3. Results
A total of 1372 consecutive patients with type 2 DM, aged 20–90 years, were recruited for this study. In all, 64 and 422 patients were excluded due to insufficient HBP and UACR data, respectively. In addition, there were 148 patients who were newly prescribed angiotensin II receptor blocker (ARB) or angiotensin-converting-enzyme inhibitor (ACE-I), or who stopped using them during follow-up. Another 263 patients who had moderately or severely increased albuminuria were also excluded.
The final sample included 424 patients with normal or mild albuminuria (
Figure 1). Among them, during 5-year follow-up period, 74 patients developed moderately increased albuminuria and 1 patient developed severely increased albuminuria.
Patient baseline demographic and clinical characteristics are presented in
Table 1 and
Table 2. Median (interquartile range) age, duration of diabetes, BMI, and levels of total cholesterol and those of HbA1C were 64.0 (59.0–70.0) years, 9.0 (4.8–15.0) years, 23.0 (21.4–25.3) kg/m2, 191 (170–212) mg/dL, and 6.6% (6.2%–7.3%), respectively. The patients in the IH-HSBP group were older than those in the high HBP group (69.6 vs. 60.6 years,
p < 0.001). The unadjusted odds ratio (OR) with 95% confidence interval (CI) of developing diabetic nephropathy, given IH-HSBP, IH-HDBP, and high HBP, was 2.68 (1.36–5.30), 0.78 (0.21–2.81), and 1.63 (0.87–3.04), respectively (
Table 3), using normal HBP as a reference. In multivariate analyses, adjusted OR (95% CI) of developing diabetic nephropathy, given IH-HSBP, was 2.36% (1.14%–4.89%,
p = 0.02) in Model 2 and 2.39% (1.15%–4.96%,
p = 0.02) in Model 3 (
Table 3).
In subgroup analyses, an adjusted OR (95% CI) for developing nephropathy, given IH-HSBP, was 1.68 (0.66–4.27) among age > 65 years (
Table 4); meanwhile, in age < 65 years, an adjusted OR (95% CI) was 3.06% (0.63%–15.0%) (
Table 4), using normal HBP as a reference.
In subgroup analysis of SBP control, in patients with equal to or more than 135 mmHg, the adjusted odds ratio (95% CI) of IH-HSBP, using normal HBP as a reference group for the development of diabetic nephropathy, was 5.39% (1.92–18.6%) (
Table 5). In patients with <135 mmHg, the adjusted odds ratio (95% CI) of IH-HSBP was 0.71% (0.32–1.35%) (
Table 5). The odds of each adjusting factor for the development of diabetic nephropathy are presented in
Table 6.
4. Discussion
In the present study, IH-HSBP was associated with an increased risk of transition to moderate or severe albuminuria in patients with type 2 DM during a 5-year follow-up period.
The results are in line with the previous 2-year cohort study [
8]. The mechanism likely to account for the association between IH-HSBP and diabetic nephropathy risk has been described elsewhere [
23,
24,
25,
26,
27,
28,
29]. Increased arterial stiffness has been associated with the development of ISH [
30]. Further arterial aging might result in additional increase of IH-HSBP, which is a risk factor for target organ dysfunction [
31] and diabetic nephropathy [
32]. The association between proteinuria and high BP is strictly related to very high risk of cardiovascular disease in type 2 diabetes [
33,
34]. In advanced type 2 diabetic nephropathy, appropriate management is of great importance [
35]. So, we should adequately man-age home SBP. In HBP management, especially, we should clarify the association be-tween albuminuria and isolated high HSBP.
In the present study, IH-HSBP was associated with an increased risk of diabetic nephropathy; however, high HBP was not. The patients in the IH-HSBP group were older than those in the high HBP group. When arterial stiffness was compared between the IH-HSBP and High-HBP groups using baPWV measurements, there appeared to be higher arterial stiffness among patients in the IH-HSBP group than in those in the High-HBP group (
Table S1) [
16,
36]. Arterial aging in IH-HSBP may be associated with increased odds for the development of diabetic nephropathy. Similarly, the isolated high HDBP group was not associated with an increased risk of diabetic nephropathy development. Those in the isolated HDBP group were younger, had a short duration of diabetes, lower baPWV, and also lower HSBP than the isolated high HSBP group (
Table S2). For these reasons, only the isolated HSBP was associated with an increased risk of diabetic nephropathy development in this study.
The effect of IH-HSBP on the development of diabetic nephropathy, defined using estimated glomerular filtration rate (eGFR), is very important. Then, we analyzed the association between IH-HSBP and the development of diabetic nephropathy, defined using eGFR, and found that there was no relationship between them. We examined the association between changes in eGFR and the factors which were associated with IH-HSBP, including duration of diabetes or baPWV, and found no association. We assume that development of diabetic nephropathy, defined using ACR but not eGFR, was associated with pathophysiology of IH-HSBP in this study, although the precise mechanism is not unclear. Moreover, in this study, the mean (standard deviation) change in eGFR over 5 years was −0.37 (7.92) mL/min/1.732, which might be too small to properly analyze the development of diabetic nephropathy.
Initiation or discontinuation of anti-diabetic medications such as sodium glucose co-transporter (SGLT2) inhibitors may affect intra-glomerular pressure and the progression of diabetic nephropathy. However, SGLT2 inhibitors were not used at the start or during the initial 2-year follow-up period of this study. Among 424 patients with type 2 DM in the present cohort, 24 patients were newly prescribed SGLT2 inhibitors during the study period. Nevertheless, use of SGLT2 inhibitors did not affect the risk of diabetic nephropathy associated with IH-HSBP. Most present study patients were prescribed these agents for less than one year, which might have reduced their impact on outcomes of interest. Further studies are needed to examine this effect.
ISH among young-to-middle-aged Japanese people is associated with premature mortality due to cardiovascular disease [
37]. In the present study, age-stratified subgroup analysis revealed that the adjusted OR was higher among patients aged <65 years than in those aged ≥65 years. These findings were consistent with those of our previous study [
8], in that the association between IH-HSBP and diabetic nephropathy was weakened in patients ≥ 65. Among the patients with IH-HSBP ≥ 65, the progression of diabetic nephropathy was observed in 9.4% (2-year) and 9.5% (5-year). The progression of diabetic nephropathy did not increase over 3 years among the patients with IH-HSBP ≥ 65. Therefore, the association between IH-HSBP and diabetic nephropathy was weakened in patients aged ≥65. Meanwhile, subgroup analyses stratified by SBP status revealed that IH-HSBP increased the risk of diabetic nephropathy only in patients with SBP ≥ 135 mmHg. Patients with IH-HSBP may be at a lower risk if their SBP measurements meet the hypertension diagnostic criteria of less than 135 mmHg [
21]. It should be noted that patients in this group were older and more likely to take antihypertensive medications than patients with SBP < 135 mmHg (
Table S3). Patients with SBP ≥ 135 mmHg had remarkable ISH, which would be associated with arterial damage and diabetic nephropathy.
To the best of our knowledge, this is the first study to evaluate the impact of IH-HSBP on the risk of diabetic nephropathy in patients with type 2 DM over the medium to long term. The results support and strengthen previous reports. In addition, the risk of younger patients with ISH was elucidated through the 5-year follow-up period.
Nevertheless, this study has several limitations, which should be considered when interpreting its findings. First, we did not have data on salt intake, protein intake, or levels of exercise, which would be associated with the development of diabetic nephropathy [
27,
38,
39,
40,
41]. In this regard, we could not clearly identify the prognostic significance of HBP for the development of diabetic nephropathy, even in a longer study. Second, only Japanese men and women were included in the study population. Therefore, these findings might not be generalized to other ethnic groups. Third, only single baseline measurements of BP were performed. This may be potential bias. However, the association of target organ damage was confirmed by BP at baseline or during follow-up [
21]. Single BP assessments would be reliable when the addition of subsequent values does not significantly alter the results. Fourth, another important issue is the ultrasound findings on kidneys in the baseline, particularly the size of kidneys, which should be hypertrophic or enlarged before a moderately increased albuminuria development. However, these were not the ultrasound findings on kidneys. Fifth, the risk of ISH, defined by home BP on developing albuminuria in diabetic patients, was similar after the follow-up period and was prolonged for 3 years. The results were essentially similar to previous findings, and thus could not add new information for clinical science. We should at least prolong the follow-up period up to 10 years or more. Finally, a non-albuminuric phenotype has for years been reported in diabetic kidney disease (DKD) of type 2 DM [
42]. Therefore, many patients with type 2 DM, despite being normoalbuminuric if they have a GRF of <60 mil/min/1.73m
2, still have DKD. In the present study, we did not include patients with a GFR of <60 mil/min/1.73m
2. Thus, we were not able to evaluate the decline in renal function in the definition of DKD in this study. Further studies will be conducted in the future.