OBSERVATIONAL STUDY ON THE INFLUENCE OF NOISE POLLUTION ON THE QUALITY OF SLEEP OF PORTO RESIDENTS COMPARED WITH THAT OF RURAL COMMUNITIES*

Exposure to noise pollution has escalated to alarming levels and is causing a number of health problems. It leads to interrupted sleep, increasing stage 1 and awake and decreasing REM and deep sleep. We sought to find a correlation between exposure to high levels of noise pollution in industrialized districts and sleeping problems reported by their inhabitants compared with rural neighbourhoods, using the Pittsburgh Sleep Quality Index (PSQI). This study was composed of 65 people, 33 living in the district of Porto and 32 in rural villages. Each participant was asked to complete two questionnaires; one where they reported their sociodemographic data and the PSQI which indirectly evaluated their sleep quality in the last month. The results from each of the studied groups were compared and analysed using binomial, T and Chi-Square tests. The present study shows that noise has an implication, although limited, on people’s sleeping patterns in the city of Porto, as it affects their daytime life and productivity with a p-value of 0.046. Preventive measures should be taken that focus on reducing nocturnal noise. The responsible departments should enforce the regulations established by the government.


Introduction
Noise pollution is defined as a group of annoying sounds to which populations are involuntarily exposed, coming from traffic (aerial, metropolitan or from cars), industrialization or social activities (Jakovljević et al., 2006;Ruparel et al., 2013).
In the past few years, industry and vehicles circulation has increased a great deal in our cities, originating traffic, and therefore, alarming levels of sound and atmospheric pollution that weren't there before (Geravandi et al., 2015). Exposure to these pollutants, especially sound pollution, has a number of damaging effects on the population living in these cities, sleep quality being one of the most affected (Geravandi et al., 2015). Being aware of this problematic, the World Health Organization (WHO, 1995) has set 40 decibels (dB) as the limit at which noise exposure becomes prejudicial (Berglund et al., 1999). Unfortunately, more than 30% of Europeans are exposed to traffic noise above 55dB at nighttime, exposure that may cause sleep problems (Berglund et al., 1999). Nocturnal noise causes arousals at levels as low as 48dB and physiological reactions, like increasing hormonal secretion, cortical arousals and body movements at levels around 33dB (Halperin, 2014).
At nighttime, the noise that originates from traffic becomes intermittent, different from daytime when it is more continuous, as more people are driving (Pirrera et al., 2010). This provokes even more sleeping problems than continuous noise does (Pirrera et al., 2010).
The effects of exposure to noise pollution in human organisms are not the same for everyone (De Paiva Vianna et al., 2015). It depends on the specific characteristics of the sound, such as frequency, exposure time, intensity and individual susceptibility (De Paiva Vianna et al., 2015). Because of these differences, in this study, we took in consideration some of these aspects, such as the levels of sound exposure in the studied areas, with help from noise maps given by the councils of each city; and accessed the data from the participants regarding the amount of time they have lived there.
A healthy sleep schedule is fundamental for a good physiologic and mental capacity throughout the day (Griefahn et al., 2004). Sleep disturbances include biological responses that may have numerous adverse effects in health and well-being that range from difficulties in falling asleep (longer latencies to sleep), alterations on sleeping patterns, arousals, increasing blood pressure and even cardiac arrhythmias (Griefahn et al., 2004). The increase in sleep latency and the interference with sleep patterns may occur even when the subject is not consciously awakened by the excessive noise (Ruparel et al., 2013). The fracture of sleep by nocturnal noise typically increases stage 1 sleep and stage awake, decreasing deep and REM sleep, causing a lighter, less restoring sleep (Halperin, 2014). This fact not only disturbs the person during the night, but has repercussions on daily life, as people living with this problem experience fatigue during the day, irritation, mood swings and decrease in cognitive function (Halperin, 2014).
The pathophysiologic basis of the sleep-noise relation is based in the stimulation of the hypothalamus-hypophysesadrenal axis, the medulla and the sympathetic nervous system with the subsequent secretion of stress hormones, like adrenalin, noradrenalin and cortisone (Maschke et al., 2004). These responses may have long term implications on the health status of exposed subjects (Jakovljević et al., 2006).  (Halperin, 2014). The last one considered the most harmful non-auditory consequence due to its effect in quality of life and daily performance (Halperin, 2014).
An European study carried out in 2015 reveals Portugal as the third country with the highest index of exposure to excessive noise in a group of 11 European countries evaluated. Porto stands out as the loudest city, followed by Lisbon and Coimbra (Albera et al., 2015). The studied rural areas gathered data shows that a good part of the municipality of Cinfães doesn't present significant noise levels, and there are no particularly critical situations from an acoustic point of view (Engineering, 2016). Buildings located in the surroundings of national roads are the most exposed but

Study Characteristics
This is an observational study be as only possible changes in sleep patterns of the sample were documented without exerting any influence on it or on the results found. It is also retrospective since the researched process has already occurred, and descriptive as it describes a reality impartially, without interference from the researchers and lastly, cross-sectional as a single evaluation was performed.

Selection of Participants
Each participant was recruited on social media platforms, such as Facebook, Instagram and Twitter, during February and March 2021, asked to fill 2 questionnaires, one where they report demographic and health-related data, in order to select the eligible participants, which contained 13 questions, and the second questionnaire, the PSQI, wich evaluates indirectly, their sleep quality in the last month. After these answers, the results from each of the studied populations were compared are still not problematic and the surroundings of the wind parks generate an important level of noise, but it does not influence nearby populations (Engineering, 2016).
The analysis of noise maps shows that the municipality of Castelo de Paiva presents, in a part of its territory, low noise levels (RNT, 2019). The highest noise levels are registered in the surroundings next to the main national roads that serve the municipality affecting, at most, the first front of houses (RNT, 2019).
In this paper we aimed to find a correlation between living in the district of Porto, and because of that, being exposed to high levels of sound pollution, and the development of bad sleeping patterns using the Pittsburgh Sleep Quality Index (PSQI). To serve as control group, a sample of people from rural areas of northern Portugal, Cinfães and Castelo de Paiva, was also evaluated.
The choice of these two rural areas was due to their greater proximity to the authors, which facilitated data collection. The secondary objective is to study the factor adjustment to sound pollution as a possible influence in the sleep quality, particularly if by adjusting ourselves to this type of pollution over the years, our sensibility to it decreases provoking a less affected sleep or, at least, a loss on the perception of the pollutant.  and analyzed. The selection of the sample used was a sampling by convenience, based on the following inclusion and exclusion criteria (Table I).

Methods to collect the data
The data from the participants was collected "via 2 online questionnaires", and the data from the levels of sound pollution in each studied area was collected from the city's responsible authorities in the subject, most of them, available online to the general public. The

Portuguese version of the PSQI was translated by Karina
Del Rio, who gave permission to use the questionnaire in the present study. The PSQI includes a scoring key to calculate seven subscores, referring to subjective sleep quality, sleep latency, sleep duration, sleep efficiency, sleeping disorders, use of sleeping pills and drowsiness and daytime dysfunctions, each of which can range from 0 to 3, with 3 being the poorer result and 0 the normal, non-pathological result, using the participant's answers related to each component. The subscores are tallied, yielding a global score that can range from 0 to 21. When the Global score is less than or equal to 5, according to Karina Del Rio, sleep is considered good, and with a score greater than 5, sleep is considered to be poor quality.

Variables
To study the possibility of adjustment, the subjects were inquired about their years of residence in the reported area. The participants were also asked to report if they considered themselves to be stressed people or not on their daily life. The reported stressed and years of residency variables were accessed in the sociodemographic questionnaire along with age, sex, know auditory disease, medication and area of residency (Table II).

Statistical Analyses
The software used for statistical analysis was Statistical Package for the Social Sciences (SPSS) version 27. The data was analyzed using descriptive statistics. To characterize the quantitative variables, we used measures of central tendency (mean and standard deviation) and, for qualitative variables, absolute and relative frequencies.
To graphically illustrate the variables, we used frequency  explained, information about the goal of the study as well as the importance of the individual's participation, always guaranteeing the confidentiality of the data. The PSQI and sociodemographic data questionnaire answers were submitted to statistical tests using SPSS version 27 that made the characterization of the sample and the assessment of the sleep quality of the participants possible, and those results are as follows.

Studied Sample description
Adjustment is an important factor that influences the effects of noise on the quality of someone's sleep (Muzet, 2011). Even though some investigators claim exposure to road traffic noise doesn't result in adjustment, some recent studies have provided evidence that it does happen sleep (Muzet, 2011). In these studies, adjustment could be observed when subjective sleep parameters were taken into account and would happen from exposures that may vary from weeks to a few months sleep (Muzet, 2011). The variable years of residency was used to document the possibility of adjustment to noise that can occur in residents and that may vary in individuals based on interindividual variations in the sensitivity to noise sleep (Muzet, 2011). As shown, the mean of this variable is above 10 years in both studied areas (Table  III). We could not find a scientific consensus on the time needed to develop adjustment, with it varying from weeks to months or even years in different studies. Thus, even though being statistically significant, the variable years of residency does not present a bias because, if occurring, the adjustment would be present after 10 years of residency. The rest of the variables taken into account do not show statistically significant differences, and so, are not considered biases either.  The results of the global score found that both groups did not have statistically significant differences. The city sample global score was 6.1 (2.4) (Mean (SD)) and the rural sample was 5.9 (2.6) (Mean (SD)) with a P-Value of 0.73. Even though the city global score was higher than the rural global score, which was expected, it represents a residual difference with no statistical significance.
Both groups showed a global score over 5 points, which indicates a poor overall sleep quality, slightly worse in the city group (Table IV).
Even though the 7 components showed a worse result, meaning a generally poorer quality of sleep in the city group, the only statistically significant association found was in the drowsiness and daytime dysfunctions component of the PSQI score (P-value=0.046). As for the 6 other components, no statistical evidence was identified, showing no indications that these parameters are more affected when living in Porto when compared to living in rural areas where the exposure to environmental noise is below 45dB.

Discussion
Our study showed that respondents from city area had significantly more drowsiness and daytime dysfunctions but did not show a bigger predisposition to difficulties falling asleep, a longer time spent in bed, a worse sleeping efficiency, sleeping disorders or use of medication to aid falling asleep (Table IV). They also did not show a significantly poorer sleep quality (PSQI total score). Both samples had a score above 5 points which means a poor sleep quality in the overall population.
In studies on this area, two different approaches and methodologies are commonly chosen, laboratory experiments and field research (Jakovljević et al., 2006), our study falls in the second category. Field research has a powerful advantage over laboratory experiments because noise is an environmental threat that is everywhere in our cities and so it makes the understanding of noise hazards in a daily life setting possible (Jakovljevic et al., 2006). But this method shows disadvantages as well, the most significant being the fact noise is usually only measured outside, making it very hard to access the  actual levels inside people's houses, as well as the fact that subjective measures are not as reliable as objective ones (Jakovljević et al., 2006).
Jakovljević and colaborators performed, in 2006, a similar study. On their research there were no significant differences reported between the samples according to the residential area in the duration of sleep, sleep latency or use of sleeping pills (Jakovljević et al., 2006). The rest of the parameters showed a significant difference (Jakovljević et al., 2006). Correlation analysis in their study showed that noise annoyance was significantly related to all sleep quality parameters except for average duration of sleep and the use of medication to fall asleep and also that subjective noise sensitivity was significantly correlated with sleep latency, tiredness after sleep and the use of said medication (Jakovljević et al., 2006). These parameters, annoyance and subjective noise sensitivity, were not evaluated in our study, which could have given important additional information to understand the results better and should be considered in all future studies in the area. The fact that there were no differences in sleep latency between respondents from the city area and the rural area could be due to adjustment to nocturnal noise, considering long-term residence from the subjects in the respective areas (Jakovljević et al., 2006). The poorer sleep quality and daytime dysfunctions such as tiredness after sleep have been previously reported (Öhrström, 2004). The changes that occur in sleep stages, for example, the shortening of time spent in stage 3 and REM sleep and subsequential increase in stages 1 and 2 may explain these late effects of a poor sleep quality (Jakovljević et al., 2006).
The non-existing difference in the use of sleep medication between the two samples may imply that resorting to this means is more of a personality trait, such as neuroticism, and not as much because of the involving environment (Jakovljević et al., 2006). Highly significant correlation coefficient between neuroticism and use of sleeping pills in Jakovljević and colaborators study supports this assumption, showing that the control of such variables as subjects' personality may have a major importance in contextualizing the results (Jakovljević et al., 2006).
Not only is the level of noise to which the residents are exposed important, but also, the time of day when this exposure takes place, being a factor influencing the quality of sleep (Jakovljević et al., 2006).
On the other hand, Han and colaborators in 2015, in a study comparing residential, construction, commercial and transportation hub areas of China, found that among the four areas, sleep quality was reported worse in the transportation hub area and there was no statistically significant difference between the other three. All areas had noise levels above the regulation limit and the transportation hub area was the worst (Han et al., 2015). This proves that within the city, different areas have different noise levels that may influence a lot the affection on sleep quality, demonstrating that an important consideration to make in future studies is to account for this fluctuation within areas and describe as good as possible where people actually live, unfortunately, in our study due to ethics and data protection polices, it was not possible to specify the participants exact area of residency.
In another investigation carried by Pirrera and colaborators in 2014 using questionnaires as PSQI and the Epworth Sleepiness Scale showed no significant difference in any of the parameters between groups (quiet and noisy). This study also found that outside noise levels were not reflected inside people's bedroom (Pirrera et al., 2014 Preventive measure should be taken focusing on the decrease of nocturnal noise to levels between 30 and 45dB maximum (Jakovljević et al., 2006). Socially responsible departments in our cities should enforce the regulation on noise established by the government and improve the available technology to reduce noise pollution as much as possible and assure the well-being of the population (Han et al., 2015). The residents themselves should selfeducate on the matter and take all measures available to reduce the hazards of this pollutant (Han et al., 2015). We hope that further research provides more information on policies/interventions as well as health impacts evaluations (Han et al., 2015).
Because this study was performed, in part, during a national lockdown due to the COVID-19 pandemic, people's sleeping patterns may be altered to some degree due to higher levels of stress, different sleeping schedules provoked by working from home, or not being at work at all, a completely different daytime and social life and so many other factors that the pandemic brought on. Also, if our study had a more representative sample of the population, perhaps the results would have been more significant, given their tendency to be worse on the city group. Not only improving the amount of people in the samples but also only using residents specifically from the city of Porto, which was not possible, could have made a difference. The loss of volunteers on the process of emailing the second questionnaire after the sociodemographic data one could have been avoided if we had presented both questionnaires on the first contact with the individual.
This study, because it was one of the first performed in Porto, opens the path to more investigation on the subject not only in this city but in Portugal, with special attention to the recommendations on this discussion. We showed that noise has an implication on people's sleeping patterns as it affects their daytime life and productivity and has to be better controlled by the authorities to give the residents the best environmental conditions possible to not affect their health negatively.