Certain patterns emerge when one takes a closer look at human behaviour.

This is to be expected, as specific stimuli tend to elicit similar responses. We also know that people can be categorised into population groups based on their shared characteristics; each group shares a common denominator and features that distinguish it from the others. Therefore, individual behaviour cannot necessarily be viewed separately from behavioural patterns in the population, given that our social context influences and directs our behavioural choices through various norms and rewards, creating behavioural opportunities and causing or reducing stress.

Since much of the population’s health loss is caused by diseases that are either avoidable or preventable (Alwan et al. 2010), social context also has an essential role in explaining why some people remain healthy while others become ill. This process is far from random: it is shaped by the interplay of resources provided by the surrounding environment and individual behavioural choices and opportunities.
The same pattern applies to health behaviour and therefore helps explain the relationship between lifestyle and mental health. However, the social environment is not something that remains constant over time. Let’s consider, for example, how our own daily life has changed over the past few decades. Since changes in the health behaviour of the population form the starting point for changes in physical and mental health, the current situation in population’s health behaviour and mental health should be viewed in the context of the developments in the past.

This article aims to describe and analyse the relationship between lifestyle and mental health in the adult population of Estonia. The empirical part of the article is based on the 1990 – 2020 data from the survey Health Behaviour among Estonian Adult Population and examines associations between lifestyle indicators and mental health outcomes over the past 30 years. By focusing on the indicators of health and risk behaviour and manifestations of their unequal distribution, this article aims to provide a brief insight into the relationships between lifestyle and mental health in adulthood and how their patterns in Estonia have changed over time.

While there have been several surveys among the Estonian population that could be used to describe the relationship between lifestyle and mental health, the survey offers the longest timeline.Health Behaviour among Estonian Adult Population (TKU). This cross-sectional, population-based survey conducted every two years among Estonian residents aged from 16 to 64 first took place in 1990. Its data allow us to describe several indicators of health status and health behaviour and their determinants over the past 30 years.

Since the question about self-reported stress¹ has been phrased in the same way throughout all the waves of the study, this indicator provides a good basis for discussing the long-term trends of mental health. The long-term prevalence trends of stress and multiple health behaviour indicators² (vt Figure 2.3.1) reveal several interesting changes. The prevalence of unbearable or excessive stress was lowest (16%) in the 1990 survey but increased rapidly in subsequent years, reaching 28% in 1996. The prevalence of stress decreased in the first years of the 2000s, reaching 18% in 2006. However, the prevalence of stress increased again in the following years correlating with the onset of economic crisi. There have been no major changes during the past decade; on average, one in five adults experiences excessive stress.

library(ggplot2)
library(tidyr)
library(scales)
library(dplyr)

## 
## Attaching package: 'dplyr'

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

#faili sisselugemine ja andmete formaadi korrigeerimine
J231=read.csv("PT2-T2.3-J2.3.1.csv",header=TRUE, encoding ="UTF-8")
J231x=J231[1:5,1:17]
J231x=pivot_longer(J231x,2:17)

J231y=J231[8:17,1:17]
J231y[2,1]=J231y[1,1]
J231y[4,1]=J231y[3,1]
J231y[6,1]=J231y[5,1]
J231y[8,1]=J231y[7,1]
J231y[10,1]=J231y[9,1]
J231y$Cl=rep(c("low","high"),5)

J231y=pivot_longer(J231y,2:17)
J231y=pivot_wider(J231y,names_from=Cl,values_from=value)

J231=cbind(J231x,J231y[,3:4])

names(J231)[1:2]=c("Faktor","Year")
J231$Year=sub("X","",J231$Year)
sildid=na.omit(J231)%>%group_by(Faktor)%>%filter(Year%in%c(min(Year),max(Year)))

#joonis
ggplot(J231,aes(x=Year,y=value,col=Faktor))+
  geom_line(aes(group=Faktor), linewidth=1.5)+
  geom_point(aes(group=Faktor),cex=2)+
  geom_label(data=sildid,aes(x=Year,y=value,label=paste(round(value),"%",sep="")),cex=3,show.legend = FALSE)+
  theme_minimal()+
  theme(legend.title=element_blank())+
  theme(legend.position = "bottom")+
  ylab("")+
  guides(color = guide_legend(nrow=2,byrow = TRUE)) +
  scale_color_manual(values=c("#1E272E","#FF3600","#6666cc","#668080","#f09d00"))+
  theme(text = element_text(color="#668080"),axis.text=element_text(color="#668080"))

## Warning: Removed 6 rows containing missing values (`geom_line()`).

## Warning: Removed 6 rows containing missing values (`geom_point()`).

If we add certain health behaviour trends to the mix, we see even more significant changes over the past three decades. In 1990, daily smokers made up 30% of the Estonian population aged from 16 to 64. This percentage remained stable until the middle of the next decade, when the share of daily smokers in the population started a consistent decline. The trend of alcohol use can be observed since the survey of 1996, when 21% of the population consumed alcohol once a week or more. Although there has been no significant change in the frequency of alcohol use over time, the proportion of frequent alcohol consumers was somewhat higher between 2006 and 2012. At that time, an average of 29% of Estonian residents aged from 16 to 64 consumed alcohol at least once a week. The prevalence of excess body weight can be examined through the total share of overweight (KMI³ = 25,0–29,9 kg/m2) and obese (KMI ≥ 30,0 kg/m2) adults in the population, and this has consistently increased over time. Meanwhile, the increase in the share of adults engaged in recreational sports indicates that general awareness of the importance of healthy behaviours has grown. Although the opposite trajectories of these two indicators may come as somewhat unexpected, they illustrate, on the one hand, the versatility of health determinants and, on the other hand, the differences in health behaviour between population groups. The increase in the proportion of overweight and obese people in the general population does not necessarily indicate a similar trend among those who have increased their physical activity.

Since the changes in the time series of these indicators at least partially overlap with several significant events or stages for Estonia, the following analysis will look at four distinct periods. The period from 1990 to 1996 marks a time of social transition and also reflects the greatest relative change in the prevalence of stress. The period from 1998 to 2004 covers transformative years for Estonia’s development, including the economic crisis of 1998 and the country’s accession to the European Union and NATO in 2004. The period from 2006 to 2012 follows the years of rapid economic growth and the severe economic crisis that followed, while the last period (from 2014 to 2020) is distinguished by the COVID-19 pandemic that reached Estonia in the spring of 2020 (at the time of data collection).

Analysing the associations presented in in Figure 2.3.2 shows that the discussed health behaviour indicators are associated with higher stress levels, but the strength of the association (and its statistical significance) varies from period to period. While the adjusted model covering the whole period shows that, compared to occasional smokers and non-smokers, daily smokers have 1.3 times the odds of having higher stress levels, the same odds ratio was the lowest and statistically non-significant for the period from 1990 to 1996, which had the highest prevalence of daily smokers. In all subsequent periods, however, daily smoking was associated with higher stress levels. For alcohol, the model spanning the entire period showed that drinking once a week or more was associated with 1.2 times the odds of greater stress compared to drinking alcohol less frequently. This relationship has been statistically significant since 2006. When it comes to body weight, the most conspicuous risk factor for stress is obesity. The subgroup of obese people had 1.2 times the odds of experiencing stress as adults with normal weight. However, if we look at this relationship across the different periods, the relationship between obesity and high stress levels was statistically significant only in the last two periods. A similar relationship appeared in connection to recreational sports. Adults who exercised less than once a week had approximately 1.2 times the odds of feeling stressed than those who exercised once a week or more. The odds ratio is the highest in the first period, while it is insignificant in the period from 1998 to 2004.