Sodium and Potassium Intake, Knowledge Attitudes and Behaviour Towards Salt Consumption Amongst Adults in Podgorica, Montenegro

Excess salt and inadequate potassium intakes are associated with high cardiovascular disease (CVD). In Montenegro, CVD is the leading cause of death and disability. There is no survey that has directly measured salt and potassium consumption in Montenegro. The aim is to estimate population salt and potassium intakes and explore knowledge, attitudes and behaviour (KAB), amongst the adult population of Podgorica. Random samples of adults were obtained from primary care centres. Participants attended a screening including demographic, anthropometric and physical measurements. Dietary salt and potassium intakes were assessed by 24 h urinary sodium (UNa) and potassium (UK) excretions. Creatinine was measured. KAB was collected by questionnaire. Six hundred and thirty-nine (285 men, 25–65 years) were included in the analysis (response rate 63%). Mean UNa was 186.5 (SD 90.3) mmoL/day, equivalent to 11.6 g of salt/day and potassium excretion 62.5 (26.2) mmoL/day, equivalent to 3.2 g/day. Only 7% of them had a salt intake below the World Health Organization (WHO) recommended target of 5 g/day and 13% ate enough potassium (>90 mmoL/day). The majority (86%) knew that high salt causes ill-health. However, only 44% thought it would be useful to reduce consumption. Salt consumption is high and potassium consumption is low, in men and women living in Podgorica.


Introduction
Non-communicable diseases (NCDs) are the leading, yet preventable, causes of death worldwide [1]. In Montenegro, NCDs are a major public health challenge undermining socio-economic development [2]. According to the World Health Organization (WHO) latest estimates [2], NCDs account for 92% of total deaths, out of which 22% are premature (occurring before the age of 70 years) and 61% due to The survey took place between 15 th September and 15 th December 2017. From the 1008 households and individuals interviewed in the sampling frame, 639 of them (63%) provided suitable data for inclusion in the survey analysis. Originally, 6 did not meet the inclusion criteria, 99 did not provide date and time of the urine collections, 12 had missing data, 6 provided urine collections with volume less than 500 mL (conventionally taken as not plausible), 237 (23.5%) provided either undercollections (<22 h) or over-collections (>26 h) and 9 had urinary creatinine excretion outside 2 standard deviations (SDs) of the sex-specific distribution of urinary creatinine in the sample ( Figure  2).  The survey took place between 15th September and 15th December 2017. From the 1008 households and individuals interviewed in the sampling frame, 639 of them (63%) provided suitable data for inclusion in the survey analysis. Originally, 6 did not meet the inclusion criteria, 99 did not provide date and time of the urine collections, 12 had missing data, 6 provided urine collections with volume less than 500 mL (conventionally taken as not plausible), 237 (23.5%) provided either under-collections (<22 h) or over-collections (>26 h) and 9 had urinary creatinine excretion outside 2 standard deviations (SDs) of the sex-specific distribution of urinary creatinine in the sample ( Figure 2). Nutrients 2018, 10, x FOR PEER REVIEW 3 of 12 those with known history of heart or kidney failure, stroke, liver disease, those who recently began therapy with diuretics (less than two weeks), pregnant women, any other conditions that would make 24 h urine collection difficult. The sample for the survey was selected with a sex-and age-stratified random sampling of men and women in six facilities of the PHC of Podgorica ( Figure 1). The survey took place between 15 th September and 15 th December 2017. From the 1008 households and individuals interviewed in the sampling frame, 639 of them (63%) provided suitable data for inclusion in the survey analysis. Originally, 6 did not meet the inclusion criteria, 99 did not provide date and time of the urine collections, 12 had missing data, 6 provided urine collections with volume less than 500 mL (conventionally taken as not plausible), 237 (23.5%) provided either undercollections (<22 h) or over-collections (>26 h) and 9 had urinary creatinine excretion outside 2 standard deviations (SDs) of the sex-specific distribution of urinary creatinine in the sample ( Figure  2).   The excluded participants had similar anthropometric characteristics and blood pressure values but were younger and has faster pulse rate than those included in the final sample (see Supplementary  Tables). The survey was carried out in accordance with the Declaration of Helsinki and Good Clinical Practice [29]. Ethical approval for the survey was obtained from the Committee of Research Ethics of the National Centre of Public Health and participants provided written informed consent to take part.

Data Collection
The examination was performed in a quiet and comfortable room, with the participants who were not allowed to smoke, exercise, eat, consume caffeine and to have a full bladder for 30 min before measurements. The survey was carried out in three steps: (a) questionnaire survey, (b) physical measurements and (c) 24 h urine collections.
The questionnaire (face-to-face interview, adapted version of the WHO STEPS Instrument for NCD Risk Factor Surveillance) [30] was used to collect data on respondent's demographic and socio-economic status (by occupation and educational attainment); diet, frequency of high salt food consumption, fruit and vegetable consumption, knowledge attitudes and behaviour on dietary salt, history of high BP, diabetes and CVDs, lifestyle advice.
Anthropometric indices, BP and heart rate were measured in all participants. Height was measured in cm with a laser stadiometer (SECA 709). Body weight was measured in kg using a digital electronic scale (Transtek, model GBS-721). Body mass index (BMI) was calculated as weight (kg) divided by height squared (m 2 ). Overweight was defined as a BMI of 25 kg/m 2 or more and obesity as a BMI of 30 kg/m 2 or more. Systolic and diastolic BP and heart rate measurements were taken three times in the right arm on a sitting position, using a universal cuff and automatic BP and heart rate monitors (Boso Medicus Uno, Bosch+Sohn GmbH, Jungingen, Germany). The first measurement was ignored, the mean of second and third measurements being taken for analysis. The measurements were taken after the participant had rested for 15 min and each with three minutes of rest between the measurements (maximum deviation of cuff pressure measurement ± 3 mmHg and of pulse rate display ± 5%). Hypertension is defined as systolic and/or diastolic BP ≥ 140/90 mmHg or regular antihypertensive treatment [31].
A single 24 h urine collection was obtained from the participants. Each participant was given a leaflet with explanations along with the necessary equipment and a record sheet on which participant noted the start and the finish times of their urine collection, any missed urine aliquots and any medication taken during the collection. The participants were carefully instructed on urine collection methodology [27,28]. In an effort to minimize bias, participants were also requested not to change their diet before or during the day of the urine collection. The first void upon waking on the day of collection was discarded. The urine volume of the 24 h collection was measured by field team-members and a urine sample was stored in a cool place for a maximum of 24 h until transportation to the laboratory. Sodium, potassium and creatinine determinations were carried out immediately [27,28]. Sodium and potassium concentration in the urine samples were determined using a Ion Selective Electrode with a Beckman Coulter Synchron CX5PRO System and expressed in mmoL/L [32]. Creatinine concentration was determined through the Creatinine (urinary) Jaffé kinetic method and expressed in mg/dL [33].

Statistical Analysis
All statistical analyses were performed using the SPSS software, version 20 (SPSS Inc., Chicago, IL, USA). To detect approximately 1 g reduction in salt intake over time using 24 h urinary sodium excretion (difference~20 mmoL/24 h), with a standard deviation of 75 mmoL/day (alpha = 0.05, power = 0.80), a minimum sample of 120 individuals per stratum is recommended [27,28]. Thus, a minimum recommended sample size of 240 was estimated per age and sex groups and adjusted for an anticipated non-response rate of 50% [27,28]. The population was stratified in groups by sex (men and women) and by age (I: 25-34 years, II: 35-49 years, III: 50-65 years). T-test for unpaired samples or analysis of variance (ANOVA) was used to assess differences between group means. Pearson chi-square test was used to test the association between categorical variables. Linear regression analysis was used to detect the association between salt and potassium consumption with socio-economic status, adjusting for age and sex. To convert urinary output into dietary intake, the urinary excretion of sodium (UNa) or potassium (UK) values (mmoL/day) were first converted to mg/day. Then, sodium values were multiplied by 1.05 (assuming that only 95% of sodium ingested is excreted), while potassium values were multiplied by 1.3 (on the assumption that only 70% of the potassium ingested in excreted in the urine [34]. The conversion from dietary sodium (Na) intake to salt (NaCl) intake was made by multiplying the sodium value by 2.542. The results were reported, as appropriate, as mean (SD and/or 95%CI) or as percentages. Two-sided p below 0.05 were considered statistically significant.

Characteristics of the Participants
The characteristics of the participants are shown in Table 1. There was no statistically significant difference in the mean age between male and female participants, however men were significantly taller and heavier than women and had a higher systolic, diastolic BP and pulse rate. The prevalence of hypertension was on average 36.4% (225/618), comparable in men (110/279 or 39.4%) and women (115/339 or 33.9%; P > 0.05).

Daily Urinary Excretions of Volume, Sodium, Potassium and Creatinine and Salt and Potassium Intake
Average urinary volume excretion was 1629 mL per day, being higher in men than women ( Table 2). Average urinary creatinine excretion was 1.41 g per day, being again higher in men than women (Table 2). Urinary sodium excretion showed a normal distribution with a tail skewed to the right (i.e., towards higher values). Mean urinary sodium was 186.5 (SD 90.3) mmoL/24 h (Table 2), equivalent to a mean consumption of 11.6 (5.6) g of salt per day (Table 2). Men excreted more sodium than women (mean difference 63.9 mmoL/24 h, P < 0.001), equivalent to 4.0 g of higher salt consumption than women. Only 46 participants (7%) met the levels of salt intake of 5g or less recommended by the WHO, significantly less in men than women (n = 11 or 4% vs n = 35 or 10%, respectively, P = 0.003). Urinary potassium excretion showed a normal distribution with a tail skewed to the right (i.e. towards higher values). Mean urinary sodium was 62.5 (26.2) ( Table 2), equivalent to a mean consumption of 3.2 (1.3) g of potassium per day ( Table 2). Results are mean (SD); † P < 0.001; * P < 0.02 vs men.
Men excreted more potassium than women (mean difference 10.9 mmoL/24 h, P < 0.001), equivalent to 0.6 g of higher potassium consumption than women. Only 79 participants (13%) met the levels of potassium intake of 90 mmoL/day or more recommended by the WHO, significantly less in women than men (n = 26 or 8% vs n = 53 or 19%, respectively, P < 0.0001). No differences in salt and potassium consumption were detected by age (see Supplementary Tables). Salt and potassium consumption varied by socio-economic status, both measured by educational attainment and by occupation (Figure 3). After adjustment of age and sex, salt consumption was significantly lower in participants with the highest educational attainment (P for trend = 0.02) and it was the highest in retired and unemployed (P for trend = 0.02), suggesting a relationship between high salt intake and low socio-economic status. On the contrary, potassium consumption showed a direct relationship with educational attainment (P for trend = 0.05) and with current employment (P for trend = 0.05), suggesting a lower potassium consumption in more disadvantaged socio-economic groups. occupation (Figure 3). After adjustment of age and sex, salt consumption was significantly lower in participants with the highest educational attainment (P for trend = 0.02) and it was the highest in retired and unemployed (P for trend = 0.02), suggesting a relationship between high salt intake and low socio-economic status. On the contrary, potassium consumption showed a direct relationship with educational attainment (P for trend = 0.05) and with current employment (P for trend = 0.05), suggesting a lower potassium consumption in more disadvantaged socio-economic groups.

Knowledge, Attitude and Behaviours Towards Salt Intake
A high proportion of participants (73.8%) reported adding salt to food at the table often or always, more amongst women compared to men (79.1 vs 67.2%; P = 0.0007) (Table 3). However, only 21.4% said they would often or always add it when cooking, especially men (25.2 vs 18.3%; P = 0.03). The majority (85.8%) knew that high salt causes serious health problems. However, only 44.2% thought it would be useful to reduce salt consumption, 35.6% felt their intake was excessive and 34.3% were doing anything to try and reduce it. Women were more pro-active than men in doing so (39.8 vs 27.6%; P = 0.001).

Discussion
This is the first survey on salt and potassium consumption carried out in adults in Montenegro, using the gold standard measure of 24 h urinary sodium and potassium excretions as biomarkers of intake. The results show unequivocally that salt consumption is high and that potassium consumption is low, both in men and in women.
Average salt intake exceeded by over twofold the WHO recommended maximum population target of 5 g per day [12]. Only 7% of the participants met the WHO salt targets. Men excreted more sodium than women. Discretionary use of salt is very common in Montenegro, with 3 out of 4 participants adding salt regularly to food and 1 in 4 also using it regularly when cooking. The majority of participants knew that high salt causes serious health problems. However, only less than half thought it would be useful to reduce its consumption, 1 in 3 felt own intake was not excessive and only 1 in 3 was reporting doing anything to reduce it. The answers to these questions reveal an insufficient level of knowledge of the real problem associated with high salt consumption amongst the participants of Podgorica and the widespread unreadiness to transfer this knowledge to behavioural changes in using discretionary salt. Average potassium intake was low in Montenegro, less than half the WHO recommended minimum population target of 90 mmoL/day or more [23]. Only 13% met WHO potassium targets.
Our main findings point out that the salt intake in Montenegro is higher than those reported in many neighbouring countries, both in men and women [35,36]. In the recent MINISAL study, the daily urinary sodium excretion of Italians was 189 for men and 147 mmoL/24 h for women [34] and in Greece 194 and 158 mmoL/24 h for male and female participants, respectively [36]. On the other hand, the potassium intake was similar to that found in Greece [36] but higher than that measured in Italy [35]. As reported in the majority of previous studies [37], men consume more salt than women. This is almost certainly explained by greater total food consumption. As expected, there are gender differences in both salt and potassium consumption, as detected in many other population surveys. Salt and potassium are expressed as total quantities rather that consumption per calorie intake, hence the gender difference is mainly explained by the larger body size of men compared to women and the corresponding total food consumption compared to women. However, there were gender differences in attitudes and behaviours towards salt consumption. Women reported more often less use of discretionary salt and more readiness to doing something to control salt intake. These could contribute, at least in part, to the differences found.
Social inequalities are important determinants of ill health globally [38,39]. People from low socio-economic background die sooner and have more disabilities. They are more likely to depend of cheaper, unhealthy food rich in salt. In the UK, their knowledge of public health recommendations is lower and the use of table salt and total salt intake is higher [34,40,41]. In England and Wales, the salt reduction of 1.4 g per day seen recently has not resulted in a reduction in socioeconomic differences [42]. "Upstream" health policies tend to reduce health inequalities [43]. Furthermore, an improvement in socio-economic position decreases the risk hypertension that is seen with low parental social status [44].
Socioeconomic status is a major determinant of excessive dietary salt and insufficient potassium intake [34,41,42]. Indeed, the healthy foods are more expensive than Western dietary patterns [45], that is, more energy-dense and nutrient-poor foods [46], that leads to a high availability of poor-quality foods. Also, low level of education unfavourably contributes to higher salt and lower potassium intake [34,41,42,47]. Hence, the interaction of low income and low levels of education contribute together to a low adherence to healthy eating pattern. Our results confirm this trend even in Montenegro, where unemployed subjects or participants with low level of education consumed more salt and less potassium. These conditions, in turn, would contribute to increase the otherwise avoidable burden of CVD. Currently, in the European region, a gradient across measures of inequalities is present even for the degree of implementation of current salt reduction policies [48].

Strengths and Limitations
Strengths of our study are the inclusion of a large random sample of male and female participants as representative of the capital city of Montenegro, the assessment of average consumption of salt and potassium by 24 h urine collection (gold-standard measure of salt and potassium intake) [28], the rigorous quality control and the careful standardized protocol of urine collections. In particular, the rigorous instructions to ensure completeness of urine collections and the strict protocol to select for analysis only those fulfilling the quality control criteria among which the length of collection time and the assessment of urinary creatinine excretion, markers of the accuracy of the collection. However, the study has some limitations. The possibility of selection bias cannot be excluded. Participants not delivering complete urines were younger than those complying. However, no other differences were seen in body size and cardiovascular parameters. Urinary sodium and potassium excretions were only assessed once. Given the large within-subject variability in consumption, it is acknowledged that a single collection is unlikely to characterize an individual's salt and potassium consumption [49]. However, group estimates are less likely to be biased by this variability. Finally, sampling was drawn from the capital city of Podgorica. Whilst this represents a valid random sample, it may not be fully representative of the entire country, consisting of a mountainous Northern Region and a small Coastal Region on the Adriatic Sea. Montenegro is a small country with an estimated population of approximately 640,000, of which approximately 150,000 resides in the capital and surroundings [50]. The recent and continuous migration to the capital city makes the population of Podgorica rather heterogeneous, although some differences in dietary habits between the different regions may still exist.

Conclusions
Salt intake is high and potassium is low in Podgorica, Montenegro. Awareness, attitudes and behaviours about salt and its implication for health suggest that there is an urgent need for intensive awareness campaigns and health promotion to improve the take up of preventive strategies aiming at reducing salt consumption, whilst at the same time increasing potassium intake by encouraging higher consumption of potassium-rich food. Awareness about hidden salt in processed food should be highlighted. A national program for reducing salt intake in Montenegro needs to be implemented through systematic efforts and multisectoral collaboration including not only public awareness and behaviour-change communication (including via health care professionals), but, more importantly, structured programs for reformulation that set the framework for the food industry to reduce salt in processed foods, major source of salt intake. The current Montenegrin approach to reducing population salt intake is based predominantly upon voluntary food supply reformulation and the Government has already introduced a regulation on maximum salt content in bakery products. Based on international experience and using national dietary intake data, the government can set food category-based salt reduction targets for the most important foods that contribute to salt intake among the Montenegrin population, with the objective of contributing to a 30% reduction in salt consumption by 2025. Only one in two people consume sufficient amount of potassium and public policies should be directed to encouraging an increase intake of fruit, vegetable, pulses and nuts by setting targets to increase consumption to 90 mmoL per day or more. In the absence of a more comprehensive national survey of habitual salt and potassium intake in Montenegro, our data provides a useful baseline against which to monitor the impact of future initiatives.
Author Contributions: F.PC. developed the protocol, trained local teams, co-ordinated quality control and data analysis and drafter the manuscript. L.D. carried out quality control and statistical analysis. M.B. and A.K. coordinated the study, carried out the fieldwork and liaised with the local laboratory. V.C. helped with the drawing of the stratified random sample from the sampling frame and contributed to statistical analysis. J.J. and J.B. contributed to the design and interpretation of the findings. All authors contributed significantly to the final version of the manuscript.