Locomotive syndrome is associated with large blood pressure variability in elderly hypertensives: the Japan Ambulatory Blood Pressure Prospective (JAMP) substudy

Abstract

Chronic pain, represented by locomotive syndrome (LS), and psychosocial factors are possible factors of blood pressure (BP) variability (BPV). The authors tested the hypothesis that there are links among LS, depression, and BPV. In 85 Japanese elderly hypertensive patients with normal daily activities, the authors performed ambulatory BP monitoring, determined the LS scale (LSS), and administered the Self‐Rating Questionnaire for Depression (SRQD). The LSS score but not the SRQD score was associated with the standard deviation (SD) and coefficient of variation (CV) of daytime systolic BP (SBP) and SD of nighttime SBP (all P<.05). Higher LSS score (in quartiles) was associated with a higher SD of daytime SBP (P=.041), even after adjusting for covariates. Regarding the components of the LSS score, movement‐related difficulty and usual care difficulty were associated with the SD and CV of daytime SBP. In elderly hypertensive patients, the LSS score was associated with exaggerated systolic BPV. The LS state could be an important determinant of systolic BPV.

1. Introduction

Hypertension is defined by an individual's blood pressure (BP) levels, and the management of hypertension is based on the average of office, home, or ambulatory BP values.1, 2, 3 However, there is growing evidence that excessive BP variability (BPV) represents a risk for organ damage and cardiovascular (CV) events.4, 5 In a large cohort study, BPV was associated with hypertensive target organ damage and CV events independent of the BP level.6 There are various types of BPV, from short‐term (beat‐by‐beat BPV and diurnal BPV), to long‐term (day‐by‐day home BPV, visit‐to‐visit office BPV, seasonal BPV, and age‐related BPV for several years) forms.7

Excessive BPV and its related organ damage are caused by the disruption of neurohumoral networks and changes in the vascular properties of small and large arteries.7 Physical and psychological stress is one of the important factors associated with BPV.8 It is well known that a sudden rise in BP, which could result in large BPV, can be caused by acute stress induced by strong pain or acute psychological stress. However, it is not yet established whether BPV can also be caused by chronic stress induced by chronic pain or by psychosocial factors, such as depression.

Physical and psychosocial problems such as the development of frailty are important concerns in older people,9 especially in aging societies such as that in Japan. Locomotive syndrome (LS) is derived from the concept of frailty,10, 11 proposed by the Japanese Orthopedic Association.12 It is a geriatric high‐risk condition that may soon require care services because of locomotive organ dysfunction. LS in the elderly is modifiable and has become a focus of attention in the field of orthopedic and geriatric medicine.13, 14

It is established that frailty in older age is associated with a two‐ to three‐fold increase in the risk of CV disease (CVD),15 but the etiology of the association between frailty and CVD is not fully understood. Depression is also frequently seen in the elderly. It was recently reported that depression and frailty often overlap,16 and it may be difficult to clinically differentiate depression from frailty in the elderly.16 Depression is also reported to be associated with short‐term BPV17 and CVD risk.18, 19

With this background, we hypothesized that exaggerated BPV due to physical and psychosocial stress is a major mechanism behind the development of CVD in elderly individuals. We thus tested our hypotheses that there are significant links among LS, depression, and large BPV in elderly hypertensive patients with normal activities of daily living (ADLs), and that frailty itself could be one of the major contributing factors for the large BPV in the elderly.

2. Methods

2.1. Study design and participants

The participants in this study were patients in the Japan Ambulatory Blood Pressure Prospective (JAMP) study.20 As a subanalysis of the JAMP study, 85 community‐dwelling residents of Munakata Oshima Island (a small and remote island in the southwestern area of Japan with a total of 780 residents) were recruited at an outpatient clinic. The entry criteria of this study were that the participants be older than 65 years, engage in normal ADLs, and have the ability to come to the clinic by themselves.

The protocol of the JAMP study is registered on the University Hospital Medical Information Network Clinical Trials Registry Web site under the trial number UMIN000020377. Briefly, the JAMP study20 is a prospective observational study to evaluate predictive values of ambulatory BP for CV events in Japanese patients with any of the following CVD risk factors: hypertension, impaired glucose tolerance or diabetes mellitus (DM), dyslipidemia, smoking habit (including chronic obstructive pulmonary disease), chronic kidney disease (CKD), atrial fibrillation, metabolic syndrome, and sleep apnea syndrome. The exclusion criteria were recent (within the prior 6 months) history of CV or cerebrovascular event, current hemodialysis treatment, chronic inflammatory disease, and malignancy.

The participants in this substudy were enrolled from April 2011 to December 2013. The institutional review board of the Jichi Medical University School of Medicine approved this study, and written informed consent was obtained from all participants. In the present study, DM was defined as having a fasting glucose level ≥126 mg/dL and/or a glycated hemoglobin level ≥6.5% (in National Glycohemoglobin Standardization Program units) or being treated with one or more antidiabetic medications. Hyperlipidemia was defined as having a total cholesterol level ≥240 mg/dL or being treated with one or more antihyperlipidemic medications. A history of CVD was defined as having had one or more of the following: angina pectoris, myocardial infarction, heart failure, aortic dissection, or stroke. Each participant's estimated glomerular filtration rate (eGFR) was calculated using the following equation21 eGFR (mL/minute/1.73 m²)=194×age (years)^−0.287×serum creatinine (mg/dL)^−1.094 (for women, ×0.739). CKD was defined as eGFR <60 mL/minute/1.73 m² and/or the presence of proteinuria. The body mass index (BMI) was calculated as weight (kg)/height² (in m²). Current smoking was defined as smoking at the time of study enrollment or within the prior year.

2.2. BP and HR measurements

Office BP and heart rate (HR) were measured at the day of study enrollment by a digital oscillometric BP monitoring device (HEM‐5001; Omron Healthcare, Kyoto, Japan) after the participant had been seated for 5 minutes. Office BP and HR were calculated as the mean of three consecutive measurements. The visit‐to‐visit variability (VVV) of the BP and HR values was evaluated using the patients’ medical records as the standard deviation (SD) and the coefficient of variance (CV) of the BP and HR values obtained at three office visits.

Noninvasive ambulatory BP monitoring (ABPM) was carried out with a validated automatic device (TM‐2433; A&D, Tokyo, Japan) that recorded the participant's BP using an oscillometric method together with the participant's HR every 30 minutes for 24 hours. Nighttime BP and HR were defined as the mean BP and HR for the period from bedtime to rising, and daytime BP and HR were defined as the mean BP and HR for the rest of the day. We used mean BP and HR levels in the daytime, nighttime, and over the 24‐hour day. BPV and HR variability were estimated by the SD and the CV of BP or HR in the respective time period.

Hypertension was diagnosed when the office systolic BP (SBP) was ≥140 mm Hg and/or diastolic BP (DBP) was ≥90 mm Hg on at least two separate occasions, based on the Japanese Society of Hypertension 2014 guideline3 or by a previous diagnosis of hypertension with current antihypertensive medication use.

2.3. LS scale

The Geriatric Locomotive Function Scale is a questionnaire developed for the early detection of LS by the Japanese Orthopedic Association12, 14 (Table S1). This LS scale is a self‐administered, relatively comprehensive measure. It covers a wide range of issues, from pain to quality of life. It consists of 25 questions, including four questions related to body pain, three questions about movement‐related difficulty, five questions concerning usual care difficulty, two questions related to cognitive function, four questions about social activities (social participation), and seven questions showing a significant association with the other components, which suggests that those seven questions can be used as a quick version of the LS scale.14 These 25 questions are graded with five‐point scales from no impairment (zero points) to severe impairment (four points), and then the points are arithmetically added to produce a total score (minimum 0, maximum 100 points). Higher scores were thus associated with worse locomotive function, and participants with LS scale scores ≥16 points were suspected to have the LS. The reliability and validity of this questionnaire have been described.14

2.4. Self‐Rating Questionnaire For Depression

The Self‐Rating Questionnaire for Depression (SRQD) is a screening tool to detect depression.22 It consists of 18 questions including six dummy questions. These 18 questions are graded on four‐point scales from no symptoms (zero points) to severe symptoms (three points), and then the points are arithmetically added to produce a total score (minimum 0, maximum 36 points). Higher scores are thus associated with depression, and individuals with an SRQD score ≥16 points are considered to be experiencing depression. The reliability and validity of this questionnaire have been reported.22

2.5. Statistical analysis

All statistical analyses were carried out with SPSS/Windows software, version 22.0 (SPSS Inc, Armonk, NY, USA). The data are expressed as the mean (±SD; continuous variables) or percentages (categorical variables). We used Pearson's correlation coefficients to calculate the correlations between parameters. We then conducted a one‐way analysis of variance (ANOVA) and an analysis of covariance to determine whether there were any significant differences between the means of scores on the LS scale and the SRQD (in quartiles). To assess the independent predictive utility of the different factors associated with the SD and CV of SBP, we performed multiple regression analyses using the forced entry method. For the independent variables, we used age, sex, BMI, office SBP, office pulse pressure, current smoking, presence of DM, presence of dyslipidemia, presence of CKD, presence of obstructive sleep apnea syndrome (OSAS), use of antihypertensive medication, number of antihypertensive medication classes, use of calcium channel blockers, use of renin‐angiotensin system blockers, use of α‐blockers, timing of the antihypertensive medication dosing, LS scale scores, and SRQD scores (in quartiles). We used SD‐SBP and CV‐SBP as dependent variables.

To determine the impact of the timing of the antihypertensive medication dosing on systolic BPV (SBPV), we performed ANOVA and Tukey analysis between groups by different timings: no medications, morning dosing, evening dosing, and morning and evening dosing. To determine whether the number of family members has an impact on LS or the SRQD scale, we then compared the LS and SRQD scale scores between the participants living alone and those who were living with family members, using the Mann‐Whitney U test. To determine whether there is a difference in LS or SRQD scale scores by number of family members, we also performed ANOVA and Tukey analysis between numbers of family members (the maximum number of family members in this study was four). P values <.05 were considered significant.

3. Results

A total of 85 hypertensive participants were enrolled in this study. The characteristics of the participants are summarized in Table 1. The mean age of the participants was 79.2±5.9 years, and 53 (62.4%) were female. The results of the ABPM parameters are summarized in Table S2.

Table 1.

Characteristics of study participants (n=85)

Age, y	79.2±5.9
Female, No. (%)	53 (62.4)
Body mass index, kg/m2	23.6±3.5
Current smoking, No. (%)	1 (1.2)
Office systolic blood pressure, mm Hg	154±19
Office diastolic blood pressure, mm Hg	75±10
Office heart rate, beats per min	70±10
Hypertension, No. (%)	81 (95.3)
History of ischemic heart disease, No. (%)	10 (11.8)
History of stroke, No. (%)	8 (9.4)
Diabetes mellitus, No. (%)	11 (12.9)
Dyslipidemia, No. (%)	51 (60.0)
Chronic kidney disease, No. (%)	20 (23.5)
Obstructive sleep apnea syndrome, No. (%)	1 (1.2)
Antihypertensive medications, No. (%)a	68 (84.0)
No. of antihypertensive medication classesb	1.5±1.1
Calcium channel blockers, No. (%)c	55 (80.9)
Angiotensin II receptor blockers, No. (%)c	41 (60.3)
Angiotensin‐converting enzyme inhibitors, No. (%)c	6 (8.8)
Diuretics, No. (%)c	14 (20.6)
α‐Blockers, No. (%)c	7 (10.3)
β‐Blockers, No. (%)c	13 (19.1)
Morning dosing of antihypertensive medication, No. (%)c	31 (45, 6)
Evening dosing of antihypertensive medication, No. (%)c	5 (7.4)
Morning and evening dosing of antihypertensive medication, No. (%)c	32 (47.1)
Regular users of nonsteroidal anti‐inflammatory drugs, No. (%)	11 (12.9)

Data are expressed as number (percentage) or mean±standard deviation. ^aThe percentages in hypertensive participants. ^bMean±standard deviation of the participants taking antihypertensive medications. ^cThe percentage of participants among those taking antihypertension medications.

The univariate analyses for the associations between LS and SRQD with ABPM parameters are shown in Table 2. The LS scale score was associated with 24‐hour SD‐SBP, 24‐hour CV‐SBP, daytime SD‐SBP, daytime CV‐SBP, and nighttime SD‐SBP. Neither the LS scale score nor the SRQD scale score was associated with any of the DBP and HR parameters (Table S3).

Table 2.

Associations between LS scale, SRQD, and ABPM parameters (SBP)

	24‐h SBP	24‐h SD‐SBP	24‐h CV‐SBP	Daytime SBP	Daytime SD‐SBP	Daytime CV‐SBP	Nighttime SBP	Nighttime SD‐SBP	Nighttime CV‐SBP
LS scale	.19	.30a	.26b	.20	.28b	.23b	.13	.22b	.18
SRQD	−.09	.13	.18	−.10	.15	.20	−.02	.002	.02

n=85. Data are expressed as correlation coefficients. Abbreviations: ABPM, ambulatory blood pressure monitoring; CV, coefficient of variance; LS, locomotive syndrome; SBP, systolic blood pressure; SD, standard deviation; SRQD, Self‐rating Questionnaire for Depression.

^a P<.01. ^b P<.05.

We next divided the participants into quartiles according to their scores on the LS and SRQD scales, and we compared BPV parameters among the four quartile groups. There was a significant trend between the LS scale score and daytime SD‐SBP (P=.041), in which daytime SD‐SBP in the highest quartile group (the fourth quartile) was significantly higher than that in the lowest quartile group (the first quartile, P=.025; Figure). The relationship remained significant (P=.035) even after adjusting for age, sex, and BMI. However, with regard to SRQD scores, there were no significant differences in any of the ABPM parameters among the four groups. There were no significant associations with the scores of LS scale or SRQD in quartiles with other ABPM parameters (Figures S1‐S3).

Figure 1.

We divided the participants according to quartiles of locomotive syndrome (LS) scale scores and Self‐Rating Questionnaire for Depression (SRQD) scores. LS scale (positive: ≥16/100 points). Q1: first quartile (0–6), Q2: second quartile (7–13), Q3: third quartile (14–30), and Q4: fourth quartile (31–100). SRQD (positive: ≥16/54 points). Q1: first quartile (0–2), Q2: second quartile (3–4), Q3: third quartile (5–6), and Q4: fourth quartile (7–54). SBP indicates systolic blood pressure; SD, standard deviation.

We then performed multiple regression analyses to assess the factors associated with SBPV parameters (Table 3). Female sex (β=−.40, P=.001), office SBP (β=.64, P=.01), office pulse pressure (β=−.89, P=.001), and LS scale score (β=.29, P=.04) were all independently associated with daytime SD‐SBP. Female sex (β=−.32, P=.01) and office pulse pressure (β=−.85, P=.002) were associated with daytime CV‐SBP. Only the LS scale score was independently associated with nighttime SD‐SBP or CV‐SBP (both β=.35, P=.04). The LS scale score remained significant when the regular use of nonsteroidal anti‐inflammatory drugs (NSAIDs) were added in the same model.

Table 3.

Factors associated with SBP variability

	Daytime SD‐SBP		Daytime CV‐SBP		Nighttime SD‐SBP		Nighttime CV‐SBP
	β	P Value	β	P Value	β	P Value	β	P Value
Age, y	.19	.14	.23	.09	−.02	.92	−.08	.61
Sex (male 1, female 0)	−.40	.001	−.32	.01	−.17	.24	−.15	.27
Body mass index, kg/m2	−.10	.46	−.06	.68	−.20	.22	−.20	.21
Office SBP, mm Hg	.64	.01	.45	.08	−.24	.42	−.44	.13
Office pulse pressure, mm Hg	−.89	.001	−.85	.002	.31	.31	.39	.19
Current smoking (yes=1, no=0)	−.05	.64	−.02	.88	.27	.06	.26	.06
Diabetes mellitus (yes=1, no=0)	.08	.54	.04	.76	−.07	.64	−.12	.45
Dyslipidemia (yes=1, no=0)	−.14	.22	−.14	.24	.08	.58	.11	.44
Chronic kidney disease (yes=1, no=0)	−.08	.49	−.02	.87	−.01	.95	.02	.87
Obstructive sleep apnea syndrome (yes=1, no=0)	.08	.50	.09	.49	.15	.31	.15	.30
Antihypertensive medications (yes=1, no=0)	−.13	.60	−.09	.72	.20	.52	.23	.45
Number of antihypertensive medication classes	−.11	.65	−.05	.86	.28	.35	.21	.47
Calcium channel blockers (yes=1, no=0)	−.19	.37	−.31	.17	−.34	.20	−.36	.17
Renin‐angiotensin system blockers (yes=1, no=0)	.22	.31	.11	.63	.01	.98	.05	.86
α‐Blockers (yes=1, no=0)	.21	.11	.18	.19	.02	.90	−.09	.55
Morning dosing of antihypertensive medications (yes=1, no=0)	.07	.72	.09	.67	−.01	.96	.02	.93
Evening dosing of antihypertensive medications (yes=1, no=0)	−.03	.81	.03	.86	−.05	.75	.04	.79
Locomotive syndrome scale (categorical)	.29	.04	.19	.19	.35	.04	.35	.04
SRQD scale (categorical)	−.11	.40	−.01	.97	−.19	.23	−.16	.28

n=85. Multiple regression analysis. Abbreviations: SBP, systolic blood pressure; SD, standard deviation; CV, coefficient of variance; SRQD, Self‐Rating Questionnaire for Depression

We next analyzed which components of the LS scale score were associated with the ABPM parameters. In the univariate analyses, body pain was associated with the daytime SBP level. Movement‐related difficulty was associated with daytime SD‐SBP, daytime CV‐SBP, nighttime SD‐SBP, and nighttime CV‐SBP. Usual care difficulty was associated with daytime SD‐SBP and daytime CV‐SBP. The “impaired social activities” component was associated with daytime SD‐SBP, daytime CV‐SBP, and nighttime SD‐SBP (Table S4). In the multiple regression analyses (Table 4), movement‐related difficulty (β=.24, P=.045) and usual care difficulty (β=.32, P=.007) were each independently associated with daytime SD‐SBP after adjusting for age, sex, BMI, mean office SBP, current smoking, presence of DM, presence of dyslipidemia, and SRQD scale (in quartiles). Usual care difficulty (β=.27, P=.018) was independently associated with daytime CV‐SBP. However, body pain did not show a significant association with the daytime SBP level. The impaired social activities component was not significantly associated with daytime SD‐SBP or with daytime CV‐SBP. None of the LS scale components was significantly associated with any nighttime BPV parameters in the multiple regression analysis.

Table 4.

Components of LS associated with daytime SBP variability

Independent Variables	Daytime SD‐SBP		Daytime CV‐SBP
	β	P Value	β	P Value
Body pain (categorical)	.16	.19	.12	.33
Movement‐related difficulty (categorical)	.24	.045	.20	.09
Usual care difficulty (categorical)	.32	.007	.27	.018
Cognitive impairment (categorical)	.09	.49	.05	.71
Impaired social activities (categorical)	.09	.49	.09	.49

n=85. Multiple regression analysis.

We put each of the components of locomotive syndrome (LS) as an independent variable, and adjusted by age, sex, body mass index, mean office systolic blood pressure (SBP), current smoking, presence of diabetes mellitus, presence of dyslipidemia, and Self‐Rating Questionnaire for Depression scale (in quartiles). Abbreviations: SD, standard deviation; CV, coefficient of variance.

Regarding the association between the timing of the antihypertensive medication dosing and SBPV, there was no significant difference in the distribution of SBPV parameters between the groups, although daytime SD‐SBP tended to be higher in the participants without antihypertensive medication (P=.072) compared with those taking antihypertensive mediation in the morning and evening.

Regarding the number of family members, 26 (30.6%) participants were living alone, 36 (42.4%) were living with two people, 20 (23.5%) were living with three people, and three (3.5%) were living with four people. There were no participants living with five or more people. The LS scale score (P=.89) and the SRQD score (P=.72) were not improved by having a spouse or daily help in the home. In addition, there were no significant differences in LS or SRQD scale scores by the number of family members.

We assessed whether the intraindividual BPV shown by ABPM is related to VVV (Table S5). We did not observe any significant association between VVV and BP or HR variability shown by ABPM except for the association between the SD of office DBP and daytime SD‐DBP (r=.25, P<.05).

4. Discussion

The results of the present study revealed that the LS scale score was independently associated with exaggerated daytime SBPV in the elderly hypertensive participants with normal ADLs. Among all of the components of the LS scale, movement‐related difficulty and usual care difficulty were associated with daytime SBPV parameters. This is the first study to show that frailty itself is associated with the short‐term daytime SBPV estimated by ABPM.

4.1. Association between LS and SBPV

In the present study, the LS scale score was independently associated with exaggerated SBPV in elderly hypertensive patients with normal ADLs. Two possible explanations for the association between LS and exaggerated SBPV are suggested. First, because physical and psychosocial stress is one of the important factors associated with BPV,8 physical stress due to LS could have resulted in sudden changes in BP in elderly participants, which was expressed as large SBPV. The second possibility is the effect of impaired BP control in the upright position in the elderly. The association between frailty and orthostatic hypotension is already established.23, 24, 25 It was recently reported that orthostatic hypertension occurs frequently among frail elderly individuals.26, 27 A study of acute hospitalized patients revealed that orthostatic hypertension was associated with favorable outcomes.26 However, in a general elderly population, orthostatic hypertension was associated with higher CV risk independently of sitting BP levels.27 We demonstrated that orthostatic hypertension is a new hemodynamic CV risk factor.28 Another recent study indicated that orthostatic hypertension is a predisposing factor for masked hypertension in healthy individuals.29 The present study was performed in elderly participants with normal ADLs, which may be generalized to a general elderly population. SBPV assessed by ABPM might reflect impaired BP regulation including orthostatic hypotension and orthostatic hypertension, and this may be a risk for CVD. This result may imply that LS underlies impaired BP regulation and may be a novel factor for masked hypertension in elderly individuals with normal ADLs.

In the present study, female sex was also associated with the SD and CV of daytime SBP. The female participants in this study were all postmenopausal (mean age 79.1±5.9 years, range 68–90 years.). It has been recognized that there is a sex difference in sympathetic neural‐hemodynamic balance.30, 31 Women, especially postmenopausal women, are at high risk for BPV.32 BPV was reported to be an independent risk factor for stroke in postmenopausal women,33 but it has not been fully clarified which factors in this association are important. Our present findings imply that not only known primary physiological changes but also musculoskeletal changes and the development of frailty may play major roles in the development of exaggerated BPV and high CVD risk in postmenopausal women.

In this study, office BP was associated with daytime SD‐SBP. Most of the participants were walking by themselves to the clinic on this small island. Although we had measured the participants’ office BP after a 5‐minute rest, 5 minutes might not have been enough of a rest period for elderly participants with LS for postexercise BP regulation. It was reported that poor postexercise SBP recovery is associated with an elevated risk for myocardial infarction.34 High office BP may reflect poor postexercise SBP recovery, and a longer rest may be needed to correctly measure the office BP of elderly individuals with LS. In addition, the white‐coat effect must be considered. Larger prospective studies are needed to clarify the causal relationship of LS and BPV in the elderly.

Among all of the components of the LS scale, movement‐related difficulty and usual care difficulty were associated with daytime SBPV parameters. It was reported that chronic pain is one of the risk factors for CVD.35 In the Framingham Heart Study, symptomatic hand osteoarthritis was associated with an increased risk of coronary events.36 Another cohort study reported a cardioprotective benefit of primary elective total joint arthroplasty,37 and the study's authors concluded that relieving pain, improvement of physical activity, and reduced use of NSAIDs after that surgery may explain the cardioprotective effect of primary elective total joint arthroplasty. However, none of these studies estimated their patients’ BPV, and the extent of BPV in musculoskeletal diseases had not yet been clarified. The finding of our present study indicates that frailty‐associated chronic pain due to impaired physical activity is linked to short‐term SBPV in elderly hypertensive patients. The results of our present study may be a key to uncovering the mechanisms underlying the association between musculoskeletal disease and the increased risk of CVD in the early stage of frailty.

4.2. Psychological factors and BPV

In the present study, the participants’ scores on the SRQD, a measure of depression, were not associated with any of the BPV parameters. Although the association between depression and BP levels has been controversial, it was recently reported that depression is a risk for short‐term BPV.17 Our present findings showed that the SRQD score was not associated with any of the ABPM parameters, and this may be due to the very low number of study participants in the depressive state.

4.3. Study strengths

This study has several strengths. First, we performed this study in a community‐dwelling population at an outpatient clinic. The participants were elderly individuals with normal ADLs who were able to come to the clinic by themselves, which is generalizable to other elderly populations at an early stage of frailty. Second, we used the LS scale, which covers a wide range of issues from pain to the ADLs as an estimation of the physical health status of the elderly. It is an established method for estimating the extent of locomotive organ dysfunction.4 The present study is important because we showed for the first time that this screening tool for frailty may also be used to determine an independent risk factor for SBPV. Our findings will contribute to the elucidation of the mechanism underlying the relationship between BPV and increased CVD risk in the elderly. Third, we compared the LS scale and psychosocial factors to investigate the association with BPV in the same population.

4.4. Study limitations

Our study has some limitations. First, the study design was cross‐sectional, and the association between the LS scale score and exaggerated SBPV does not imply any causality. Second, the sample size was relatively small. Third, this study of residents of one remote Japanese island does not address the potential ethnic differences in BPV that have been reported.38 Fourth, although we identified an independent association between LS and BPV, we did not collect precise information on potential underlying age‐related confounding factors including vascular stiffness, autonomic nervous dysfunction (eg, postural BP dysregulation, Parkinsonism), sarcopenia and frailty, physical inactivity, or psychocognitive factors such as depression and cognitive function. The findings of our present pilot study provide a foundation for larger prospective studies that would collect information about various confounding factors in order to clarify any causal relationship that exists between frailty and BPV in the elderly.

5. Conclusions

In the present study, the LS scale score was independently associated with exaggerated SBPV in elderly hypertensive patients with normal ADLs. Among the components of the LS scale, movement‐related difficulty and usual care difficulty were significantly associated with short‐term daytime SBPV. Lowering elderly hypertensive individuals’ LS scale score, for example, by treating their chronic pain and treating factors that contribute to movement‐related difficulty and usual care difficulty, may help reduce these individuals’ SBPV.

Supporting information

 

 

 

 

Imaizumi, Y. , Eguchi, K. , Murakami, T. , Saito, T. , Hoshide, S. and Kario, K. (2017), Locomotive syndrome is associated with large blood pressure variability in elderly hypertensives: the Japan Ambulatory Blood Pressure Prospective (JAMP) substudy. Journal of Clinical Hypertension, 19:388–394. doi: 10.1111/jch.12946