Download Seasonal Effects and Blood Pressure

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Jehovah's Witnesses and blood transfusions wikipedia , lookup

Hemorheology wikipedia , lookup

Blood type wikipedia , lookup

Men who have sex with men blood donor controversy wikipedia , lookup

Blood bank wikipedia , lookup

Transcript
Seasonal Effects and Blood Pressure
Blood Pressure Response to Patterns of Weather
Fluctuations and Effect on Mortality
Louise Aubinière-Robb,* Panniyammakal Jeemon,* Claire E. Hastie, Rajan K. Patel,
Linsay McCallum, David Morrison, Matthew Walters, Jesse Dawson, William Sloan, Scott Muir,
Anna F. Dominiczak, Gordon T. McInnes, Sandosh Padmanabhan
Downloaded from http://hyper.ahajournals.org/ by guest on May 4, 2017
Abstract―Very few studies have looked at longitudinal intraindividual blood pressure responses to weather conditions.
There are no data to suggest that specific response to changes in weather will have an impact on survival. We
analyzed >169 000 clinic visits of 16 010 Glasgow Blood Pressure Clinic patients with hypertension. Each clinic visit
was mapped to the mean West of Scotland monthly weather (temperature, sunshine, rainfall) data. Percentage change
in blood pressure was calculated between pairs of consecutive clinic visits, where the weather alternated between 2
extreme quartiles (Q1–Q4 or Q4–Q1) or remained in the same quartile (Qn–Qn) of each weather parameter. Subjects were
also categorized into 2 groups depending on whether their blood pressure response in Q1–Q4 or Q4–Q1 were concordant
or discordant to Qn–Qn. Generalized estimating equations and Cox proportional hazards model were used to model
the effect on longitudinal blood pressure and mortality, respectively. Qn–Qn showed a mean 2% drop in blood pressure
consistently, whereas Q4–Q1 showed a mean 2.1% and 1.6% rise in systolic and diastolic blood pressure, respectively.
However, Q1–Q4 did not show significant changes in blood pressure. Temperature-sensitive subjects had significantly
higher mortality (1.35 [95% confidence interval, 1.06–1.71]; P=0.01) and higher follow-up systolic blood pressure
(1.85 [95% confidence interval, 0.24–3.46]; P=0.02) compared with temperature-nonsensitive subjects. Blood pressure
response to temperature may be one of the underlying mechanisms that determine long-term blood pressure variability.
Knowing a patient’s blood pressure response to weather can help reduce unnecessary antihypertensive treatment
modification, which may in turn increase blood pressure variability and, thus, risk. (Hypertension. 2013;62:190-196.)
Online Data Supplement
•
Key Words: blood pressure
■
hypertension
T
here is growing evidence that outdoor temperature is a
major determinant of the observed seasonal fluctuations
in blood pressure (BP) with higher and lower BP in winter and
summer, respectively.1–6 An inverse association between ambient temperature and BP has been observed in several studies.7–9 Thermoregulatory vasoconstriction, which increases
arterial BP significantly,10 is an adaptive response to provide
enhanced circulatory function for the protective mechanisms
that are activated to maintain temperature in cold weather
(nonshivering thermogenesis and increased metabolic rate).2,11
Elevation of BP induced by a longer period of cold exposure
is not reversible after return to a thermo-neutral temperature in
animal studies12 and may result in ­cold-induced hypertension.
Although several studies explored the effects of seasonal
variations on BP, few studies have looked at longitudinal BP
changes in relation to fluctuations in weather patterns. To
our knowledge, there are no studies that examined the role
■
mortality
■
temperature
■
weather
of sunshine, rain, or air frost on BP. It is unclear whether
BP response to weather parameters like temperature, rainfall, frost, and sunshine is similar in everybody. If there is
heterogeneity in weather-related BP response, it would be
important to know whether intraindividual and interindividual responsiveness to weather changes can predict long-term
risk. The aim of this study was to determine the within-subject changes in BP in response to a range of weather patterns and test whether individual BP response to the weather
is predictive of long-term mortality and BP control in a large
hypertensive cohort.
Methods
Study Population
The Glasgow Blood Pressure Clinic provides secondary and tertiary level service to individuals with hypertension from the West of
Scotland. The details of the study population and settings, clinical
Received December 18, 2012; first decision January 8, 2013; revision accepted April 15, 2013.
From the BHF Glasgow Cardiovascular Research Centre, Institute of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow, United
Kingdom.
*L.A.-R. and P.J. contributed equally to this article.
This paper was sent to Gerald DiBona, Consulting editor, for review by expert referees, editorial decision, and final disposition.
The online-only Data Supplement is available with this article at http://hyper.ahajournals.org/lookup/suppl/doi:10.1161/HYPERTENSIONAHA.
111.00686/-/DC1.
Correspondence to Sandosh Padmanabhan, Institute of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow G12 8TA, United
Kingdom. E-mail [email protected]
© 2013 American Heart Association, Inc.
Hypertension is available at http://hyper.ahajournals.org
DOI: 10.1161/HYPERTENSIONAHA.111.00686
190
Aubinière-Robb et al Weather and Mortality 191
measurements, and outcome assessment are explained in the onlineonly Data Supplement and also described previously.13
individuals to show the BP changes were over and above the expected
annual treatment–induced changes in BP.
Results
Weather Data
The monthly average weather data for the West of Scotland were
obtained from the UK meteorologic office (www.metoffice.gov.uk).
The UK meteorologic office provides monthly average data sets of
13 climate variables that were generated for the periods 1961–2000
using a consistent analysis method. Values were produced for each
station in the meteorologic Office’s observing network and for a 1×1
km grid of points covering the United Kingdom.14 The data generated
for the West of Scotland region, available free of cost for research
purposes, were used in the current study.
Subject Visit Classification by Weather Conditions
Downloaded from http://hyper.ahajournals.org/ by guest on May 4, 2017
Each clinic visit of the patient was mapped to the mean monthly weather
(air frost, mean temp, sunshine, and rainfall) of the West of Scotland
from meteorologic data. The mean monthly weather variables were
categorized into quartiles. Further analyses were restricted to pairs of
consecutive clinic visits within a 12-month period where the weather
alternated between 2 extreme quartiles or remained in the same quartile. Thus, 3 groups were generated, Q1–Q4, where the first visit weather
variable was in the lowest quartile (Q1) and the subsequent visit in the
highest quartile (Q4); Q4–Q1, where the first visit was in the highest
weather variable quartile (Q4) and the second visit in the lowest quartile
(Q1); Qn–Qn, where both the first and the second visits were in the same
weather variable quartile. All the clinic visits of each patient were thus
classified throughout the entire follow-up period. Thus, BP changes in
relation to weather changes were calculated every year, making these
calculations specific for the weather conditions of the annual time-frame
and BP control status of the patient in the same time frame.
Statistical Analysis
The baseline characteristics of the subjects were compared by independent t test for continuous variables and χ2 test for categorical variables.
The traits analyzed were the percentage change in systolic BP, diastolic
BP (DBP), and heart rate (HR) between 2 consecutive clinic measurements ≥30 days apart but within a 12-month period. The upper limit
of time interval between each pair of BP measurements was 6 months.
The direction and magnitude of ΔBP (ΔHR) in Qn–Qn were considered the weather independent effect on BP (ΔHR) between 2 visits and
can reflect intrinsic BP (ΔHR) variability and effect of clinic interventions on BP (ΔHR). The ΔBP (ΔHR) values within each year were considered as paired measurements and were compared using paired t tests.
Subjects were categorized into 2 groups based on the direction and
magnitude of response in the Q1–Q4/Q4–Q1 groups compared with
Qn–Qn. The 2 groups included a concordant response (same direction of response in Qn–Qn and Q1–Q4/Q4–Q1) group and a discordant
response (opposite direction of response in Qn–Qn and Q1–Q4/Q4–Q1)
group. The former group was termed the temperature, sunlight, or
rain nonsensitive, and the latter group was termed the temperature,
sunlight, or rain sensitive. Generalized estimating equations were
used to model the longitudinal change, during the first 5 years of follow-up, in BP and HR in the weather-sensitive and weather-nonsensitive groups after adjustment for baseline variables. Kaplan–Meier
survival and Cox proportional hazards models were set up to analyze
differences in survival between the weather-sensitive and weathernonsensitive groups. The covariates included in the Cox proportional
hazards models were baseline age, sex, body mass index, smoking
status, systolic BP (SBP) or DBP, alcohol use, tobacco use, estimated glomerular filtration rate, and cardiovascular comorbidity.
Because this is a real-life clinic setting, our strategies to ensure that
the BP differences observed were over and above treatment-related
BP changes included the following: (1) using pairwise BP differences
obtained within a 12-month period, (2) using Qn–Qn BP change as
the reference indicating the treatment-induced change in BP within
the 12-month period, (3) restricting each pairwise comparison with
the same 12-month period, and (4) using generalized estimating
equations in temperature-sensitive and temperature-nonsensitive
Baseline Characteristics of the Cohort
The characteristics of the study cohort are presented in Table 1.
More than half (53%) of the participants were women. The
sample was middle aged (mean age, 51 years) and overweight
(mean body mass index, 28 kg/m2). The average duration of
follow-up was 6.5±5.8 years, during which BP changes in
relation to weather fluctuations were analyzed. Twenty-eight
percentage of patients attained target BP of <140/90 and
maintained it at that level for ≥1 year.
Effect of Weather Conditions on BP
Trends in BP change by weather patterns (air frost, mean
temperature, sunshine, and rainfall) are presented in Table 2,
and annual SBP change by temperature is shown in Figure 1.
There were 16 411 to 19 049 pairs of consecutive visits in the
Qn–Qn group, 3115 to 6656 in the Q4–Q1 group, and 3916 to
6445 pairs of consecutive visits in the Q1–Q4 group.
Qn–Qn showed an average of 2.1% (95% confidence interval
[CI], 1.9–2.3), 2.2% (95% CI, 2.0–2.4), 1.7% (95% CI, 1.5–
1.9), and 2.2% (95% CI, 2.0–2.3) drop in SBP for air frost,
temperature, rainfall, and sunshine, respectively. Similar drop
in DBP was also observed for each weather variable.
Q4–Q1 showed an average 2% increase in BP in response
to temperature (2.1; 95% CI, 1.6–2.6) and sunshine (2.3; 95%
CI, 1.9–2.6), but air frost and rainfall showed no significant
difference. Q1–Q4 showed an average 1.4% (95% CI, 0.9–1.8)
and 0.8% (95% CI, 0.5–1.1) increase in BP in response to air
frost and rainfall, respectively. The temperature response of
BP in Q1–Q4 was inconsistent across different years.
Using the BP response in Qn–Qn as the reference trait, the
differences in BP response to weather variation were assessed
using paired t tests (Table 3). Compared with Qn–Qn, the change
from the highest to the lowest quartile (Q4–Q1) elicited a 6% rise
Table 1. Characteristics of the Cohort
Variables
Men (n=7574) Women (n=8436) Total (n=16 010)
Age at first visit, years,
mean (SD)
50.00 (13.40)
BMI, kg/m2, mean (SD)
27.64 (5.19)
51.57 (15.60)
50.82 (14.63)
27.63 (6.27)
27.63 (5.79)
165.44 (30.76)
163.90 (29.23)
98.09 (15.21)
96.33 (23.11)
97.16 (19.79)
Total cholesterol,
mmol/L, mean (SD)
5.83 (1.56)
6.03 (1.34)
5.94 (1.45)
eGFR <60 mL/min per
1.73 m2, n (%)
1158 (18.71)
1939 (28.36)
3097 (23.78)
Alcohol use, n (%)
5001 (74.85)
3452 (46.00)
8453 (59.59)
Tobacco use, n (%)
3393 (49.15)
3128 (40.32)
6521 (44.48)
CVD, n (%)
1361 (19.21)
1172 (14.81)
2533 (16.89)
88 806.04
104 076.55
192 882.6
SBP, mm Hg, mean (SD) 162.18 (27.32)
DBP, mm Hg, mean (SD)
Time at risk
(person-years), total
BMI indicates body mass index; CVD, cardiovascular disease; DBP, diastolic
blood pressure; eGFR, estimated glomerular filtration rate; and SBP, systolic
blood pressure.
192 Hypertension July 2013
Table 2. Change in Blood Pressure and Heart Rate in Response to Weather Conditions
Qn–Qn
Response
SBP
DBP
HR
Q4–Q1
Q1–Q4
Environment
n
Mean % Change (95% CI)
n
Mean % Change (95% CI)
n
Mean % Change (95% CI)
Air frost
17 418
–0.021 (–0.023 to –0.019)
6656
–0.002 (–0.006 to 0.002)
3916
0.014 (0.009 to 0.018)
Mean temp
16 411
–0.022 (–0.024 to –0.02)
3115
0.021 (0.016 to 0.026)
6445
0.001 (–0.004 to 0.005)
0.007 (0.004 to 0.011)
Rain
19 378
–0.014 (–0.016 to 0.012)
4316
–0.016 (–0.02 to –0.012)
6226
Rainfall
19 049
–0.017 (–0.019 to 0.015)
3837
–0.017 (–0.021 to –0.013)
6357
0.008 (0.005 to 0.011)
Sunshine
17 431
–0.022 (–0.023 to 0.02)
5113
0.023 (0.019 to 0.026)
4688
–0.006 (–0.012 to –0.001)
Air frost
17 416
–0.019 (–0.021 to 0.017)
6654
3916
0.009 (0.003 to 0.015)
Mean temp
16 410
–0.021 (–0.023 to 0.019)
3114
0.016 (0.009 to 0.023)
6444
–0.001 (–0.004 to 0.002)
Rain
19 376
–0.014 (–0.015 to –0.012)
4316
–0.014 (–0.018 to –0.01)
6225
0.008 (0.005 to 0.011)
Rainfall
19 049
–0.016 (–0.018 to –0.015)
3836
–0.014 (–0.018 to –0.01)
6355
Sunshine
17 429
–0.019 (–0.021 to –0.018)
5112
0.019 (0.015 to 0.024)
4688
–0.007 (–0.011 to –0.004)
–0.003 (–0.006 to 0)
0.007 (0.004 to 0.01)
Downloaded from http://hyper.ahajournals.org/ by guest on May 4, 2017
Air frost
7545
0.005 (0.002 to 0.008)
3258
0.007 (0.002 to 0.011)
2088
0.003 (–0.002 to 0.009)
Mean temp
6929
0.006 (0.003 to 0.009)
1630
0.004 (–0.002 to 0.01)
3245
0.006 (0.001 to 0.01)
Rain
8760
0.005 (0.002 to 0.007)
1487
0.006 (–0.001 to 0.013)
2238
0.007 (0.001to 0.012)
Rainfall
8353
0.003 (0.001 to 0.006)
1452
0.003 (–0.004 to 0.01)
2139
0.006 (0 to 0.011)
Sunshine
7704
0.007 (0.004 to 0.009)
2575
0.007 (0.002 to 0.012)
2322
0.006 (0 to 0.011)
CI indicates confidence interval; DBP, diastolic blood pressure; HR, heart rate; n, number of pairs of BP measurements; and SBP, systolic blood pressure.
in SBP (mean difference [95% CI], 6.2 [5.1–7.4]; P<0.0001) for
temperature and sunshine (6.0 [5.0–7.0]; P<0.0001), and 4%
rise in SBP for air frost (4.1 [3.0–5.1]; P<0.0001). DBP showed
significant differences only for temperature and sunshine.
Intraindividual differences in BP response between Qn–Qn
and Q1–Q4 (change from the lowest to the highest quartile
of weather) showed 2.0% to 6.6% increases in SBP for air
frost, temperature, rainfall, and sunshine, which were statistically significant. For DBP, similar statistically significant
responses were seen for temperature, rainfall, and sunshine
(Table 3).
Effect of Weather Variation on HR
HR did not show any significant intraindividual differences
for any weather pattern (Table 3).
Longitudinal Outcomes by Response to Weather
Changes
Tables S1 to S3 in the online-only Data Supplement present the
sample characteristics stratified according to the SBP response
to temperature, sunshine, and rainfall. There were no major differences in baseline characteristics between study groups.
Generalized estimating equation analyses performed for
longitudinal BP change, during the first 5 years of followup after adjustment for all conventional covariates, showed a
4/2 mm Hg annual decrease in SBP/DBP, which reflects the
average effect of treatment in the clinic aimed at lowering BP.
Comparing Qn–Qn with Q4–Q1, temperature-sensitive individuals (discordant response between Qn–Qn and Q4–Q1) had a
2.68 mm Hg SBP rise (95% CI, 0.61–4.75; P=0.01) and 1.84
mm Hg DBP rise (95% CI, 0.78–2.91; P=0.001) longitudinally
>5 years compared with the temperature-nonsensitive subjects
after adjustment for all baseline covariates (Table 4). Similarly, a
1.31 mm Hg SBP rise (95% CI, –0.25 to 2.87) and a 1.22 mm Hg
DBP rise (95% CI, 0.42–2.01) were observed for sunlight.
Comparing Qn–Qn with Q4–Q1, temperature-sensitive individuals (discordant SBP change between Qn–Qn and Q4–Q1)
showed a 1.85 mm Hg SBP rise (95% CI, 0.24–3.46) longitudinally >5 years for temperature, and 2.04 mm Hg SBP rise
(95% CI, 0.82–3.27) for sunlight in comparison with temperature- or sunlight-nonsensitive individuals after adjustment for
baseline covariates (Table 4).
Kaplan–Meier survival analyses (Figure 2A), temperaturenonsensitive individuals (Qn–Qn versus Q4–Q1) showed longer
Figure 1. Patterns of systolic blood
pressure (SBP) change annually across
sequential visits. Each point is the mean
percentage change in BP between 2
sequential clinic visits for 3 temperature
groups.
Aubinière-Robb et al Weather and Mortality 193
Table 3. Intraindividual Change in Blood Pressure and Heart Rate in Response to Weather Conditions
Trait
SBP
DBP
HR
Weather
n Pairs
QnQn–Q4Q1
QnQn–Q1Q4
Q1Q4–Q4Q1
Mean % Change (95% CI)
P Value* n Pairs Mean % Change (95% CI) P Value* n Pairs
Mean % Change (95% CI)
P Value*
Air frost
1690
0.041 (0.03 to 0.051)
<0.001
1568
0.048 (0.038 to 0.058)
<0.001
672
0.018 (0.001 to 0.035)
0.04
Mean temp
1124
0.062 (0.051 to 0.074)
<0.001
1434
0.055 (0.043 to 0.066)
<0.001
542
–0.018 (–0.037 to 0.001)
0.07
Rain
2014
0.006 (–0.003 to 0.015)
0.20
2704
0.029 (0.021 to 0.036)
<0.001
1324
0.023 (0.012 to 0.035)
<0.001
Rainfall
1755
0.004 (–0.005 to 0.013)
0.33
2856
0.028 (0.02 to 0.035)
<0.001
1078
0.028 (0.016 to 0.041)
<0.001
Sunshine
1480
0.06 (0.05 to 0.07)
<0.001
1330
0.034 (0.023 to 0.045)
<0.001
834
–0.046 (–0.061 to –0.032)
<0.001
Air frost
1690
0.033 (0.023 to 0.043)
<0.001
1568
0.035 (0.026 to 0.044)
<0.001
672
0.012 (–0.003 to 0.027)
0.12
Mean temp
1124
0.046 (0.036 to 0.057)
<0.001
1434
0.05 (0.04 to 0.061)
<0.001
542
–0.015 (–0.033 to 0.003)
0.11
Rain
2014
0.005 (–0.003 to 0.013)
0.25
2703
0.027 (0.02 to 0.034)
<0.001
1324
0.024 (0.013 to 0.035)
<0.001
Rainfall
1755
0.004 (–0.004 to 0.012)
0.35
2855
0.025 (0.018 to 0.032)
<0.001
1077
0.023 (0.011 to 0.034)
<0.001
Sunshine
1480
0.053 (0.043 to 0.062)
<0.001
1330
0.04 (0.03 to 0.051)
<0.001
834
–0.036 (–0.05 to –0.022)
<0.001
Downloaded from http://hyper.ahajournals.org/ by guest on May 4, 2017
Air frost
610
0.012 (–0.006 to 0.03)
0.185
762
–0.007 (–0.02 to 0.006)
0.29
328
0.008 (–0.013 to 0.03)
0.44
Mean temp
552
–0.016 (–0.032 to –0.001)
0.07
539
–0.007 (–0.025 to 0.011)
0.47
250
–0.011 (–0.037 to 0.015)
0.41
–0.004 (–0.028 to 0.02)
Rain
515
0.003 (–0.014 to 0.021)
0.69
766
–0.008 (–0.022 to 0.006)
0.27
314
Rainfall
506
0 (–0.017 to 0.017)
0.99
791
0.001 (–0.013 to 0.014)
0.91
268
Sunshine
699
–0.013 (–0.029 to 0.004)
0.12
575
0.001 (–0.016 to 0.018)
0.93
406
0.012 (–0.013 to 0.038)
0 (–0.021 to 0.02)
0.75
0.35
0.98
CI indicates confidence interval; DBP, diastolic blood pressure; HR, heart rate; and SBP, systolic blood pressure.
*Bonferroni P=0.0008.
Discussion
survival in comparison with temperature-sensitive individuals (log rank P=0.015). In the Qn–Qn versus Q1–Q4 comparison, the results were not statistically significant (Figure 2B).
Similar results were seen in BP response to sunshine.
In the Cox proportional hazard model, temperature-sensitive individuals (SBP change between Qn–Qn and Q4–Q1)
showed a 35% (HR [95% CI], 1.35 [1.06–1.71]; P=0.01)
increased all-cause mortality compared with temperaturenonsensitive individuals (Table 5). A similar effect was seen
for sunshine. No significant difference was observed for
Qn–Qn versus Q1–Q4.
In this study, we show for the first time the magnitude of
change in BP attributable to changes in weather between
sequential clinic visits of treated patients with hypertension. The results for temperature and sunshine were similar,
whereas those for rainfall were in the opposite direction to
temperature. The BP response to temperature observed in our
study is consistent with the findings of other observational
studies. For example, in asymptomatic individuals every 10°C
decrease in the minimum temperature was associated with a
1.85 and 1.18 mm Hg increase in SBP and DBP, respectively.7
Table 4. Blood Pressure Response to Weather Conditions in the First Year of Treatment and On-Treatment Changes in Blood
Pressure
Temperature
Response
SBP, mm Hg (Qn–Qn vs Q4–Q1)
n/N
Weather nonsensitive
Weather sensitive
0.011
1.31 (–0.25 to 2.87)
1.00
1.85 (0.78 to 2.91)
0.31 (–1.37 to 2.00)
1.20 (–0.08 to 2.48)
0.066
609/2834
1.00
0.001
0.715
1.00
0.286
755/3511
470/2192
P
1136/5384
1.12 (–0.94 to 3.19)
1.00
GEE β (95% CI)
1.00
0.101
1.00
0.946
487/2241
Weather sensitive
n/N
609/2834
474/2219
0.07 (–1.88 to 2.01)
Weather nonsensitive
P
1.00
470/2192
Weather sensitive
GEE β (95% CI)
Rain
755/3511
2.68 (0.62 to 4.75)
Weather nonsensitive
DBP, mm Hg (Qn–Qn vs Q1–Q4)
n/N
1.00
Weather sensitive
DBP, mm Hg (Qn–Qn vs Q4–Q1)
P
487/2241
Weather nonsensitive
SBP, mm Hg (Qn–Qn vs Q1–Q4)
GEE β (95% CI)
Sunlight
1.22 (0.42 to 2.01)
1.00
0.003
474/2219
0.88 (–0.01 to 1.78)
0.053
1136/5384
1.00
1.00
0.52 (–0.48 to 1.52)
1.27 (0.20 to 2.34)
1.00
0.020
0.65 (–0.03 to 1.33)
0.063
Individuals with ≥3 annual BP readings are included in the GEE analyses. GEE analyses is adjusted for age, sex, body mass index, smoking status, alcohol use status,
eGFR, chronic kidney disease status, and year of blood pressure assessment. CI indicates confidence interval; DBP, diastolic blood pressure; eGFR, estimated glomerular
filtration rate; GEE β, generalized estimating equations regression coefficient; n/N=number of groups/number of observations; Q1–Q1, consecutive visits in the same
weather conditions; Q4–Q1, consecutive visits in contrasting weather conditions (high-low); Q1–Q4, consecutive visits in contrasting weather conditions (low-high); and
SBP, systolic blood pressure.
194 Hypertension July 2013
Downloaded from http://hyper.ahajournals.org/ by guest on May 4, 2017
Figure 2. A, Kaplan–Meier plot of all-cause mortality in relation to
systolic blood pressure (SBP) response to temperature (Q4–Q1).
B, Kaplan–Meier plot of all-cause mortality in relation to SBP
response to temperature (Q1–Q4).
BP changes with temperature in similar magnitude have been
observed in other cohorts as well.8,15,16 We also show that the
effect of temperature on BP varies between individuals and,
based on response, patients can be classified as either temperature sensitive or temperature nonsensitive. We show that
temperature-sensitive individuals have a higher follow-up BP
and poorer survival compared with temperature-­nonsensitive
individuals. To our knowledge, this is the first time that
weather parameters, such as temperature, sunshine, and
­rainfall, have been shown to be determinants of BP changes
in the population. However, only temperature response was
an independent predictor of mortality, emphasizing the prognostic role of temperature changes.
There are several physiological explanations for the effect
of temperature on BP. Cold exposure activates the sympathetic nervous system, which, in turn, increases the activity
of the renin–angiotensin system.17 The renin–angiotensin
system suppresses endothelial NO synthase expression and
decreases NO production,18 which contributes to the development of cold-induced hypertension. The renin–angiotensin
system also mediates the cold-induced increase in endothelin-1 production,11 a potent regulator of vascular tone and BP.
Cold exposure upregulates endothelin-A but downregulates
endothelin-B receptors and mediates the thermoregulatory
vasoconstriction. The Pressioni Arteriose Monitorate E Loro
Associazioni (PAMELA) study showed that summer–winter
differences in BP were seen not only in normotensive subjects, but also in untreated and treated subjects with hypertension, indicating that weather is a powerful determinant
of BP variability (BPV).6 Other mechanisms postulated to
account for the BP increase in cold weather include alterations in skin vasomotor tone, resulting in a marked increase in
vascular peripheral resistance; decreased sweating and, therefore, salt loss, increasing the load of sodium on the kidneys,
thus, further contributing to the increase in BP; increased
norepinephrine and epinephrine concentrations in plasma and
urine accompanying the BP increase during the cold season;
increased erythrocyte deformability and blood viscosity, a
major determinant of systemic vascular resistances; a reduced
intensity in ultraviolet light during winter has been shown to
reduce the epidermal photosynthesis of vitamin D3 and parathyroid hormone, which was shown in turn to be associated
with elevated BP levels.19–22
Table 5. Cox-PH Model for All-Cause Mortality and Blood Pressure/Pulse Rate Response to Weather Conditions
Temperature
Response
SBP (Qn–Qn vs Q4–Q1)
Deaths (n)
Weather nonsensitive
Weather sensitive
0.01
1.18 (0.98–1.44)
1.00
1.17 (0.94–1.47)
0.95 (0.82–1.10)
1.00
1.14 (0.90–1.46)
1.00
0.28
0.83 (0.68–1.00)
0.06
772
1.00
0.72
0.48
405
1.00
0.18
0.38
1.00
188
1.04 (0.82–1.32)
0.92 (0.75–1.09)
0.37
265
284
P Value
772
0.90 (0.71–1.13)
1.00
HR (95% CI)
1.00
0.09
1.00
0.49
291
Weather sensitive
Deaths (n)
405
297
0.92 (0.73–1.16)
Weather nonsensitive
P Value
1.00
284
Weather sensitive
HR (95% CI)
Rain
391
1.35 (1.06–1.71)
Weather nonsensitive
DBP (Qn–Qn vs Q1–Q4)
Deaths (n)
1.00
Weather sensitive
DBP (Qn–Qn vs Q4–Q1)
P Value
291
Weather nonsensitive
SBP (Qn–Qn vs Q1–Q4)
HR (95% CI)
Sunlight
0.97 (0.73–1.30)
1.00
0.86
1.03 (0.86–1.25)
0.72
HRs are adjusted for age, sex, alcohol use, smoking status, estimated glomerular filtration rate, chronic kidney disease status, and baseline systolic blood pressure. CI
indicates confidence interval; Cox-PH, Cox proportional hazards; DBP, diastolic blood pressure; HR, hazard ratio; Q1–Q1, consecutive visits in the same weather conditions; Q4–
Q1, consecutive visits in contrasting weather conditions (high-low); Q1–Q4, consecutive visits in contrasting weather conditions (low-high); and SBP, systolic blood pressure.
Aubinière-Robb et al Weather and Mortality 195
Downloaded from http://hyper.ahajournals.org/ by guest on May 4, 2017
Our finding that temperature-sensitive individuals have
higher mortality is in line with current data on visit-tovisit BPV, which is associated with increased mortality.23–25
It is difficult to determine whether the high risk seen in
temperature-sensitive individuals is attributable to BPV or
to the physiological mechanisms activated in response to
temperature changes. There is evidence for both mechanisms.
Temperature influences BPV,26 and BPV has an impact on
survival.27 In summer, a reduction in cardiovascular mortality
has also been observed with a reduction in the occurrence of
stroke in patients with hypertension.28–31 The cardiovascular
morbidity and mortality associated with seasonal changes may
be mediated through BP response to temperature changes. In
a study conducted among 1222 consecutive individuals who
sought medical consultation in hypertension outpatient clinics,
lower ambient temperature was associated with increased
aortic pulse pressure and poor subendocardial viability ratio.32
Low environmental temperature has been strongly associated
with increased hospital admissions for acute myocardial
infarction, stroke, and higher cardiovascular mortality.33,34
The strengths of the current study include the following:
a large cohort of ≈16 000 adults with hypertension, >169 000
clinic visits, the availability of repeat BP measurements, 35
years of follow-up for mortality with median survival time of
32 years, and the ability to link clinic visits with mean monthly
weather changes. Although the previous studies compared the
BP changes recorded in different subjects in different seasons,
we have compared changes in the same individual. We were
strictly limited to analyze the weather-related variables in
their range observed in the region of West of Scotland and
the availability of informative BP data based on clinic visits
occurring in the right sequence with each weather condition.
Our study has limitations. We have used the average of
2 clinic BP measurements obtained under indoor clinic
conditions. This is likely to have underestimated the weather
effects on BP. Also ambient conditions in clinic will be different
from those at work and at home for each patient. We are also
limited by the unavailability of ambulatory BP monitoring
or home BP readings to correlate out of office BP with clinic
BP measurements and their response to weather conditions.
Ambulatory BP monitoring would offer a more reliable
assessment of the actual effect of changing weather conditions
on daily life BP, including the separate assessment of weatherinduced changes on daytime and nighttime BP.6,26 Although we
used BP measurements within a 12-month period, the limitation
of studying a real-life cohort of patients is the absence of a
fixed interval for assessment of BP, because BP changes over
a longer period of time may be significantly influenced by
seasonal weather changes in contrast to BP changes over a
shorter period of time. The age-related increase in BP and the
effect of other baseline variables may confound longitudinal
assessment of BP response to temperature, despite our strategy
to use BP difference within each year. We have not been able to
study the role of antihypertensive agents in modulating the BP
response to weather changes because of incomplete prescribing
data available for this analysis. Furthermore, we did not have
reliable physical activity data to include in our models. Finally,
our study cohort comprised treated patients with hypertension,
and the results may not be generalizable.
Perspectives
Response to weather (especially temperature) can be reflected
in BP and is specific to the individual. BP response to temperature may be one of the underlying mechanisms that
determine long-term BPV. Knowing a patient’s BP response
to weather can help reduce unnecessary antihypertensive
treatment modification, which may in turn increase BPV and,
thus, risk. It remains to be established whether BP response
to temperature or BPV is the causal determinant of increased
risk, or whether it is the underlying autonomic function status
of the individual that is causal and BPV is just a reflection
of this.‍
Acknowledgments
We thank the patients and staff at the Glasgow Blood Pressure Clinic
at the Western Infirmary in Glasgow and National Health Service
Greater Glasgow and Clyde.
Sources of Funding
P. Jeemon is supported by a Wellcome Trust Capacity Strengthening
Strategic Award to the Public Health Foundation of India and a consortium of UK universities.
Disclosures
None.
References
1.Brook RD, Weder AB, Rajagopalan S. “Environmental hypertensionology” the effects of environmental factors on blood pressure
in clinical practice and research. J Clin Hypertens (Greenwich).
2011;13:836–842.
2. Adamopoulos D, Vyssoulis G, Karpanou E, Kyvelou SM, Argacha JF,
Cokkinos D, Stefanadis C, van de Borne P. Environmental determinants of
blood pressure, arterial stiffness, and central hemodynamics. J Hypertens.
2010;28:903–909.
3.Giaconi S, Ghione S, Palombo C, Genovesi-Ebert A, Marabotti C,
Fommei E, Donato L. Seasonal influences on blood pressure in high normal to mild hypertensive range. Hypertension. 1989;14:22–27.
4. Kent ST, Howard G, Crosson WL, Prineas RJ, McClure LA. The association of remotely-sensed outdoor temperature with blood pressure levels in
REGARDS: a cross-sectional study of a large, national cohort of AfricanAmerican and white participants. Environ Health. 2011;10:7.
5. Lewington S, Li L, Sherliker P, et al; China Kadoorie Biobank study collaboration. Seasonal variation in blood pressure and its relationship with
outdoor temperature in 10 diverse regions of China: the China Kadoorie
Biobank. J Hypertens. 2012;30:1383–1391.
6.Sega R, Cesana G, Bombelli M, Grassi G, Stella ML, Zanchetti A,
Mancia G. Seasonal variations in home and ambulatory blood pressure
in the PAMELA population. Pressione Arteriose Monitorate E Loro
Associazioni. J Hypertens. 1998;16:1585–1592.
7.Hong YC, Kim H, Oh SY, Lim YH, Kim SY, Yoon HJ, Park M.
Association of cold ambient temperature and cardiovascular markers. Sci
Total Environ. 2012;435–436:74–79.
8.Alpérovitch A, Lacombe JM, Hanon O, Dartigues JF, Ritchie K,
Ducimetière P, Tzourio C. Relationship between blood pressure and outdoor temperature in a large sample of elderly individuals: the Three-City
study. Arch Intern Med. 2009;169:75–80.
9.Barnett AG, Sans S, Salomaa V, Kuulasmaa K, Dobson AJ; WHO
MONICA Project. The effect of temperature on systolic blood pressure.
Blood Press Monit. 2007;12:195–203.
10.Greif R, Laciny S, Rajek A, Doufas AG, Sessler DI. Blood pressure
response to thermoregulatory vasoconstriction during isoflurane and desflurane anesthesia. Acta Anaesthesiol Scand. 2003;47:847–852.
11. Sun Z. Cardiovascular responses to cold exposure. Front Biosci (Elite Ed).
2010;2:495–503.
12.Shechtman O, Papanek PE, Fregly MJ. Reversibility of cold-induced
hypertension after removal of rats from cold. Can J Physiol Pharmacol.
1990;68:830–835.
196 Hypertension July 2013
Downloaded from http://hyper.ahajournals.org/ by guest on May 4, 2017
13. Paul L, Jeemon P, Hewitt J, McCallum L, Higgins P, Walters M, McClure
J, Dawson J, Meredith P, Jones GC, Muir S, Dominiczak AF, Lowe G,
McInnes GT, Padmanabhan S. Hematocrit predicts long-term mortality in
a nonlinear and sex-specific manner in hypertensive adults. Hypertension.
2012;60:631–638.
14. Perry M, Hollis D. The development of a new set of long-term climate
averages for the UK. Int J Climatol. 2005;25:1023–1039.
15. Kimura T, Senda S, Masugata H, Yamagami A, Okuyama H, Kohno T, Hirao
T, Fukunaga M, Okada H, Goda F. Seasonal blood pressure variation and
its relationship to environmental temperature in healthy elderly Japanese
studied by home measurements. Clin Exp Hypertens. 2010;32:8–12.
16. Madsen C, Nafstad P. Associations between environmental exposure and
blood pressure among participants in the Oslo Health Study (HUBRO).
Eur J Epidemiol. 2006;21:485–491.
17. Sun Z, Cade R, Morales C. Role of central angiotensin II receptors in
cold-induced hypertension. Am J Hypertens. 2002;15(1 pt 1):85–92.
18.Sun Z, Cade R, Zhang Z, Alouidor J, Van H. Angiotensinogen gene
knockout delays and attenuates cold-induced hypertension. Hypertension.
2003;41:322–327.
19.Hata T, Ogihara T, Maruyama A, Mikami H, Nakamaru M, Naka
T, Kumahara Y, Nugent CA. The seasonal variation of blood pressure in patients with essential hypertension. Clin Exp Hypertens A.
1982;4:341–354.
20. Rostand SG. Ultraviolet light may contribute to geographic and racial
blood pressure differences. Hypertension. 1997;30(2 pt 1):150–156.
21. Sharma BK, Sagar S, Sood GK, Varma S, Kalra OP. Seasonal variations
of arterial blood pressure in normotensive and essential hypertensives.
Indian Heart J. 1990;42:66–72.
22. Cuspidi C, Ochoa JE, Parati G. Seasonal variations in blood pressure: a
complex phenomenon. J Hypertens. 2012;30:1315–1320.
23. Mancia G, Bombelli M, Facchetti R, Madotto F, Corrao G, Trevano FQ,
Grassi G, Sega R. Long-term prognostic value of blood pressure variability in the general population: results of the Pressioni Arteriose Monitorate
e Loro Associazioni Study. Hypertension. 2007;49:1265–1270.
24. Rothwell PM, Howard SC, Dolan E, O’Brien E, Dobson JE, Dahlöf B,
Sever PS, Poulter NR. Prognostic significance of visit-to-visit variability,
maximum systolic blood pressure, and episodic hypertension. Lancet.
2010;375:895–905.
25.Rothwell PM, Howard SC, Dolan E, O’Brien E, Dobson JE, Dahlöf
B, Poulter NR, Sever PS; ASCOT-BPLA and MRC Trial Investigators.
Effects of β blockers and calcium-channel blockers on within-individual variability in blood pressure and risk of stroke. Lancet Neurol.
2010;9:469–480.
26. Modesti PA, Morabito M, Bertolozzi I, Massetti L, Panci G, Lumachi C,
Giglio A, Bilo G, Caldara G, Lonati L, Orlandini S, Maracchi G, Mancia
G, Gensini GF, Parati G. Weather-related changes in 24-hour blood pressure profile: effects of age and implications for hypertension management.
Hypertension. 2006;47:155–161.
27. Muntner P, Shimbo D, Tonelli M, Reynolds K, Arnett DK, Oparil S. The
relationship between visit-to-visit variability in systolic blood pressure
and all-cause mortality in the general population: findings from NHANES
III, 1988 to 1994. Hypertension. 2011;57:160–166.
28. van Rossum CT, Shipley MJ, Hemingway H, Grobbee DE, Mackenbach
JP, Marmot MG. Seasonal variation in cause-specific mortality: are
there high-risk groups? 25-year follow-up of civil servants from the first
Whitehall study. Int J Epidemiol. 2001;30:1109–1116.
29. Donaldson GC, Robinson D, Allaway SL. An analysis of arterial disease
mortality and BUPA health screening data in men, in relation to outdoor
temperature. Clin Sci (Lond). 1997;92:261–268.
30. Jakovljević D, Salomaa V, Sivenius J, Tamminen M, Sarti C, Salmi
K, Kaarsalo E, Narva V, Immonen-Räihä P, Torppa J, Tuomilehto J.
Seasonal variation in the occurrence of stroke in a Finnish adult population. The FINMONICA Stroke Register. Finnish Monitoring Trends and
Determinants in Cardiovascular Disease. Stroke. 1996;27:1774–1779.
31. Sheth T, Nair C, Muller J, Yusuf S. Increased winter mortality from acute
myocardial infarction and stroke: the effect of age. J Am Coll Cardiol.
1999;33:1916–1919.
32. Adamopoulos D, Vyssoulis G, Karpanou E, Kyvelou SM, Argacha JF,
Cokkinos D, Stefanadis C, van de Borne P. Environmental determinants of
blood pressure, arterial stiffness, and central hemodynamics. J Hypertens.
2010;28:903–909.
33. Danet S, Richard F, Montaye M, Beauchant S, Lemaire B, Graux C, Cottel
D, Marécaux N, Amouyel P. Unhealthy effects of atmospheric temperature and pressure on the occurrence of myocardial infarction and coronary
deaths. A 10-year survey: the Lille-World Health Organization MONICA
project (Monitoring trends and determinants in cardiovascular disease).
Circulation. 1999;100:E1–E7.
34.Dilaveris P, Synetos A, Giannopoulos G, Gialafos E, Pantazis A,
Stefanadis C. CLimate Impacts on Myocardial infarction deaths in the
Athens TErritory: the CLIMATE study. Heart. 2006;92:1747–1751.
Novelty and Significance
What Is New?
Summary
• The effect of temperature on blood pressure varies between individuals
Response to weather (especially temperature) can be reflected in
BP and is specific to the individual. Individuals with temperaturesensitive BP are at higher risk of early death. It remains to be established whether BP response to temperature or BP variability
induced is the causal determinant of increased risk.
and, based on response, patients can be classified as either temperature
sensitive or temperature nonsensitive. Temperature-sensitive individuals
have a higher follow-up blood pressure (BP) and poorer survival compared with temperature-nonsensitive individuals.
What Is Relevant?
• Intraindividual BP change in response to ambient temperature may identify patients with hypertension at high risk.
Blood Pressure Response to Patterns of Weather Fluctuations and Effect on Mortality
Louise Aubinière-Robb, Panniyammakal Jeemon, Claire E. Hastie, Rajan K. Patel, Linsay
McCallum, David Morrison, Matthew Walters, Jesse Dawson, William Sloan, Scott Muir, Anna
F. Dominiczak, Gordon T. McInnes and Sandosh Padmanabhan
Downloaded from http://hyper.ahajournals.org/ by guest on May 4, 2017
Hypertension. 2013;62:190-196; originally published online May 6, 2013;
doi: 10.1161/HYPERTENSIONAHA.111.00686
Hypertension is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2013 American Heart Association, Inc. All rights reserved.
Print ISSN: 0194-911X. Online ISSN: 1524-4563
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://hyper.ahajournals.org/content/62/1/190
Data Supplement (unedited) at:
http://hyper.ahajournals.org/content/suppl/2013/05/06/HYPERTENSIONAHA.111.00686.DC1
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Hypertension can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located,
click Request Permissions in the middle column of the Web page under Services. Further information about
this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Hypertension is online at:
http://hyper.ahajournals.org//subscriptions/
Supplementary Material
Blood Pressure Response to Patterns of Weather Fluctuations and Effect on Mortality
Louise Aubinière-Robb,* Panniyammakal Jeemon,* Claire E Hastie, Rajan K Patel, Linsay
McCallum, David Morrison, Matthew Walters, Jesse Dawson, William Sloan, Scott Muir ,
Anna F Dominiczak, Gordon T McInnes, Sandosh Padmanabhan
Expanded methods:
Study Population
The Glasgow Blood Pressure Clinic (GBPC) provides secondary and tertiary level service to
individuals with hypertension from the West of Scotland. All patients referred to the BP
clinic had a diagnosis of hypertension made in primary care using definitions of hypertension
based on contemporaneous guidelines and if appropriate treatment commenced in primary
care. All patients were treated at GBPC until their BP control was stabilized with continuing
follow-up at the BP clinic or in primary care. The frequency of visits to GBPC mainly
depended on individual patient BP levels and presence of other co-morbidities. The West of
Scotland research ethics service (WoSRES) of National Health Service approved the study in
GBPC database (11/WS/0083).
Clinical Measurements
The GBPC employs specialist hypertension nurses who are experienced and highly trained in
BP measurement. The procedure required subjects to rest for 5 minutes in a seated position
before BP was manually measured using standardized mercury sphygmomanometers. Three
BP measurements were performed, one minute apart, and the mean of the second and third
measurements was recorded. Blood pressure was measured consistently in the same arm for
each patient at every visit. No assessment of inter-arm BP difference was possible as the data
was not available in the database. Heart rate was measured over one minute manually.
Patients attending the clinic were advised to take their regular medications as usual. Each
patient attended the same clinic, therefore at each visit their BP measurement would occur in
the same 3 hour time window either in the morning or afternoon. Consumption of food and
drink, and level of physical exertion before each clinic appointment, could not be controlled.
Drug substitution, addition and dose adjustment occurred during follow-up and in accordance
with clinical guidelines.14,15 Prescribed medications were cross checked with patients at each
clinic visit, and they were advised to comply with their treatment at all times. However
formal concordance testing was unavailable.
Outcome Assessment
Record linkage with the General Register Office for Scotland ensured notification of a
subject’s death (provided that it occurred in the United Kingdom) together with the primary
cause of death according to the International Classification of Diseases, 10th Revision,
Version for 2007 (ICD-10) codes. We considered cardiovascular deaths (CVD mortality;
ICD-10 codes I00-I99), ischemic heart disease deaths (IHD mortality; ICD-10 codes 120I25), and stroke deaths (stroke mortality; ICD-10 codes I60-I69) in the analysis. Deaths other
than due to cardiovascular causes are classified as non-CVD deaths. Mortality data were
collected up to April 2011 allowing a maximum of 35 years for participants who had been
under follow up for the longest time.
Table S1: Characteristics of study populations stratified by groups based on blood pressure response to temperature
Clinical Variable
SBP mmHg (Qn-Qn Vs Q4-Q1)
SBP mmHg (Qn-Qn Vs Q1-Q4)
Same Direction Opposite
Same Direction
Opposite Direction
(n=473)
Direction (n=544) P Value
(n=532)
(n=596)
P Value
51.97 (14.16)
53.18 (13.04)
0.156
51.30 (14.32)
52.40 (14.71)
0.203
Age at first visit (years), mean (SD)
230 (48.63)
283 (52.02)
0.281
254 (47.74)
289 (48.49)
0.802
Men, n (%)
27.77 (5.55)
27.67 (6.04)
0.791
28.07 (6.82)
27.95 (5.38)
0.753
BMI (Kg/m2), mean (SD)
167.06 (29.06)
167.73 (28.94)
0.714
162.69 (26.09)
162.10 (29.21)
0.726
SBP (mmHg), mean (SD)
98.02 (14.89)
99.41 (14.76)
0.136
96.57 (13.49)
96.23 (14.97)
0.691
DBP (mmHg), mean (SD)
5.94 (1.33)
0.232
5.79 (1.21)
5.92 (1.24)
0.106
Total cholesterol (mmol/l), mean (SD) 5.84 (1.13)
118 (25.17)
0.436
107 (22.86)
118 (22.82)
0.988
eGFR <60 mL/min per 1.73 m2, n(%) 108 (27.62)
231 (57.32)
287 (61.59)
0.201
276 (59.74)
315 (60.11)
0.905
Alcohol Use, n (%)
Tobacco Use, n (%)
185 (44.79)
242 (46.69)
0.143
202 (42.17)
235 (43.84)
0.591
106 (25.30)
113 (23.01)
0.422
92 (18.81)
119 (21.64)
0.259
CVD, n (%)
147.63 (24.37)
150.42 (23.92)
0.067
145.01 (21.47)
147.99 (21.78)
0.021
Achieved SBP (mmHg), mean (SD)
88.13 (13.53)
88.81 (12.63)
0.407
86.92 (10.90)
87.32 (12.18)
0.566
Achieved DBP (mmHg), mean (SD)
SBP=systolic blood pressure, SD=standard deviation, BMI=body mass index, DBP=diastolic blood pressure, eGFR=estimated glomerular
filtration rate, CVD=cardiovascular disease.
Table S2: Characteristics of study populations stratified by groups based on blood pressure response to sunshine
Clinical Variable
SBP mmHg (Qn-Qn Vs Q4-Q1)
SBP mmHg (Qn-Qn Vs Q1-Q4)
Same Direction Opposite
Same Direction
Opposite Direction
(n=650)
Direction (n=710) P Value
(n=493)
(n=541)
P Value
51.02 (14.47)
52.21 (13.97)
0.125
52.80 (14.91)
51.98 (13.20)
0.351
Age at first visit (years), mean (SD)
297 (45.69)
332 (46.76)
0.693
229 (46.45)
274 (50.65)
0.177
Men, n (%)
27.73 (5.50)
27.39 (5.23)
0.247
27.90 (6.10)
27.79 (5.46)
0.752
BMI (Kg/m2), mean (SD)
165.33 (28.55)
166.14 (28.68)
0.603
168.06 (27.50)
164.63 (30.01)
0.057
SBP (mmHg), mean (SD)
97.64 (14.90)
98.22 (14.29)
0.459
98.02 (14.87)
98.44 (14.67)
0.648
DBP (mmHg), mean (SD)
5.91 (1.25)
0.727
5.86 (1.31)
6.02 (1.24)
0.058
Total cholesterol (mmol/l), mean (SD) 5.93 (1.26)
180 (26.71)
0.927
112 (26.23)
130 (26.80)
0.845
eGFR <60 mL/min per 1.73 m2, n(%) 161 (26.48)
358 (57.84)
412 (60.77)
0.283
262 (59.68)
311 (63.60)
0.22
Alcohol Use, n (%)
Tobacco Use, n (%)
271 (42.54)
314 (45.31)
0.31
190 (41.39)
223 (44.07)
0.401
129 (20.03)
164 (23.40)
0.135
105 (22.53)
95 (18.48)
0.116
CVD, n (%)
146.17 (21.12)
148.91 (22.22)
0.02
148.43 (23.89)
150.17 (23.11)
0.236
Achieved SBP (mmHg), mean (SD)
87.40 (11.77)
88.47 (11.52)
0.093
87.90 (11.55)
89.40 (11.61)
0.039
Achieved DBP (mmHg), mean (SD)
SBP=systolic blood pressure, SD=standard deviation, BMI=body mass index, DBP=diastolic blood pressure, eGFR=estimated glomerular
filtration rate, CVD=cardiovascular disease.
Table S3: Characteristics of study populations stratified by groups based on blood pressure response to rainfall
Clinical Variable
SBP mmHg (Qn-Qn Vs Q4-Q1)
SBP mmHg (Qn-Qn Vs Q1-Q4)
Same Direction Opposite
Same Direction
Opposite Direction
(n=581)
Direction (n=631) P Value
(n=982)
(n=1182)
P Value
51.19 (13.57)
50.51 (13.64)
0.385
51.02 (13.94)
51.56 (13.76)
0.366
Age at first visit (years), mean (SD)
275 (47.33)
312 (49.45)
0.462
483 (49.19)
587 (49.66)
0.825
Men, n (%)
27.65 (6.11)
27.37 (5.28)
0.398
27.48 (5.42)
27.36 (5.46)
0.622
BMI (Kg/m2), mean (SD)
171.98 (26.41)
168.56 (29.28)
0.034
171.16 (28.42)
170.38 (29.51)
0.532
SBP (mmHg), mean (SD)
99.73 (13.78)
101.36 (14.28)
0.261
100.50 (14.72)
100.88 (15.65)
0.567
DBP (mmHg), mean (SD)
6.20 (1.31)
0.682
6.10 (1.33)
6.09 (1.23)
0.867
Total cholesterol (mmol/l), mean (SD) 6.17 (1.35)
129 (21.72)
0.768
226 (25.34)
268 (24.32)
0.601
eGFR <60 mL/min per 1.73 m2, n(%) 117 (21.01)
374 (68.25)
400 (66.33)
0.49
593 (64.95)
740 (66.31)
0.522
Alcohol Use, n (%)
Tobacco Use, n (%)
254 (42.28)
302 (49.27)
0.171
456 (48.82)
515 (45.02)
0.084
127 (22.36)
135 (21.74)
0.797
223 (23.60)
275 (23.83)
0.901
CVD, n (%)
149.19 (20.88)
150.04 (22.28)
0.493
149.46 (21.59)
151.72 (23.65)
0.022
Achieved SBP (mmHg), mean (SD)
87.63 (9.95)
88.90 (11.25)
0.038
88.97 (11.62)
89.68 (12.27)
0.169
Achieved DBP (mmHg), mean (SD)
SBP=systolic blood pressure, SD=standard deviation, BMI=body mass index, DBP=diastolic blood pressure, eGFR=estimated glomerular
filtration rate, CVD=cardiovascular disease.