Community awareness of dental fluorosis as a health risk associated with fluoride in improved groundwater sources in Mangochi District, Malawi

Fluoride contamination in drinking water is a serious concern globally because of its long-term effects on human health. Fluoride consumption is considered good for healthy development of teeth and prevention of tooth decay in children if consumed in concentrations of 0.5–1.5 mg/L. Consumptions below 0.5 mg/L have negative health outcomes in form of dental caries development (Siaurusevičiūtė & Albrektienė 2021). Excessive ingestion of fluoride in concentrations higher than 1.5 mg/L, which mainly is through drinking water consumption, may cause dental, skeletal and neurological disorders (Woldeyes et al. 2004; Sajidu et al. 2008; Dong & Wang 2016). Table 1 summarises the fluoride concentration ranges and their effects on human health.

Fluoride is one of the two priority chemicals, arsenic being the other, mainly being considered when it comes to monitoring the progress of sustainable development goal (SDG) target for drinking water (WHO/UNICEF 2017). The World Health Organisation (WHO) regulates the fluoride intake by setting 1.5 mg/L as the guideline value for fluoride in drinking water (WHO 2017). On the other hand, the Malawi Standard (MS733-2005) provides for 6 mg/L as the maximum limit for fluoride concentration for boreholes and shallow wells. Although fluoride is traced in many waters, concentrations of up to 10 mg/L, or even higher, are usually found in groundwater which interacts with minerals that harbour fluoride, such as fluorspar, cryolite and fluorapatite in the earth's subsurface (WHO 2017). Fluoride concentration depends on a number of factors namely geological, chemical and physical characteristics of the aquifer, porosity and acidity of the soil and rocks, temperature and depth of the wells (Sajidu et al. 2008) and residence time (Brunt et al. 2004).

Groundwater is the most relied-upon water source for Sub-Saharan Africa (SSA) and the rest of the developing world (Xu et al. 2019). However, studies have reported the presence of fluoride in groundwater from SSA, with the highest levels being registered in the East African Rift System (EARS) (Onipe et al. 2020). Malawi, being part of the EARS, is therefore susceptible to fluoride contamination, although it is not well documented (Addison et al. 2020a). This shows that groundwater sources such as boreholes and protected shallow wells, although categorised as improved water sources, may be susceptible to fluoride contamination if constructed in zones of high groundwater fluoride concentrations. By definition, improved water sources refer to sources that are potentially capable of delivering safe water by nature of their design and construction, which include piped water, boreholes or tube wells, protected dug wells and protected springs (WHO and UNICEF 2017). Literature has shown that an improved water source may not always be safe (Bain et al. 2014; Shaheed et al. 2014), yet the water users may perceive their water source to be safe, especially if it is the only reliable water source for them (Grupper et al. 2021).

Past studies revealed the presence of fluoride in concentrations higher than 1.5 mg/L in some groundwater sources such as boreholes and shallow wells in Mangochi district in Malawi (Sajidu et al. 2008; Chimphamba & Phiri 2014; Government of Malawi 2021). However, the studies have been patchy and localised, such that their findings may not be generalised but only act as a pointer to a serious problem that needs further investigation. In addition, there was no study that assessed the community awareness of fluoride present in drinking water and its associated health risks with a focus on dental fluorosis in Mangochi district. This study, therefore, aimed at filling the knowledge gap by assessing the level of groundwater fluoride contamination and community awareness of associated health risks in Mangochi district. This study was necessary for better characterisation of peoples' exposure to fluoride contamination, thereby informing policy and programming that aims at mitigating the health risks.

Mangochi is a lakeshore district at the southern tip of Lake Malawi. The district is located between latitudes 13° 28′ 59″ S and 14° 47′ 53″ S, and longitudes 34° 43′ 20″ E and 35° 46′ 11″ E, with minimum, maximum and average elevations of 473, 1980 and 726 meters above sea level (masl), respectively (topographic-map.com n.d.). The hydrogeology of the district is characterised by two main aquifer types namely: (1) unconsolidated alluvium which mainly lies in the lakeshore plains of Lake Malawi and Lake Malombe as well as along the Shire river; and (2) basement aquifers which lie mainly in the upland areas extending eastwards and westwards of the district and some hilly land portions that extend into Lake Malawi (Upton et al. 2018). Mangochi district has a total land area of 6,729 km², constituting the largest portion at 21.2% of the Southern Region's land area and 7.1% of Malawi's total land area. The district's population, which was 1,148,611 as of 2018 and projected to 1,323,159 as of 2022, is the largest in the Southern Region and second largest in Malawi, constituting 14.8 and 6.5%, respectively (National Statistical Office 2019).

Mangochi district was chosen because of its susceptibility to high fluoride levels in its groundwater for being in the alluvial plain of the upper Shire Valley (British Geological Survey 2004). At the same time, the district has been investing highly in groundwater infrastructure development (Andreah 2022) with support from different partners. Ideally, this scenario calls for attention as argued by Salem et al. (2022) that before developing more new infrastructure, there is a need to enhance effectiveness in water interventions.

In order to assess the prevalence of dental fluorosis and community awareness, users of the water points with fluoride concentrations exceeding Malawi standard value of 6 mg/L were identified for interviews through snowball sampling method because it was difficult to acquire the list of users of the water points. In snowball sampling, the researcher uses the first identified individuals as informants to identify other qualifying individuals, who in turn also identify others, after the required information is collected from them. Snowball method is useful where access is difficult, the topic is sensitive or little is known about the group under study (Cohen et al. 2000; Kumar 2011).

The ExStik FL700 fluoride meter from ExTech Instruments was used to conduct onsite fluoride testing of water. Total ionic strength adjustment buffer (TISAB) tablets were used as reagents. For values exceeding the meter's operating range of 0.00–9.99, deionised water was used for dilution. Based on the instrument's user guide, the fluoride meter was calibrated between 1 and 10 ppm fluoride ion by using 1 and 10 ppm fluoride standard solutions which were prepared by dissolving one TISAB tablet in 20 mL of 1 ppm fluoride standard and another tablet in 20 mL of 10 ppm fluoride standard, respectively. In order to obtain representative groundwater samples, sufficient well purging was done before any sample was collected. Samples were collected and analysed in triplicates; however, the triplicate samples were considered as ‘single sample’ after analysis and averaging. For instance, there were 82 water points coded from WP001 to WP082, each visited twice for water testing, representing 164 visits or 164 samples. The testing for fluoride was conducted in March for the wet season and in October for the dry season. The waterpoint codes were used throughout the study to avoid bias and the actual area details were only revealed after analysis.

The study further examined the prevalence of dental fluorosis as a qualitative way of validating the fluoride testing results. Although dental fluorosis can still be observed in areas with concentrations above 1.5 mg/L and below 6 mg/L, a decision was made to focus only on the cases that would be above 6 mg/L as Malawi's standard maximum value for fluoride in boreholes and shallow wells. With permission from and in presence of parents at the water points, physical examinations were conducted on the children that were available at the time of the study in order to examine the prevalence of dental fluorosis. A total of 75 children aged between 3 and 13 were checked for brownish teeth discolouration.

Parents of the 75 children were interviewed to assess the level of their awareness of the cause of dental fluorosis in their children and a total of 41 parents (1 male and 40 females) participated. As part of the interviews, the participants were asked to confirm if the respective water point with high fluoride concentration was their only source of drinking water and for how long they had been using it so that there could be meaningful inference.

Data were analysed using Statistical Package for Social Sciences (SPSSv20). Descriptive statistics, especially frequencies, were used. Having realised that the fluoride concentration data were not normally distributed, non-parametric tests such as one-way analysis of variance (ANOVA) on ranks using Kruskal–Wallis Test and independent samples Mann–Whitney U Test were performed to test for spatial and seasonal variations respectively. Kruskal–Wallis test is used to statistically measure significant differences between three or more independent samples, while Mann–Whitney U test is used to statistically measure any significant difference between two independent samples (Cohen et al. 2000). In both cases, the analyses are based on rankings, in which the data are jointly ranked as belonging to a single sample, by assigning rank 1 to the item with the lowest value, rank 2 to the next higher item and so on until a summed-up rank is obtained (Kothari 2004).

In this study, each water point was supposed to have two sample sets, one for the wet season and one for the dry season; however, 162 sample sets (n = 162) were analysed instead of 164 (82 × 2) because two boreholes were found non-functional in October. In this study, each sample set was treated and analysed individually and later corresponded to individual water points.

Fluoride testing was conducted in March 2020, which was considered as the saturation point for the wet season, and in October 2020 as the peak of the dry season. Examination of children and interviews with parents were later conducted in July 2021 as a follow-up study.

Further analysis through box plotting showed that there were four extreme outlier cases, and these are values for both wet and dry seasons at WP055 (Mulopa-Balamanja) and WP081 (Kwipulumbu). The results show that these boreholes had extremely high fluoride concentrations in both seasons as compared to the rest of the waterpoints. WP055 registered 19.2 and 17.2 mg/L while WP081 registered 27.2 and 19.8 mg/L in wet and dry seasons, respectively.

Of the 75 children aged between 3 and 13 that were present at the time of the follow-up study, dental fluorosis was evident among 62 children representing 82.7% prevalence rate, with 68% being those aged between 6 and 13 years. All the 41 respondents (100%) said they were not aware of any relationship between the water they were using and fluorosis. There was clear evidence that the community members were not aware that the dental fluorosis development in the teeth of their children was as a result of the water they use. Up to 56.1% of the respondents felt that dental fluorosis was mere tinting of teeth caused by eating a wild tropical fruit known by its scientific name as Zanha africana and locally called Ntalawanda. The fruit is usually available in December and kids love to go to the bush to harvest. A 42-year-old father at WP055 in Balamanja village said:

About 14.6% of the participants suspected that the kids' teeth were not properly brushed, as such, they could use sand to polish up their children's teeth as explained by a 37-year-old mother at WP058 in Mankhambira village:

The last group of 12, representing 29.3% completely had nothing in mind, although some thought it was generational changes since older generations did not experience this, probably because they were drinking direct from Lake Malawi and not using groundwater as presented by a 52-year-old mother and grandmother at WP050 in Kholowere village:

The study findings on the levels of fluoride concentrations in some water sources, on one hand, contradict with the prediction of groundwater vulnerability to geogenic fluoride nation-wide in Malawi by Addison et al. (2020b), in which it was predicted that the highest risk zone in Mangochi is ‘elevated geological fluoride risk zone’ where there is more than 60% likelihood of fluoride exceeding 1.5 mg/L and 0% likelihood of exceeding 4 mg/L. This prediction implies that fluoride concentrations for groundwater sources (boreholes and shallow wells) in Mangochi district would fall within Malawi Standard value of 6 mg/L and that the highest that can be recorded is 4 mg/L. This contradiction could be attributed to the fact that the prediction by Addison et al. (2020b) was based on a statistical model, while this study was based on actual observations and measurements. According to Xie (2011), ‘all statistical models have limitations’, and St-Pierre (2016) argues that ‘model predictions and observations are not on the same scale – they are not measuring the same thing’. This would entail results of a prediction model may not be 100% the same as a result from actual measurement. Further to this, the inadequacy of data used in the prediction model may have contributed to its limitation, as alluded to by Addison et al. (2020b) that ‘this prediction may change with increased sampling and testing of the map’. As such, there is a likelihood that some data values from fluoride testing results would still be outside the predicted ranges.

That is why, the findings of this study, on the other hand, agree with other past studies that found fluoride concentration levels above 4 mg/L in some boreholes in Mangochi district. A study by Chimphamba & Phiri (2014) reported borehole water quality data collected by the Ministry responsible for Water Affairs in Malawi since 1980s, and established that out of 5,324 boreholes assessed for the whole Malawi, 139 boreholes had fluoride concentrations exceeding WHO limit of 1.5 mg/L, out of which 63 cases, representing 45%, were registered in Mangochi district with a maximum of 7 mg/L. This study, however, did not indicate how many boreholes had values exceeding Malawi's standard maximum value of 6.0 mg/L. According to Government of Malawi (2021), fluoride concentrations ranging from 1.15 to 14.5 mg/L were detected in 11 boreholes that were tested in TA Chowe and TA Chiunda in the district. Out of the 11 boreholes, 7 had values above 6 mg/L representing 64%, with values ranging from 7.2 to 14.5 mg/L. Results of this nature in Mangochi district are expected due to its location within the alluvial plains. A report by the United Nations in 1989, as cited in Mapoma & Xie (2014), reported fluoride concentrations between 2 and 10 mg/L from the alluvial regions in Malawi. The high fluoride concentrations in groundwater could jeopardise efforts of achieving safe drinking water services in the district in line with the SDG target on drinking water.

The fact that two water points, coded WP055 and WP081, were found as extreme outlier cases in terms of their fluoride concentrations that went as high as between 17.2 and 27.2 mg/L, points to the possibility that these water sources could be located within hot spring zones, although no study has established yet. Studies have shown the relationship between the hot springs and high groundwater fluoride concentrations (Haji et al. 2018; Addison et al. 2020b; Addison et al. 2021). Haji et al. (2018) found that fluoride concentrations in the hot springs of Bilate River Basin of Southern Main Ethiopian Rift could go as high as 57.4 mg/L. It is worth noting that not all hot springs would appear on the surface, rather some may be hidden as the hydrothermal groundwater from depth fails to discharge directly at ground surface and instead discharges at the sediment base and remains buried and hidden from view, thereby mixing with shallow groundwater (Addison et al. 2021).

Despite the Malawi standard value for fluoride in boreholes and shallow wells (6 mg/L) being out of the health range of 0.5–1.5 mg/L, one may argue as to why boreholes with fluoride concentrations higher than 6 mg/L still existed and are in use, against the existence of a standard. This exposes some lapses in monitoring systems and enforcements of such standards. A major challenge is that while the conceptualisation of ‘improved water source’ remains elusive (Manda 2009), monitoring is lacking. A study by Sajidu et al. (2007) also noted that there was no monitoring of chemical water quality in Machinga despite the prevalence of fluorosis in the district being common among both children and adults.

The statistically insignificant seasonal variability could mean that water users have an almost constant level of exposure to fluoride concentration. However, there were higher values in the wet season than in the dry season as also confirmed by the higher mean ranking during the Independent Sample Mann–Whitney U Test. One would expect to see higher concentrations in the dry season and low concentrations in the wet season because it is often attributed to the dilution effect of aquifer recharge from rain input (Okoo 2007; Dibal et al. 2017; Satheeshkumar et al. 2017). However, this finding agrees with findings from other studies that also found higher fluoride concentrations in the wet season than in the dry season (Sreedevi et al. 2006; Battaleb-Looie & Moore 2010; Ghahramani et al. 2020). In the neighbouring district of Machinga, Sajidu et al. (2007) also found seasonal variations in groundwater fluoride concentrations with 26% of the water points under study having higher values in the wet season than in the dry season. Although no further probing was done, the scenario of higher fluoride concentrations in the wet season would in one way be related to fluoride-bearing substances being found at shallower depths such that interaction between the water and the contaminants is more prominent in the wet season due to the rise in water levels; or contamination by rainwater infiltrate that interacts with the fluoride-bearing substances located at a shallower depth and circulates further into the main aquifer. This is achieved through the process of leaching and dissolution of fluoride during the circulation of water in fluoride-bearing rocks and soils, resulting in fluoride enrichment of the groundwater (Brunt et al. 2004; Battaleb-Looie & Moore 2010). However, this suggested reason may still be challenged or confirmed through a further study that aims at identifying the fluoride sources.

The statistically significant spatial variation between TA Katuli and the rest of the TAs could be as a result of differences in geochemical and environmental characteristics of the study area while the insignificant differences among Chilipa, Namavi and Nankumba could be as a result of homogenous geochemical and environmental characteristics within the three TAs (Nkemdirim et al. 2016). Although this study did not look into the possible sources of groundwater fluoride in the study area, which in itself presents a gap for further investigation, other studies have shown that fluoride enrichment in groundwater is mainly a result of the interaction between water and fluoride-bearing minerals in the host rocks, such as fluorspar (fluorite), apatite (fluorapatite) and phosphorite (Prajapati et al. 2017). Geological processes such as weathering of the fluoride-bearing minerals and cation exchange, especially under high pH and low calcium conditions, are the common factors that influence fluoride concentration in ground water (Prajapati et al. 2017; Suneetha et al. 2018; Hao et al. 2021). Any variations in the existence of fluoride-bearing minerals, geological and geochemical processes would, therefore, result in variations in fluoride concentrations in the groundwater. However, a detailed analysis specific to the study area would be necessary.

The high prevalence of dental fluorosis among children in the study area signifies high-level exposure to health risks associated with high fluoride ingestions. The high temperature experienced in Mangochi which at times can go as high as 40 °C (Mangochi District Council 2017) lead to a high intake of fluoride-concentrated drinking water. Despite the health risk associated with exposure to high fluoride concentrations in drinking water, low-level awareness makes communities attribute traits of dental fluorosis that are heavily present in children to Zanha africana. The Zanha africana misconception was a general consensus among the parents because most of the children with pronounced dental fluorosis were in the age range of 6–13, which is the age group that likes adventuring in the bushes and enjoying wild fruits. This agrees with Akuno et al. (2019) who suggest that ‘Dental fluorosis in children also seems to be more prevalent and evident in permanent teeth than in primary teeth, further suggesting that in children over 10 years of age, dental fluorosis is more visible and prevalent than those below the age of eight’. Sajidu et al. (2007) also observed low level of awareness of the dangers of fluorosis in Machinga District in Malawi. The study by Etta (2020) in the Far North Region of Cameron found that 90.2% of the respondents did not know the cause of teeth discoloration (dental fluorosis). In a study in Kenya, 80% of the respondents attributed the staining of teeth to drinking water; however, 59% thought it was due to drinking salty water while only 21% identified fluoride as the causative agent (Gevera et al. 2022). This demonstrates that this study finding agrees with other past studies that a lack of community awareness on the linkage between dental fluorosis and highly fluoride-contaminated drinking water is a serious problem in different areas, both local and international.

The results of this study portray that there is a problem, both in terms of the fluoride concentrations as well as the prevalence of fluorosis in the study area, which could be a sign of people's exposure to fluoride-related health risks. However, the severity of the risks could better be established through a human health risk assessment (HHRA) following standard HHRA approaches like the one suggested by the United States Environmental Protection Agency (US-EPA), which considers hazard quotient (HQ) as a ratio of the estimated daily intake (EDI) to a reference dose (RfD) (Keshavarz et al. 2015; Karami et al. 2019; Dobrinas et al. 2022). By definition, HHRA is ‘the process to estimate the nature and probability of adverse health effects in humans who may be exposed to chemicals in contaminated environmental media, now or in the future’(US-EPA 2014). The HHRA in this case would, among others, assess the chances that people would experience different types of health problems when exposed to different levels of fluoride, and the people's susceptibility based on factors such as age, genetics, pre-existing health conditions, ethnic practices or gender.

The study concludes that fluoride contamination in groundwater is an issue in Mangochi district. High fluoride concentrations in excess of 10 mg/L exist in some ‘improved’ groundwater sources such as boreholes and protected shallow wells in the district, with TAs Nankumba and Katuli as the most and least vulnerable respectively, among the four TAs in the study. Dental fluorosis development is also a problem where there is high groundwater fluoride concentration. However, there is low awareness of health risks associated with fluoride contamination which was evident when community members attributed dental fluorosis to Ntalawanda (Zanha africana), a wild fruit in the area.

The study exposes a gap in terms of water quality monitoring and enforcement of standards by those in authority. There is, therefore, need for efforts to: (1) enhance water quality monitoring; (2) enforce water quality standards right from the construction stage and (3) enhance community awareness on water quality issues. It is also recommended that Government of Malawi and its development partners should seriously consider using solar powered groundwater reticulated water systems which have the ability to transport water from contamination-free sources to the areas with high ground fluoride concentrations, starting with the areas that this study has already exposed, in a bid to discontinue usage of the existing highly fluoride-rich water sources. In these reticulated systems, one borehole drilled in fluoride contamination zone can supply a number of villages through communal taps. There is also a great need to conduct a detailed HHRA due to the registered elevated fluoride concentrations in drinking water in the study area. In addition, further studies to investigate the sources of high fluoride concentrations on Mulopa Balamanja and Kwipulumbu boreholes and the possibility of the existence of hidden hot springs around the two boreholes need to be conducted. Lastly, it is recommended that similar studies should be conducted in other areas in order to have a broader understanding of groundwater fluoride distribution and the health risks involved in Mangochi district.

This study is part of PhD research at Mzuzu University and was funded by the Embassy of Iceland in Malawi. Mangochi district Council provided logistical support.

Free and informed consent of the participants or their legal representatives was obtained and the study protocol was approved by the appropriate Committee for the Protection of Human Participants [Malawi National Commission for Science and Technology – National Committee on Research in the Social Sciences and Humanities, Ref No: NCST/RTT/2/6, dated 5 May 2020].

K.A. conceptualized the whole article, developed the methodology and provided the software; K.A., M.M., M.T. validated the data; K.A. conducted a formal analysis; K.A. brought the resources; K.A. wrote the original draft; K.A., M.M., and M.T. reviewed and edited the article; M.M. and M.T. supervised the article; K.A. conducted funding acquisition. All authors have read and agreed to the published version of the manuscript.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.