Sex-Specific Association of Ambient Temperature With Urine Biomarkers in Southwest Coastal Bangladesh

Introduction Men are vulnerable to ambient heat-related kidney disease burden; however, limited evidence exists on how vulnerable women are when exposed to high ambient heat. We evaluated the sex-specific association between ambient temperature and urine electrolytes, and 24-hour urine total protein, and volume. Methods We pooled a longitudinal 5624 person-visits data of 1175 participants' concentration and 24-hour excretion of urine electrolytes and other biomarkers (24-hour urine total protein and volume) from southwest coastal Bangladesh (Khulna, Satkhira, and Mongla districts) during November 2016 to April 2017. We then spatiotemporally linked ambient temperature data from local weather stations to participants' health outcomes. For evaluating the relationships between average ambient temperature and urine electrolytes and other biomarkers, we plotted confounder-adjusted restricted cubic spline (RCS) plots using participant-level, household-level, and community-level random intercepts. We then used piece-wise linear mixed-effects models for different ambient temperature segments determined by inflection points in RCS plots and reported the maximum likelihood estimates and cluster robust standard errors. By applying interaction terms for sex and ambient temperature, we determined the overall significance using the Wald test. Bonferroni correction was used for multiple comparisons. Results The RCS plots demonstrated nonlinear associations between ambient heat and urine biomarkers for males and females. Piecewise linear mixed-effects models suggested that sex did not modify the relationship of ambient temperature with any of the urine parameters after Bonferroni correction (P < 0.004). Conclusion Our findings suggest that women are as susceptible to the effects of high ambient temperature exposure as men.

Table S1: Piecewise linear regression model showing associations between maximum temperature and concentrations of urine biomarkers for different maximum temperature segments using individual, household, and community level random intercepts.Pvalue indicates the overall significance of maximum temperature and sex interaction for all segments of ambient temperature using the Wald test.Regression co-efficient and 95% confidence intervals represent geometric mean ratio in relation to 5℃ increases in maximum ambient temperature.Footnote: Level of significance is 0.004 (α = 0.05/12 or 0.004) Table S2: Piecewise linear regression model showing associations between maximum temperature and 24-hour excretion of urine biomarkers for different maximum temperature segments using individual, household, and community level random intercepts.P-value indicates the overall significance of maximum temperature and sex interaction for all segments of ambient temperature using the Wald test.¥ Indicate change in mean biomarkers due to 5℃ increases in minimum ambient temperature.
Table S5: Piecewise linear quantile model showing association between daily average temperature and the concentration of urine biomarkers for different average temperature segments using individual level random intercepts.P-value indicates the overall significance of average temperature and sex interaction for all segments of ambient temperature using the Wald test.Footnote: Level of significance is 0.004 (α = 0.05/12 or 0.004).
Table S6: Piecewise linear quantile model showing association between daily average temperature and the 24-hour urine biomarkers for different average temperature segments using individual level random intercepts.P-value indicates the overall significance of average temperature and sex interaction for all segments of ambient temperature using the Wald test.However, women are also involved in strenuous outdoor activities in the vicinity of their homes and fields in many cultures, making them similarly vulnerable to ambient heat.Sex could be a biological variable that influences dimorphic physiological responses against ambient heat exposure, making men more vulnerable (Refer to the Introduction section).

Supplementary figures:
Objectives 3 State specific objectives, including any prespecified hypotheses 5 Unlike men, the health impact of chronic exposure to ambient heat on women is inconclusive, which demands much attention.Disaggregated data by sex and sex-stratified analyses considering the similar heat exposure across population may inform vulnerability and guide clinical and public health interventions to protect all groups affected from adverse health consequences of raised ambient temperature.In this study, we explored whether men and women have similar concentration and excretion of urinary electrolytes (sodium, potassium, calcium, magnesium, chloride), 24-hour excretion of urine total protein, and volume when exposed to a range of ambient temperatures spanning through the winter and summer months in a tropical coastal region of Bangladesh.(Refer to the last paragraph of the Introduction).

Study design 4
Present key elements of study design early in the paper 6 We used urine electrolytes (sodium, potassium, chloride, calcium, and magnesium), total protein, and volume data from a stepped wedge randomized controlled trial conducted in three southwest coastal districts (Khulna, Satkhira, and Bagerhat) of Bangladesh from November 2016 to April 2017, covering both dry winter and early dry summer.The trial followed 1,175 participants from 542 households.We collected urine samples during five visits to evaluate the health effects of providing a low salinity drinking water intervention provided by managed aquifer recharge (MAR).
All participants gained access to the intervention water supply at some point in the study period during the second to fifth visits.Urine samples were collected from both intervention and control phases in all five consecutive months during the study period, making a total data of 5624 person visits.Detail of the study design, selection and enrolment of participants has been reported elsewhere. 22Refer to the "Study setting and design" part of Methods section).( The findings of this secondary analyses (sex-stratified mixed-effect linear regression) are described in the Result section.Table 2 and 3 reported the associations between average ambient temperature and urine biomarkers, Table S1 and S2 reported estimates about maximum ambient temperature and urine biomarkers, and Table S3 and Table S4 reported estimates on the associations between minimum ambient temperature and urine biomarkers.Sensitivity analyses are reported in Table S5 and   S6.

Key results
18 Summarise key results with reference to study objectives 13 (1 st paragraph in the Discussion section) We found no sex difference in the statistical associations between ambient temperature and urine biomarkers, particularly sodium, chloride, potassium, calcium, magnesium, total protein, and urine volume.In contrast to many epidemiological studies highlighting men particularly vulnerable to high ambient heat, our results suggest women are equally susceptible to ambient temperature-related urinary biomarkers changes.We found higher urine sodium and chloride concentrations but

16
We believe this study findings are generalizable to the population with similar coastal and climatic contexts, including saltwater intrusion affected areas where communities experience drinking water salinity (e.g., Ganges River delta, Mekong, and Red River delta).
(Refer to last paragraph in the discussion).

FFootnote:
Level of significance is 0.004 (α = 0.05/12 or 0.004) $ Regression co-efficient and 95% confidence intervals represent geometric mean ratio in relation to 5℃ increases in minimum ambient temperature.

Figure S2 :
Figure S2: Histogram of the concentration and 24-hour excretion of urinary electrolytes (Sodium, Potassium, Chloride, Calcium, and Magnesium) including 24hour excretion of urine total protein and volume.

Figure S3 :
Figure S3: Relationship between ambient temperature (minimum, average, and maximum) and 24-hour urine volume for male and female.

Figure S4 :
Figure S4: Relationship between ambient temperature (minimum, average, and maximum) and 24-hour urine volume for different age-groups.

Figure S10 :
Figure S10: Restricted cubic spline plots with their 95% confidence bands using linear mixed-effect model demonstrating sex-stratified association between the concentration and 24-hour excretion of urinary electrolytes (Sodium, Potassium, Chloride, Calcium, and Magnesium) and 24-hour excretion of urine total protein and volume with minimum ambient temperature, adjusted for age, BMI, physical exercise, smoking, alcohol consumption, sleep duration, religion, household wealth, time of visit, drinking water salinity, and humidity.

Figure S11 :
Figure S11: Restricted cubic spline plots with their 95% confidence bands using linear mixed-effect model demonstrating sex-stratified association between the concentration and 24-hour excretion of urinary electrolytes (Sodium, Potassium, Chloride, Calcium, and Magnesium) and 24-hour excretion of urine total protein and volume with maximum ambient temperature, adjusted for age, BMI, physical exercise, smoking, alcohol consumption, sleep duration, religion, household wealth, time of visit, drinking water salinity, and humidity.

Figure S12 :
Figure S12:Restricted cubic spline plots with their 95% confidence bands using linear mixed-effect model demonstrating sex-stratified association between average temperature and urine electrolytes and total protein after excluding outliers with adjusting for age, BMI, physical exercise, smoking, alcohol consumption, sleep duration, religion, household wealth, time of visit, drinking water salinity, and humidity.

Figure S13 :
Figure S13:Restricted cubic spline plots with their 95% confidence bands using linear mixed-effect model demonstrating sex-stratified association of average ambient temperature with urine electrolytes, 24-hour urine total protein, and volume among participants who did not report any comorbidities after adjusting for age, BMI, physical exercise, smoking, alcohol consumption, sleep duration, religion, household wealth, time of visit, drinking water salinity, and humidity.

Figure S14 :
Figure S14: Restricted cubic spline (RCS) plots with their 95% confidence bands using a linear mixed-effect model demonstrating sex-stratified associations between the concentration and 24-hour excretion of urinary electrolytes (Sodium, Potassium, Chloride, Calcium, and Magnesium) and 24-hour urine total protein and volume with average ambient temperature among participants with complete 24-hour urine collection based on creatinine index >0.7.The RCS plots were adjusted for age, BMI, physical exercise, smoking, alcohol consumption, sleep duration, religion, household wealth, time of visit, drinking water salinity, and humidity.

Figure S15 :
Figure S15: Restricted cubic spline plots with their 95% confidence bands using linear mixed-effect model demonstrating age-stratified association amongst women between the concentration and 24-hour excretion of urinary electrolytes (Sodium, Potassium, Chloride, Calcium, and Magnesium) and 24-hour excretion of urine total protein and urine volume with average ambient temperature, adjusted for other covariates.
the study setting, locations, and relevant dates were described in "Study setting and design", "Ambient temperature data", and "Urine sample collection and processing" part of the Methods section.Data was collected from a stepped wedge randomized controlled trial conducted in 3 southwest coastal districts of Bangladesh.The detail of the study design, recruitment, follow-up, and data collection were published elsewhere. 22Participants 6 (a) Cohort study-Give the eligibility criteria, and the sources and methods of selection of participants.Describe methods of follow-up Case-control study-Give the eligibility criteria, and the sources and methods of case ascertainment and control selection.Give the rationale for the choice of cases and controls Cross-sectional study-Give the eligibility criteria, and the sources and methods of selection of participants 6 This is a secondary analysis of 5624 person visits data collected from stepped wedge randomized controlled trial.Detail of the study design, selection and enrolment of participants has been reported elsewhere. 22(Refer to "Study setting and design" part of the Methods section).
lower 24-hour excretions for both sexes when ambient temperature increased, this is consistent with prior evidence and suggest substantial sweating by both men and women (Refer to the 1 st paragraph in Discussion section).Limitations 19 Discuss limitations of the study, taking into account sources of potential bias or imprecision.Discuss both direction and magnitude of any potential bias 15-16 (7 th and 8 th paragraph in Discussion) We have discussed the plausible limitations of this study including its potential impact (Refer to the 7 th and 8 th paragraph in Discussion section).the overall interpretation of the results with supporting evidence, limitations, results from similar studies and other relevant evidence in the Discussion section.Generalisability 21 Discuss the generalisability (external validity) of the study results

Urine Potassium excretion (mmol/24 hour) ¥
¥ Indicate change in mean biomarkers due to 5℃ increases in maximum ambient temperature.

Table S3 :
Piecewise linear regression model showing associations between minimum temperature and concentrations of urine biomarkers for different minimum temperature segments using individual, household, and community level random intercepts.Pvalue indicates the overall significance of minimum temperature and sex interaction for all segments of ambient temperature using the Wald test.Regression co-efficient and 95% confidence intervals represent geometric mean ratio in relation to 5℃ increases in minimum ambient temperature.

Urine Chloride concentration (mmol/L)
Piecewise linear regression model showing associations between minimum temperature and 24-hour excretion of urine biomarkers for different minimum temperature segments using individual, household, and community level random intercepts.P-value indicates the overall significance of minimum temperature and sex interaction for all segments of ambient temperature using the Wald test.

Urine sample collection and processing" and "Statistical analyses" in the Methods section).
The median age of the sample population was 41 years [Interquartile Range (IQR): 31-54].The majority were women (59%), married (96%), and belonged to the Hindu religion (58 %).Fifty-one percent (51%) of study participants were never smokers, and 97% did not drink alcohol.An overview of the characteristics of participants, including sexstratified distribution, is presented in Table1.In the study area during the 5-month study period (December 2016-April 2017), the highest daily mean temperature was 30.4°C in April, while in January the lowest mean temperature was recorded (16.6 °C) (Refer to the 1 (Refer to the last paragraph of "Statistical analyses" in the Methods section).

st paragraph of the Results, Table 1, and Figure 1).
The estimates of average ambient temperature and urine biomarkers in different models (model 1-4) are given inTable 2 and 3. Furthermore, estimates of maximum and minimum ambient temperature and urine biomarkers in different statistical models (model 1-4) are reported in TableS1, S2, S3, and S4.The detail description of different models is provided in the "Statistical analyses" in Methods section.