension seems to be a low risk for vascular complications (11,12). Second, O’Brien et al. (13) and Imai et al. (14) reported that nocturnal decline in blood pressure in patients with essential hypertension is often diminished (nondipper hypertension) and sometimes inverts to become a nocturnal elevation (inverted dipper hypertension). Nondipper hypertension, particularly inverted dipper hypertension, accelerates vascular disturbances (15,16), including microalbuminuria (17). Many studies have reported that type 2 diabetic patients have nondipper hypertension (18–23). Therefore, it seems that blunted nocturnal and/or inverted dipper hypertension may cause micro- and macrovascular complications in type 2 diabetic patients. Third, a morning surge in blood pressure may be related to these events. A number of reports indicate that the early morning surge in blood pressure acts as a trigger for vascular events (22,23). Most diabetic patients have the morning surge (23). These phenomena in diabetic patients are considered to be caused by many neuroendocrine and hematological factors, including autonomic neuropathy (18–23), which may result in glomerular hyperfiltration, hypercoagulability, and hypofibrinolysis, promoting micro- and macrovascular disturbances (17–23). In fact, a high prevalence of these phenomena was observed in MH but not in MN. In addition, the severity of MH in CN patients tended to be greater than in CH patients. Moreover, the relation of MBP and CBP levels was a greater range, indicating true and white coat hypertension, and MBP level in some patients was higher than the corresponding CBP level, indicating that reverse dipping hypertension might occur, although we did not measure 24-h ambulatory blood pressure. It is hypothesized that treatment with antihypertensive drugs reduced daytime blood pressure but did not restore blunted nocturnal hypertension, did not decrease nocturnal hypertension, and could not attenuate the morning surge in blood pressure (21,24). The greater range in relation of MBP and CBP, and the negative association between events of nephropathy and clinic DBP may be partially explained by the effect of treatment with antihypertensive drugs, as hypothesized above.

Analysis by ROC curves also indicates that HBP has a stronger predictive power than CBP, especially in nephropathy. The cutpoints of 130/85 mmHg have higher sensitivity in morning measurement than in clinic measurement, although specificity in the cutpoint of 130 mmHg SBP in the morning measurement was lower than in the clinic measurement. Accordingly, measurement of HBP in the morning is a useful method of determining these phenomena, as indicated by the Ohasama study (2–5,14,16,24), and high HBP levels in the morning in type 2 diabetic patients may be related to micro- and macrovascular complications of diabetes.

All findings indicate that high BP levels in the morning, obtained by means of self-measurement in type 2 diabetic patients, should be treated as hypertension.

In summary, in patients with type 2 diabetes, hypertension based on self-measurement of BP in the morning after awakening is strongly related to micro- and macrovascular complications, especially nephropathy. It is concluded that control of MH may prevent vascular complications in type 2 diabetic patients.

We thank the nurses in our clinic for measurement of blood pressure in the patients. We also thank Yutaka Imai, MD (Tohoku University Graduate School of Medicine and Pharmaceutical Science) for helpful comments.