Seasonal variation of decompensated heart failure admissions and mortality rates in sub-Saharan Africa, Cameroon

Decompensated heart failure (HF) and related deaths are common in cardiology units. Epidemiological data in north hemisphere suggest higher rates during winter. We aimed to describe the seasonal variation of decompensated HF admission and mortality rates in a country with equatorial climate. We conducted a retrospective cross-sectional study from 2016 to 2018 in cardiology unit of the Yaoundé Central Hospital, collecting data on HF morbidity and mortality from registers and patients’ les. Climate data of the corresponding period were obtained from the meteorology oce of the Cameroonian ministry of transports. Analysis of variance and Chi square test were respectively used to compare the continuous and categorical data between seasons. Correlation between continuous data was assess with the Pearson correlation coecient.


Abstract
Background Decompensated heart failure (HF) and related deaths are common in cardiology units. Epidemiological data in north hemisphere suggest higher rates during winter.

Objective
We aimed to describe the seasonal variation of decompensated HF admission and mortality rates in a country with equatorial climate.

Methods
We conducted a retrospective cross-sectional study from 2016 to 2018 in cardiology unit of the Yaoundé Central Hospital, collecting data on HF morbidity and mortality from registers and patients' les. Climate data of the corresponding period were obtained from the meteorology o ce of the Cameroonian ministry of transports. Analysis of variance and Chi square test were respectively used to compare the continuous and categorical data between seasons. Correlation between continuous data was assess with the Pearson correlation coe cient.

Conclusion
Decompensated HF admissions and deaths increases in rainy season where the temperature is lower.
This suggest a need of effective methods to mitigate the effect of cold on cardiovascular diseases.

Background
Cardiovascular diseases (CVDs) are a global pandemic leading to about 17.9 million deaths each year worldwide, representing 31% of all global deaths [1]. Among these CVDs, heart failure (HF) which is characterize by the inability of the heart to pump enough blood to meet metabolic demands, affect at least 37.7 million people worldwide [2]. Its prevalence increases with age to about 10% in patients aged above 75 years [3]. In sub-Saharan Africa, HF represent 9.4-42.5% of all medical admissions, and 25.5-30% of all admissions in cardiology unit. The mortality all-cause mortality rate of hospitalized patient at one year following discharge rage from 21.9 to 57.9% [4]. Most of patients hospitalized for heart failure are presenting with signs and symptoms of congestion and uid retention, in the context of pre-existing cardiomyopathy. A potential precipitating factor for this so-called decompensated heart failure was identi ed in 88.5% of patients. Among these factors are the noncompliance with diet, lack of adherence to prescribed medications, infections, arrhythmias, acute coronary ischemia, uncontrolled hypertension and other miscellaneous causes [5].
Climate and seasons variations play a role in onset and mortality of CVDs. In many countries, there is a clear seasonal pattern in mortality rates of CVDs and myocardial infarction, being higher in winter than in other seasons. Low ambient temperature, physical inactivity, air pollution and infections were the main factors associated to these variations [6]. Previous studies have described similar seasonal variations for heart failure. Highest admission and mortality rates were reported in winter. In addition, mean length of stay and median hospitalization cost were highest in winter [7]. Many hypothesis have been suggested to explain this burden of admissions in winter, among which : (i) increased cardiovascular risk due to lower physical activity; (ii) lower vitamin D synthesis linked to lower light exposition; (iii) increased blood pressure and decreased uid lost related to lower temperature; (iv) higher pollution levels and u-like infection frequency [8]. However, climate and meteorological parameters in northern hemisphere are different from those observed in equatorial region, and data on seasonal variation of CVDs in this later region are really scarce. This study aimed to describe variations of decompensated HF hospitalizations and mortality rates between the different seasons and the in uence of meteorological parameters on these admissions and deaths in Cameroon, a sub Saharan Africa country located in the equatorial region.

Study population and setting
We conducted a retrospective cross-sectional study in cardiology department of the Yaoundé Central Hospital, including all adult patients admitted for decompensated HF between January 1st, 2016 and December 31st, 2018. The diagnosis of decompensated HF was based on signs and symptoms, and transthoracic echocardiography results, according to the 2016 European Society of Cardiology Guidelines for the diagnosis and treatment of acute and chronic heart failure [9]. Patients with incomplete les were excluded.

Data collection
Participants' data were collected from admission registers of the unit and patients' les. This included the total number of admissions for each month and year, potential precipitating factors of decompensated HF, survival outcome of the subject.
We used climate data obtained from the meteorology o ce of the Cameroonian ministry of transports.
These were the monthly data of precipitation, humidity and temperature over the period of 2016-2018. From previous geographical studies done in Yaoundé, we de ned four major seasons which were the long dry season (November, December, January, February and March), the short rainy season (April and May), the short dry season (June and July) and the long rainy season (August, September and October) [10]. We also de ned dry and wet months as month with a RI less than one ore greater than one respectively.

Statistical analysis
Categorical data were described with number and percentages, while continuous variables were reported as mean ± standard deviation for normally distributed data.

Seasonal variation of decompensated HF admissions and deaths
Admission rate for decompensated HF was higher during the long rainy season (38.2%) compare with all the other season (36.7%, 33.1% and 35% respectively for long dry, short rainy and short dry seasons). We observed a similar result for mortality and lethality rates -with highest values during the long rainy season. All these differences were not statistically signi cant (Fig. 1). Admission rates, Lethality rates and mortality rates were respectively 38.2%, 21.9% and 8.2% for coolest months, 36.6%, 12.1% and 5.2% for hottest months: these differences were not statistically signi cant. Precipitating factors -identi ed in admitted patients-of decompensated HF were similarly distributed between the different seasons ( Table 1).

Correlation between meteorological parameters and rates
Admission, lethality and mortality rates were inversely correlated to monthly average temperatures, with p values of 0.070, 0.030 and 0.020 respectively. There was no statistically signi cant linear correlation between these rates and precipitations nor wet averages ( Table 2).

Discussion
Cardiovascular diseases morbidity and mortality are increasing over time, -decompensated HF being one of these diseases -and constituted a burden for countries health systems, especially in sub Saharan Africa. Identi cation of factors contributing to this burden help to understand their pathophysiology, and to nd right solution. Climate and meteorological parameters have been suggested as factors that affect admissions and mortality related to CVDs. We found that admission rates and mortality indicators were higher during long raining seasons (where humidity level is higher, and temperature is lower) and were inversely correlated to temperature.
Admission for decompensated HF represented 36.2% of total admissions during our study period; more than half of deaths in the cardiology unit were related to decompensated HF. Boombhi et al. observed a similar result in the same unit few years ago, with admissions rate of 30% and 33.3% respectively in 2014 and 2015 [11]. These high admission rates can be explained by the increasing rate of cardiovascular risk factors -among these risk factors are hypertension, obesity, smoking-in sub Saharan Africa; knowing that HF is one of the most common outcome of these risk factors [12]. Furthermore, there is a low treatment awareness, medical adherence and disease control for these CVDs. Medication nonadherence and uncontrolled hypertension were among the three most frequent precipitating factors for decompensated HF in our study. Therefore, more sensibilization is needed to reduce the burden of cardiovascular risk factors and thus HF burden. Mortality rate was lower in a group of patients admitted in US for HF between 2011 and 2013 -the overall mortality rate there was 3.1% [7]. This suggest that sub-Saharan Africa health systems need more resources and capacity building to decrease HF mortality.
Admission, mortality and lethality rates were inversely correlated to the monthly mean temperature, and they seem higher during the long rainy season. This result is consistent with the result of Ansa et al., who observed a higher admission rate of heart failure during the rainy season [13]. Similar results were also observed in the north hemisphere, where higher rates of heart failure admissions were obtained during the winter -a season where temperature are lower and humidity is higher compare with other seasons [7]. The main explanation of this higher admission rate seems to be related to lower temperatures in these seasons, through various mechanism. Exposition to cold increases the sympathetic and reninangiotensin -aldosterone systems activities, leading to higher heart rate, peripherical vascular resistance and blood pressure. This exposition also decreases vasodilator effect of azote monoxide. All these effects increase heart post charge and reduce the cardiac output, resulting in higher probability of decompensation of pre-existing HF [14]. Patient's education must therefore emphasize on the need of more attention to the respect of therapeutic recommendations during cold periods like rainy seasons. Health workers and stakeholders must also be aware of this increase of needs in cardiology unit during rainy seasons.
The higher mortality rate observed during the rainy season may also be explained by external factors to the patient like the availability of heath human resources in hospitals during rainy seasons, especially in low and middle incomes countries like Cameroon. Most of the health worker in our context doesn't own a private car which can permit them to get at work on time when there is rain. The health care workforce availability may thus be reduced during rainy seasons, leading to higher mortality rates as it have been showed in other context [15].
Our study may have underestimated the real burden of decompensated HF, as it was a hospital-based study. In our country, patients recourse to other therapeutic means like traditional healer, prayers and selfmedication. Another limit is that autopsy wasn't perform for death patient, and they may be other death causes for our patient that HF.

Conclusion
Decompensated HF admission and mortality rates remain high in sub Saharan Africa. Rainy season is the period where most decompensation and deaths occur, and this may be related to lower temperatures in that season. However, external factors like availability of health work force during the rainy period may also play a role in this higher mortality during rainy season. Effective methods to mitigate the effect of cold on cardiovascular system and thus CVDs are needed.
Abbreviations ANOVA: Analysis of variance; CVDs: Cardiovascular diseases; HF: Heart failure; RI: Rainfall index Declarations Ethics approval and consent to participate Before the start of the study, an ethical clearance was obtained from the Regional Ethical Committee for Centre region's representation of the Cameroonian National Ethical Committee for Research in Human Health (Reference number: CE N°-783-/CRERSHC/2019). Furthermore, this study was conducted in accordance with the ethical principles found in the 1964 Declaration of Helsinki and its later amendments. Inform consent was not request for this retrospective study, and data were handled anonymously.

Consent for publication
Not applicable Availability of data and materials The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.