Estimating short and longer-term exposure of domestic cats to dietary iodine fluctuation
This observational study was designed to first assess what cat owners routinely fed their cats to ascertain how variable the intake of dietary iodine may be. Second, we analysed the iodine content of a range of cat feeds. Finally, we measured iodine in cat urine and in cat hair as potential markers of variability in acute dietary iodine intake (cf. urine) and as a speculative marker of longer-term iodine exposure (cf. hair). Through a survey of cat owners, we found that the diet of cats is regularly changed on a daily-to-weekly basis, usually by the type and flavour of feed. More often than not, cats are fed either ad-lib or twice-a-day dry kibble, mixed with a pouch of wet food of varying flavour. Iodine content of cat feeds was broadly compliant with EU guidelines, but of the 22% that were non-compliant, the majority tended to be low in iodine and were dry kibble. Different batches of the same cat feeds had between 14 and 31% variability in iodine content. Domestic cats could therefore experience low and/or variable dietary iodine, which urinalysis of iodine confirmed; urine iodine varied by orders of magnitude. It was of interest that, reported for the first time, hair iodine content was significantly lower in cats being treated for hyperthyroidism, relative to controls. The implications of this finding warrant further investigation. The domestic cat is likely to experience variable excursions of dietary iodine that could facilitate pre-disposition to hyperthyroidism in combination with other goitrogens present in the modern, domestic environment. Such an aetiopathogenesis would mirror the development of toxic nodular goitre in humans.
What and how are cats fed
According to our survey, which was predominantly, but not exclusively, of UK-based cat owners, nearly all were fed commercial ‘branded’ pet feed; indeed, 95.4% consumed at least half of their dietary intake as commercial food, similar to that reported previously (98.8%40). As fed daily, the majority of commercial feed is given as a mix (approx. 50:50) of wet and dry food, as previously described41. If one type of food is ever given solely, then similar to the USA42, it is likely to be dry kibble (24% exclusively fed dry food). In our survey, in the 2020s, it was rare for cats to only receive wet food. Interestingly, similar surveys of cat owners conducted in the 1970s reported that only 39% of cats received any commercial dry food at all. Indeed, only 2% were fed dry food exclusively43 (cf. 24% in our survey). By the 1980/90s, dry food was becoming more popular with > 40% cats reportedly being fed only commercial wet/dry food40. These data clearly indicate changed feeding practices of companion animals, such as the cats in our survey. A greater proportion are now fed dry food exclusively, which is likely to facilitate modern working practices. Supporting these data, a large proportion of cats in our survey were reported as being predominantly indoor, but with access to the outdoor environment. A recent survey of cat owners in America42 found that 82% of cats lived almost exclusively indoors. This would suggest that cats in the 2020s have a greater reliance on commercial food, supplementing their diet with prey species caught outdoors infrequently.
The type of wet foods that are fed to cats, usually mixed with dry, are most often in the form of pouches. 40% of cats received tinned foods as part of their wet food intake. Decades ago when pouches were uncommon, 85% of cats were fed wet food from tins41,43. Around the millennium, 30% of cat owners reported feeding tinned food, with less than 10% doing so exclusively40. Taken together, these results again suggest altered feeding practices over the last few decades, with a gradual drift away from feeding tinned wet food to cats. Feeding tinned foods has been linked with increased risk of hyperthyroidism in cats6,8,9,11, which has risen sharply over a similar time-period. Thus, although tinned, and the linings of some dry food containers, may contain goitrogenic contaminants, such as BPA and PBDEs6,44, we would suggest this is unlikely a direct cause of the increased rates of hyperthyroidism in cats. In contrast, fish-flavoured foods have been reported to have high iodine content, and are regularly fed to cats32, likely due to a belief that cats prefer fish-flavoured foods45,46,47. Nevertheless, our survey data suggested that fish-flavoured foods comprised a much smaller proportion of the cat’s diet than meat flavours. If only one flavour of cat food was fed, it was uncommon for this to be fish, but rather white meat. Despite regular (i.e. daily) changes in dietary flavour, but not in brand—the latter was rarely changed—cat owners in the UK are brand loyal and only feed fish-flavoured foods uncommonly as part of a varied, weekly change in the diet flavour.
Pet foods have variable, but not often high, iodine content
Analysis of complete pet foods revealed wide variability in the iodine content of both cat and dog foods. To our knowledge, this is the first time that the iodine content of pet foods that are commercially available in the UK has been investigated, but the results are broadly congruous with similar studies conducted in the USA31, New Zealand30 and Germany26. Using recommended methods (ICP-MS with alkaline-digestion;48,49) we report that relatively few feeds (~ 20%) were outside of current EU guidelines34. Certainly, many more cat feeds were non-compliant in comparison to dog feeds, but many were close to the lower or upper nutritional guideline i.e. the cats were unlikely to be exposed to widely varying iodine content. Nevertheless, in humans, TNG occurs mostly in people with mild to moderate, not severe, iodine deficiency18. Therefore, if the pathogenesis of feline hyperthyroidism is akin to that of human TNG, the iodine-deficient foods that we have currently analysed may pose a risk for hyperthyroidism in cats if they were consumed over long periods of time. But, of course, currently fed foods as tested here may not necessarily reflect the iodine concentrations in foods fed during the long aetiopathogeneis of feline hyperthyroidism.
Greater variability in iodine content was found among wet, relative to dry, foods; all foods that exceeded the legal maximum were wet. Similarly, previous work had indicated that wet food tended to have greater variability in mineral content50,51. Nevertheless, in the current study, iodine content did not significantly vary between types of packaged wet foods (e.g. tins vs pouches, trays) and to a large extent the cats in our survey were usually fed a mix of wet and dry, or dry food only, which has a more consistent iodine content. Thus, with respect to cat foods, both iodine deficient and iodine excessive ‘complete’ diets are available. If such diets happen to be fed concurrently, it is theoretically possible for a cat to experience reasonable fluctuations in iodine intake, which, even when in the range of ‘normal’, could affect thyroid hormone levels in cats25. Therefore, our data does suggest that the routine feeding of cats with commercially available cat foods could inadvertently contribute toward feline hyperthyroidism through varied iodine intake. Determining causality, however, would require a suitably powered randomised, longer-term clinical trial, beyond the scope of the present study.
Wet pet feed batch-to-batch variation in iodine content
The current, and other, studies have ordinarily only determined iodine content in a range of pet foods. To our knowledge, only one study determined iodine content in different batches—two—of the same feeds30. In that study, the two batches of feed were broadly similar. Here, we determined iodine content in eight different complete, wet foods for adult cats across four-to-six different batches and found an average of 22% variation between batches, particularly in those foods with higher iodine content. An individual cat fed one of these higher iodine feeds over a long period of time, may experience reasonable variation in iodine intake. Enough to underpin a TNG-like hyperthyroidism? The current study cannot answer this question but does provide the evidence for perhaps closer and repeat monitoring of thyroid levels of cats fed specific diets. Toxic nodular goitre, the form of hyperthyroidism in humans that resembles feline hyperthyroidism, is induced by excessive iodine intake after a period of iodine deficiency16. Nevertheless, no wet food repeatedly analysed in the current study was found to have both iodine deficient and iodine excessive batches, only reasonable variation around high or low guideline amounts34.
Should we feed cats fish? Iodine intake
A previous study has reported higher iodine content of fish, relative to egg or dairy products32. Indeed, our pre-selected high iodine diets all had fish as the main flavour. Cat owners in our study reported feeding fish-flavoured feeds to many of the cats. However, in our hands, the iodine content of the currently analysed feeds was not significantly higher than feeds without fish as the main flavour. While some epidemiological studies have reported an association between consumption of fish-containing foods and risk of feline hyperthyroidism8,14, we suggest this is unlikely to be a cause-effect relationship, via iodine content. Continued analyses of further batches of feeds with fish as the main flavour would be required. Ultimately, whether any cat has consistently high or low iodine intake is better reflected through measurement of iodine excretion in urine28,29.
Urine iodine is variable in cats
In the current study, domestic cats did not appear to have chronically low or high iodine intake as reflected by analysis of their urine (cf. dogs, see Figure S1b). The kidneys have a major role in maintaining iodide homeostasis by excreting > 90% of dietary iodine intake, with iodine excretion being linearly proportional to intake28,29, also observed in cats26. Furthermore, the urinary iodine concentration of healthy cats has been shown to respond to dietary iodine changes, reducing significantly following dietary iodine restriction27 and increasing significantly when dietary iodine is increased25,52. Equally, cats fed a consistent level of iodine experience little change in urinary iodine concentration27. Thus, urinary iodine concentrations are a useful indicator of short-term (e.g. 1–2 days) dietary iodine intake52,53. In this study, feline urinary iodine was highly variable, likely reflecting, therefore, highly variable dietary intake. For all the pathological specimens we had access to, unfortunately we did not have dietary histories and thus a cause-effect relationship is beyond the scope of this study. Notwithstanding the cause of death and thus reason for necropsy in the current study, the similarity of urinary iodine between cats and dogs in the current study, does not suggest that cats have very different iodine intake to dogs, at least in the short-term. Marked variability in urinary iodine within individual cats has been previously reported54,55. Of interest, cats with a diagnosis of hyperthyroidism but sampled prior to treatment, were found to have lower urinary iodine excretion compared to euthyroid cats55. In the current study, urinary iodine concentrations were measured exclusively in outwardly healthy animals. However, as a potential biomarker of longer-term iodine exposure we did sample hair in known healthy-control (free from thyroid disease) or in cats currently receiving treatment for hyperthyroidism.
Hair iodine content is lower in cats being treated for hyperthyroidism
Since no study had previously measured iodine content of cat hair, we first validated our methods using human hair (with a human hair CRM). In our hands, human hair iodine was similar to previously published values i.e. 0.385 (0.209–0.535) μg/g iodine (median [interquartile range])39. No study has related hair iodine, as a longer-term marker of intake, with a diagnosis of hyperthyroidism. In cats, we report for the first time that cats being treated for hyperthyroidism had significantly lower hair iodine content than healthy cats without any known thyroid condition. Note that cats with a diagnosis of hyperthyroidism, but sampled prior to treatment, had significantly lower urinary iodine excretion compared to euthyroid cats55. Children with goitre have enlarged thyroid glands that correlates with higher hair iodine content56, which is believed to arise from chronic, excessive iodine intake57. However, both feline hyperthyroidism and TNG are conditions associated with an elderly population and so these data in children are of questionable relevance. Our novel data also require further validation: we were, for example, unable to determine whether hair iodine is lower in cats being treated for hyperthyroidism and thus reflect success of treatment or ‘biomarks’ a chronically low or variable-low dietary iodine status.
The study does have some limitations which should be acknowledged when interpreting the results. We were unable to retrospectively obtain diet histories for the cats submitted to pathology in our study, and thus the data for each of groups and outcomes are a cross-sectional sample, and should be treated with a degree of caution. Nevertheless, these first data do suggest that a longer-term study of euthyroid and hyperthyroid cats taking dietary, urine and hair samples is warranted to better understand the relationship between intake, uptake and bioaccumulation. Hair has been used previously as an easily obtained biomarker of longer-term mineral status, because many trace elements are permanently deposited in the hair matrix during formation58,59. Furthermore, other variable factors in our observational study that could be addressed in a prospectively-designed experimental study could include a possible effect of age on hair mineral analyses, as our control versus hyperthyroid cat groups were not age-matched. In addition, other comorbidities that could have an influence on iodine status but were not recorded on our database such as inflammatory bowel disease, could influence interpretation of results.
In summary, the present study has characterised estimated exposure of domestic cats to low and variable iodine intake, a dietary pattern associated with toxic nodular goitre in humans. Based on our survey data, feeding practices of domestic cats have changed considerably over the last few decades; most owners now feed commercial food, mainly dry kibble with varying types/flavours of additional wet pouches. Owners are brand loyal and tinned food is on the decline. We report good (78%) compliance of diets to European guidelines for iodine content. The 22% non-compliant feeds were mostly below nutritional minimum and were likely to be dry kibble. Feeds and urinalysis indicated a likely wide variation in dietary iodine intake in the domestic cat population. Significantly low iodine in the hair of cats being treated for hyperthyroidism may reflect success of treatment or biomark chronically low dietary intake. Labelling feeds with iodine content would empower owners to make informed decisions when selecting foods for their cats.