Presented in part at the 29th Annual Meeting of the American Society for Bone and Mineral Research, September 16–19, 2007, Honolulu, Hawaii, USA.
The authors state that they have no conflicts of interest.
Introduction: Skeletal fluorosis (SF) can result from prolonged consumption of well water with >4 ppm fluoride ion (F−; i.e., >4 mg/liter). Black and green teas can contain significant amounts of F−. In 2005, SF caused by drinking 1–2 gallons of double-strength instant tea daily throughout adult life was reported in a 52-yr-old woman.
Materials and Methods: A 49-yr-old woman developed widespread musculoskeletal pains, considered fibromyalgia, in her mid-30s. Additionally, she had unexplained, increasing, axial osteosclerosis. She reported drinking 2 gallons of instant tea each day since 12 yr of age. Fluoxetine had been taken intermittently for 5 yr. Ion-selective electrode methodology quantitated F− in her blood, urine, fingernail and toenail clippings, tap water, and beverage.
Results: Radiographs showed marked uniform osteosclerosis involving the axial skeleton without calcification of the paraspinal, intraspinal, sacrotuberous, or iliolumbar ligaments. Minimal bone excrescences affected ligamentous attachments in her forearms and tibias. DXA Z-scores were +10.3 in the lumbar spine and +2.8 in the total hip. Her serum F− level was 120 μg/liter (reference range, 20–80 μg/liter), and a 24-h urine collection contained 18 mg F−/g creatinine (reference value, <3). Fingernail and toenail clippings showed 3.50 and 5.58 mg F−/kg (control means, 1.61 and 2.02, respectively; ps < 0.001). The instant tea beverage, prepared as usual extra strength using tap water with ∼1.2 ppm F−, contained 5.8 ppm F−. Therefore, the tea powder contributed ∼35 mg of the 44 mg daily F− exposure from her beverage. Fluoxetine provided at most 3.3 mg of F− daily.
Conclusions: SF from habitual consumption of large volumes of extra strength instant tea calls for recognition and better understanding of a skeletal safety limit for this modern preparation of the world's most popular beverage.
Skeletal fluorosis (SF) is caused by chronic ingestion or inhalation of high amounts of fluoride ion (F−). This disorder features, however, a considerable range of severity and has been classified into four clinical phases. Preclinical SF is generally asymptomatic despite slight radiographic osteosclerosis. Clinical phase 1 SF is characterized by axial osteosclerosis and sometimes pain or stiffness in joints. Clinical phase 2 SF manifests with more extensive osteosclerosis, joint pain, and calcification of ligaments and may lead to limitation of mobility. Clinical phase 3 SF, often called “crippling fluorosis,” is associated with excrescences of bone at the attachments of tendons, ligaments, etc. near joints and vertebrae, skeletal deformities, and significantly reduced mobility, and may include debilitating muscle wasting and neurological deficits caused by spinal cord compression. The osteosclerosis of SF is most apparent in the axial skeleton. When advanced, it is accompanied by aching of the ribs and spine, impaired cervical and lumbar motion, kyphosis, and painful lower limbs. Long bone fractures may occur because of poor-quality, dense, fluorotic osseous tissue that can sometimes represent a form of osteomalacia.
Worldwide, SF affects millions and is most prevalent where high levels of F− contaminate well water (>4 ppm; i.e., >4 mg/liter), primarily regions of China, India, and Africa. SF also occurs when chronic F− exposure derives from certain industrial dusts or fumes (e.g., arc welding). Unprocessed foods contain too little F− to cause disease. Processed tea, however, can be F− rich, and many people in Asia suffer SF where poor-quality “brick” tea is comprised of the mature leaves, twigs, and berries of the tea plant, Camellia sinensis. Additional causes of SF are rare. In developed countries, SF has resulted from certain tainted wines or mineral waters, fluoridated toothpaste, and other unusual exposures.
Skeletal F− levels typically increase lifelong. Nevertheless, information in 2007 from a detailed case report of crippling SF showed that, after F− exposure stops, symptoms can improve within 1–2 yr, significant reduction in radiographic osteosclerosis may take a decade, and F− levels in bone can diminish appreciably over years.
Tea is the world's most popular beverage and is often said to promote health. However, in 2005, we reported phase 1 SF in a 52-yr-old American woman who had consumed 1–2 gallons of double-strength instant tea daily throughout her adult life. Also, we documented significant F− levels in several brand name instant teas. Subsequently, we found substantial F− concentrations in some ready-to-drink bottled teas.
Here, we describe SF in middle age from habitual consumption of large volumes of extra strength instant tea since childhood.
MATERIALS AND METHODS
A 49-yr-old woman from Illinois, USA, was referred for chronic, widespread, musculoskeletal aches and pains accompanying dense bones.
At age 31 yr, she underwent hysterectomy without oophorectomy and had not yet experienced symptoms of menopause. In her mid-30s, she became “tired and sore” and was eventually diagnosed with fibromyalgia by a rheumatologist who reported no inflammatory changes and a positive tender point exam in a fibromyalgic distribution. She was also found to have osteoarthritis (OA) that was particularly severe in her knees.
In 1997, at age 39 yr, DXA (Hologic, Waltham, MA, USA) documented elevated BMD in her lumbar (L1–L4) spine (Z-score, +4.3), whereas her left total hip BMD was normal (Z-score, +0.5). At age 48 yr, in 2006, BMD had increased 37% and 19% at these sites (Z-scores, +9.1 and +2.5, respectively). Radiographs also showed axial osteosclerosis that had progressed, and bone scintigraphy showed increased radioisotope uptake at several joints (see below).
Reportedly, she had sustained an atraumatic, nondisplaced, right distal tibial fracture at age 48 yr. There were no risk factors for hepatitis C infection. Daily oral medications included either 20 mg fluoxetine or 75 mg citalopram oxalate intermittently for 5 yr, 0.1 mg thyroxine, 25 mg hydrochlorothiazide, 1 mg ropinirole for restless legs syndrome, 10 mg cyclobenzaprine, 50 mg atenolol, and 50 mg diphenhydramine for insomnia, as well as 75 mg pregabalin twice a day and a combination of propoxyphene napsylate 100 mg/acetaminophen 650 mg three times a day for pain. Also, she took a supplement of 500 mg calcium with 200 units of vitamin D twice a day despite good dietary calcium intake.
Physical examination revealed a 5′ 4″-tall, 207-lb white woman who appeared suntanned in midwinter. Vital signs were unremarkable. Her tooth enamel was not mottled. There was no torus palatinus, and the mandible was not broad. Positive findings included tenderness to light fist percussion over her cervical spine and areas of her thoracic and lumbar spine. Soreness occurred from gentle squeezing of her hands and forearms and during compression of her arms, where discomfort in some places seemed to emanate more from soft tissue than from bone. Fist percussion over the femora caused no pain, but there were areas of tenderness during two-finger percussion of both pretibial regions. Audible crepitations occurred when she moved her knees. Ankles rotated silently without discomfort.
To explore SF as a potential explanation for her symptoms and elevated BMD, we learned that she had never worked in industry, did not consume alcohol-containing beverages or mineral water, or did not use Teflon-coated cookware. She had a vegetable garden and lived over a coal mine and near an abandoned zinc smelter but provided recent reports from the Environmental Protection Agency (EPA) and Department of Health for Illinois, which judged her topsoil (using X-ray fluorescence for metals) to be safe. She had not been exposed to fertilizers, insecticides, or chemical fumes, and she used fluoridated municipal tap water lifelong, which her provider told us was formulated to contain 1.0–1.1 ppm F− throughout the year. She brushed her teeth with fluoridated toothpaste twice daily and rinsed and spit out the dentifrice. She had not had dental F− treatments or used F−-containing mouthwash.
However, when asked if she drank tea, she promptly reported consuming 2 U.S. gallons (i.e., 7.57 liters total) of extra strength instant tea (Nestea; Nestlé, USA, Glendale, CA, USA) daily since age 12 yr. She purchased the unsweetened preparation and added aspartame (l-aspartyl-l-phenylalanine, Equal; Merisant, US, Chicago, IL, USA) beginning when this product was first marketed in 1982, gradually increasing to 24 1-g packets per gallon. She had switched to the decaffeinated preparation of this instant tea brand 10 yr earlier because of insomnia. Her husband mentioned that during the summertime she sometimes drank 3 gallons daily. He consumed none. Reportedly, her siblings also drank considerable volumes of instant tea each day during childhood, but chose other beverages by early adult life and did not have skeletal symptoms. Her mother had osteoporosis. Her two adult children were said to be healthy, including in infancy after her pregnancies during which she drank the instant tea without aspartame.
Skeletal radiographs available for review spanned 2003–2007 and showed a generalized increase in BMD (Figs. 1A–1F). The osteosclerosis was most apparent in the spine where the normal trabecular pattern was replaced with a uniform chalky density. OA spurring and cartilage loss affected multiple sites, most notably the knees.
The cranial vault was diffusely dense but not enlarged. The facial bones and mandible were radiographically normal (Fig. 1 A). The spine was shaped correctly (Figs. 1B–1D). Minimal degenerative disease, seen as disc narrowing and small marginal spurs, affected the midthoracic and upper lumbar spine but was typical for age (Figs. 1C and 1D). No ligamentous calcification was present. Osteosclerosis involved the glenoid and subglenoid portions of both scapulae.
The arms appeared normal. Calcifications were noted in the common flexor and extensor tendon origins at both elbows, likely related to chronic epicondylitis. OA changes were present in both hands, most severe at the first carpometacarpal joints bilaterally. Less severe joint narrowing and spurring was seen throughout the interphalangeal joints, indicating mild OA typical for her age.
The pelvis was diffusely osteosclerotic, especially in the posterior iliac wings (Fig. 1E), but somewhat less than in the spine. The corticomedullary junctions of the superior pubic rami were obscured by the osteosclerosis. Mild OA spurring affected both hips, but without significant cartilage loss. Minor degenerative changes were present at both sacroiliac joints, typical for age.
The femurs and tibias were shaped normally (Figs. 1E and 1F). Severe tricompartmental OA affected both knees, which contained joint bodies (data not shown). Some periosteal ossification involved the syndesmotic membrane attachment to both tibias.
Ligamentous calcification was found at both ankles in a pattern typically seen after ankle sprains. A small amount of periosteal bone excrescence was present proximally along the tibia and fibula bilaterally. Bilateral plantar calcaneal spurs were noted.
Overall, the major radiographic disturbance was markedly increased axial BMD of a nonspecific pattern, without bone deformity or expansion. Degenerative changes were slightly accelerated, especially in the upper extremities, except for severe OA in the knees.
MRI of the head and neck in 2006 showed a C5–C7 syrinx, yet otherwise unremarkable spinal cord. There was a Chiari I malformation but no other intracranial abnormality. Cervical degenerative disk disease was present, with mild disk bulging at C5–C6 and C6–C7. No ligamentous calcification was identified.
Whole body skeletal scintigraphy at age 48 yr showed no focal abnormality (Fig. 2). Normal radionuclide uptake was noted in the skull, spine, and pelvis. Increased tracer accumulation was present at sites of known OA.
DXA (Hologic Discovery C) BMD of the lumbar spine (L1–L4) was 2.104 g/cm2 (Z-score, +10.3; i.e., 216% age-matched controls). Left total hip BMD was 1.234 g/cm2 (Z-score, +2.8; i.e., 139% age-matched controls).
Before referral, her serum was studied at Mayo Medical Laboratories (Rochester, MN, USA): calcium, 9.1 mg/dl (reference range, 8.4–10.2 mg/dl); inorganic phosphate, 2.5 mg/dl (reference range, 2.5–4.5 mg/dl); creatinine (crt), 1.1 mg/dl (reference range, 0.7–1.2 mg/dl); intact PTH, 55 pg/ml (reference range, 15–72 pg/ml); alkaline phosphatase (ALP), 177 U/liter (reference range, 38–128 U/liter); bone-specific alkaline phosphatase, 39 μg/liter (reference range, 4.5–16.9 μg/liter for premenopausal women); osteocalcin, 47 ng/ml (reference range, 11–50 ng/ml); 25-hydroxyvitamin D, 42 ng/ml (reference range, 20–52 ng/ml); 1,25-dihydroxyvitamin D, 36 pg/ml (reference range, 15–75 pg/ml) (ARUP Laboratories, Salt Lake City, UT, USA). Electrolyte and thyroid-stimulating hormone levels, serum protein electrophoresis (Quest Laboratories), and hemogram with leukocyte differential counting were unremarkable. Anti-nuclear antibody and rheumatoid factor were negative, and the erythrocyte sedimentation rate was 24 mm/h (reference range, 0–20 mm/h).
F− was quantitated (Quest Diagnostics, Nichols Institute, Chantilly, VA, USA) in serum and in a 24-h urine specimen collected just before the patient stopped drinking instant tea. To determine the F− concentration in an aliquot of her instant tea beverage that she fortuitously brought to the clinic, as well as in her municipal tap water, we commissioned two independent testing facilities (St Louis Testing Laboratories, St Louis, MO, USA, and Kiesel Environmental Laboratories, St Louis, MO, USA), which both use F−-selective electrode (with known additions) methodology.
To assess for chronic F− exposure, F− levels in fingernail and toenail clippings (provided a few days before she stopped drinking instant tea and once again 3 mo afterward) were measured using an F−-selective electrode after hexamethyldisiloxane-facilitated diffusion extraction. Control specimens, provided for the initial F− determinations, consisted of concurrent nail clippings from three 46- to 48-yr-old premenopausal women who, like the patient, did not use nail polish. The specimens were anonymized and analyzed in duplicate, triplicate, or quadruplicate depending on the amount of tissue available.
F− levels at the time of diagnosis are summarized in Table 1. In serum, the F− concentration of 120 μg/liter (reference range, 20–80 μg/liter) was distinctly elevated (reference range, 15–50 μg/liter for adults with 1 ppm F− in their drinking water). The 24-h urine specimen contained 4.26 mg F−/liter (reference range, 0.20–3.20 mg F−/liter). Considering the substantial urine volume of 4.9 liters, excretion of F− was clearly increased at 20.9 mg/d or 17.7 mg F−/g crt.
Table Table 1.. Fluoride Levels at Diagnosis
The patient stated that she prepared her instant tea beverage using 0.75 to 1 cup of the tea powder added to 1 gallon of tap water (manufacturer's label recommendation is 0.67 cup/gallon). Analysis in duplicate at both testing laboratories showed an identical mean F− concentration of 5.8 ppm in the instant tea beverage brought to clinic, with her tap water contributing ∼1.2 ppm F−. Therefore, the 2 gallons of instant tea consumed each day provided 44 mg F− (5.8 mg F−/liter × 7.57 liter). The F− from the instant tea powder (i.e., tap water F− was excluded) was ∼35 mg daily.
The F− levels (mean ± SE) in the fingernail clippings from the patient (just before stopping instant tea) and from the control subjects were 3.50 ± 0.13 and 1.61 ± 0.18 mg/kg, respectively (p < 0.001). In the toenail clippings, the F− levels were 5.58 ± 0.12 and 2.02 ± 0.23 mg/kg, respectively (p < 0.001).
Soon after SF was suspected, the patient stopped drinking instant tea and switched without difficulty to ∼2 qt (1 qt = 0.946 liter) of F−-free bottled water daily. She also ceased using fluoridated toothpaste and fluoxetine, which contains 3.3 mg F− per tablet.
Three months after stopping instant tea, etc., F− levels in her fingernail and toenail clippings were 36% and 26% lower, but still elevated at 2.24 and 4.12 mg/kg, respectively. After 4 mo, the serum F− concentration was <20 μg/liter (reference range, 20–80 μg/liter), and a 24-h urine specimen of 2.2 liters contained a normal F− level of 1.21 mg/liter (2.7 mg F−/d or 2.3 mg F−/g crt).
Assessment of her urine calcium level (4 mo after F− exposure stopped, and while she was not taking her calcium supplement) showed just 6 mg calcium/g crt (verified by Alverno Clinical Laboratories, Hammond, IN, USA). Although it seemed that this could reflect avid calcium uptake by her skeleton caused by a persisting anabolic effect of the F− or healing of SF osteomalacia (see below), she was taking hydrochlorothiazide. DXA 8 mo after F− exposure stopped showed unchanged BMD in the lumbar spine and total hip (2.120 and 1.241 g/cm2, respectively), and therefore no evidence that BMD was still increasing. She stated that skeletal symptoms were unchanged; axial discomfort was being dealt with by increases in her medications. On physical examination, she seemed less tender in several areas of her skeleton.
Our patient developed SF from habitual consumption of large volumes of extra strength instant tea.
F− metabolism and skeletal effects
Fluoride (F−), the ionic form of the trace element fluorine, is the 13th most abundant element in the earth's crust (∼0.07%). Adults typically consume <0.5 mg of F− daily in food, whereas water fluoridation (∼1.0 mg/liter) adds ∼1–3 mg/d. In the absence of high dietary levels of certain divalent cations, especially calcium, >80% of ingested F− is rapidly absorbed (t½ ∼ 30 min) from the intestinal tract. On entering plasma, F− quickly establishes a steady-state distribution between the extracellular and intracellular fluids of soft tissues. Because of its negative charge and large hydration radius, F− does not readily cross cell membranes. Instead, it distributes across these barriers according to the magnitude of the pH gradient and the diffusion equilibrium of the highly diffusible and permeating molecule, HF (pKa = 3.4). Tissue-to-plasma concentration ratios span ∼0.1 in fat to 0.8 in liver and do not exceed 1.0 unless ectopic calcification occurs as in atherosclerotic plaques, near-term placenta, or the pineal gland. In fact, ∼50% of the F− absorbed from the gut deposits in the skeleton where it is firmly, but not irreversibly, bound; most of the remainder is excreted in the urine.
In fingernails and toenails, F− enters at their growth ends in concentrations proportional to, but much higher than, those in plasma. Hence, F− levels in nail clippings reflect average circulating F− concentrations prevailing 3–4 mo previously. The mechanism for F− accumulation in this keratin is not known but probably involves diffusion from plasma.
The skeleton contains ∼99% of the body burden of F−. Because F− has the same charge and virtually the same size as OH−, F− substitutes for OH− in hydroxyapatite to form hydroxyfluorapatite crystals and lesser quantities of fully substituted fluoroapatite. These crystals are more compact, less soluble, and more stable than hydroxyapatite, and resist skeletal resorption. Furthermore, F− uncouples bone remodeling by promoting osseous tissue formation, leading to a net accumulation of skeletal mass. Osteoblasts proliferate because F− enhances mitogenic signals or growth factor effects or because it inhibits phosphotyrosine phosphatase increasing tyrosyl phosphorylation, prompting recruitment of precursor cells to become osteoblasts. In fact, sodium fluoride (NaF) given for osteoporosis increases lumbar spine BMD linearly, even after 6 yr of treatment.
Unfortunately, chronic F− toxicity can produce poor-quality, brittle bones, and osteomalacia is sometimes seen on histopathologic examination of SF. Painful lower limbs may result from microfractures, and our patient reportedly sustained an atraumatic tibial fracture 1 yr before referral.
Skeletal fluorosis and our patient
Initially, our patient had high serum levels of both total and bone-specific ALP, as well as high-normal levels of serum osteocalcin, consistent with enhanced osteoblast activity and her increasing BMD over the previous decade. Radiographs showed spinal and pelvic osteosclerosis typical of SF, but there was a relative lack of the commonly seen bone excrescences at ligament and muscle attachments, periosteal thickening, or calcification of the sacrotuberous or intraspinal ligaments (sometimes with ossification of the anterior longitudinal ligament compressing the spinal cord) expected in severe SF. Only minimal calcification was noted at her distal tibial attachments of the syndesmotic membranes. Hence, it was important to at least consider hepatitis C–associated osteosclerosis and the high bone mass phenotype (LRP5 activation) as possible explanations for her increasingly dense bones (Table 2). Instead, her radiographic findings, despite chronic and very high levels of F− intake, seem most like phase 1, rather than phase 2 or 3, SF. However, DXA showed that our patient's axial BMD was actually markedly elevated. BMD was considerably higher in her spine and hip compared with values after 4 yr of NaF therapy for postmenopausal osteoporosis. Based on the F− levels in our patient's fasting serum and 24-h urine collection before stopping tea, her renal clearance of F− was ∼121 ml/min and probably equaled her glomerular filtration rate. For people with normal kidney function, which she apparently had, F− clearance typically ranges 30–50 ml/min. Perhaps her high F− clearance reflected a serum F− level that was relatively low before drinking tea that day. Clearly, there was no evidence that renal compromise impaired her ability to excrete F−.
Table Table 2.. Causes of Increasing High Bone Mass in Adults
Although the 500 mg of calcium supplementation taken twice a day by our patient may have benefited her by (1) diminishing F− absorption from the gut (after complexing with this anion) or (2) mineralizing osteoid, calcium did not abolish her gut absorption of F− because F− levels in her serum, urine, and fingernail/toenail clippings were elevated. Deficiency of dietary calcium or vitamin D has been regarded as an exacerbating factor for SF, and calcium and vitamin D supplementation for SF perhaps mineralizes or prevents excessive osteoid formation. Conversely, however, calcium supplementation could diminish bone resorption by suppressing PTH secretion and therefore preserve the fluorotic skeleton or exacerbate hypercalciuria caused by negative bone balance that can follow cessation of F− exposure (see below). Our patient developed SF with seemingly adequate dietary calcium and levels of vitamin D.
In 2007, we reported a case of phase 3 SF, ostensibly from habitual ingestion of fluoridated toothpaste, and documented that substantial symptomatic improvement from advanced SF is possible. Decreased symptoms can occur within 1–2 yr after stopping F− exposure, but significant reduction in osteosclerosis can take decades, and complete correction of dense bones might never be achieved for severe SF diagnosed in middle age. In fact, the t½ of F− in the adult skeleton is ∼7 yr and probably increases in advanced age. For osteoporosis, cessation of NaF therapy leads in most patients to loss of the acquired bone mass at approximately the same rate it increased during treatment. Calcium or estrogen supplementation has been reported not to preserve the gains. Recently, however, Cundy described a woman whose high BMD from NaF therapy for postmenopausal osteoporosis persisted until she stopped hormone replacement. We anticipate that our patient will lose fluorotic bone mass soon as she enters menopause, and we will monitor her for transition from hypocalciuria to hypercalciuria.
Fluoride levels in instant tea
In 2005, we reported that the beverage made from some commercial instant tea powders available in the United States and prepared according to the manufacturers' recommendations (but using distilled water) exceeded the Public Health Service (PHS) and even Food and Drug Administration (FDA) limits for F− in fluoridated water and bottle beverages, respectively, and one mixture surpassed the EPA primary standard for F− in drinking water (4 ppm). Subsequently, in 2006, we found that F− levels in some ready-to-drink, bottled teas in the United States exceeded the PHS and FDA limits. We recognized for both studies that our data reflected only a single look in time at one sample of each tea preparation, knowing that F− levels in teas can vary according to geographic origin, time of year when harvested, and other factors. We therefore concluded that further study was indicated.
Instant tea and skeletal fluorosis
Appreciation of the significant amounts of F− in some modern preparations of tea suggested to us that SF caused our patient's high bone mass when she recounted her remarkable volume of instant tea consumed over three decades. In fact, her beverage, made extra strength in tap water with ∼1.2 ppm F−, contained 5.8 ppm F− on the day we examined her. In the United States, this F− concentration exceeds the EPA's primary standard (enforceable) of 4.0 ppm F− for drinking water, the FDA's limits spanning 1.4–2.4 ppm F− for bottled water or beverages, and the PHS's optimum levels ranging from 0.7 to 1.2 ppm F− for community water fluoridation. The World Health Organization's guideline limit for F− in well water is 1.5 ppm. To minimize the risk of dental fluorosis, the PHS recommends community water fluoridation with optimum F− levels that depend on average air temperature. Similarly, the FDA requirement for F− in bottled beverages or water packaged in the United States depends on the annual average maximum daily air temperature where the product is sold.
Although the exceptional volume of instant tea that our patient consumed daily was the primary factor in her SF, an extra strength mix using fluoridated water was also contributory. Her tap water accounted for ∼9 mg F− ingested each day. Intermittently, over the past few years, at most, another 3.3 mg F− daily may have come from fluoxetine, depending on how much F− is released during the drug's metabolism. Nevertheless, the instant tea powder was clearly the principal source of her F− exposure, accounting for ∼35 mg F−/d.
Intake of at least 10 mg of F− daily for 10 yr seems necessary for “preclinical SF.” Hence, this exposure is considered the “no-observed-adverse-effect level” for adults. The “lowest effect level” for stage 1 or stage 2 SF is 20 mg/d with continuous exposure for at least 20 yr. Our patient's beverage contained 5.8 ppm F−. Accordingly, drinking 2 qt/d (1.9 liters) for 10 yr would exceed the no-observed-adverse-effect level for F− (5.8 ppm F− × 1.9 liters = 11.0 mg/d). Drinking 1 gallon (4 qt) daily (21.9 mg F−/d) for 20 yr would exceed the lowest effect level for phase 1 or phase 2 SF.
Staging our patient's skeletal fluorosis
Despite our patient's prolonged intake of high amounts of F−, significant symptoms, and marked axial osteosclerosis, her age-appropriate number and size of exostoses, lack of periostitis, etc. are atypical for phase 3 SF reported worldwide. In fact, she and the other American patient with SF from instant tea lacked these radiographic features of crippling SF. Clinicians and radiologists must now be aware of this presentation for SF from instant tea. Our first patient, whose F− intake ranged from ∼37 to 74 mg/d throughout her adult life, would have phase 1 SF, but our current patient is difficult to classify because fibromyalgia shares symptoms with SF. Difficulties that improve over the next 1–2 yr might be recognized in retrospect to be from SF.
There are relatively few reports of SF, at any clinical phase, in the United States. However, DXA screening for osteoporosis and routine evaluation of BMD after fractures will now surely unmask additional hyperdense skeletons here and in many other countries worldwide. SF in our patient was readily detected using DXA, which now represents an additional and widely available quantitative means (beyond assay of bone F− levels) to document the severity of SF. In our patient with phase 3 SF from toothpaste and our patient with phase 1 SF from instant tea, BMD Z-scores were markedly elevated at +14.3 and +9.9 in their L2–L4 and L1–L4 spines, respectively (+6.6 and +1.3 at the femoral neck and total hip, respectively). Our current patient's BMD Z-scores at diagnosis of SF were +10.3 in her L1–L4 spine and +2.8 in her total hip, fitting with intermediate severity compared with the other two patients. Although direct quantitation of F− levels in bone represents the “gold standard” for identifying SF, SF is typically diagnosed from an increased F− level in a 24-h urine collection. We confirm here that measuring F− levels in fingernail and toenail clippings is useful for revealing the level of chronic exposure to F−.
Other cases of SF from instant tea
Recent telephone contact with the first reported patient with SF from instant tea, now age 61 yr, showed that 8 yr after abruptly stopping this beverage, low back pain has returned. She attributes this to more physical work and cessation of physical therapy. Symptoms of rheumatoid arthritis in her hands are controlled with several medications. For approximately the past year, she consumed 2 cups of brewed (bag) tea daily. She also continued esterified estrogens plus methyltestosterone, and ethinyl estradiol and norethindrone for the past 5 yr. DXA (Norland Medical Systems, Fort Atkinson, WI, USA) recently documented essentially no decrease in BMD during this time (L2–L4 and femoral neck Z-scores of +9.1 and +1.8, respectively). Perhaps hormone therapy is preserving her fluorotic bone mass.
From our understanding of F− levels that can be found in commercial instant teas and experience with these patients, we surmise that habitual consumption of 3 qt daily of some regular-strength preparations for >10 yr, especially if made with fluoridated water, could cause clinically significant SF. For example, if the beverage contained just 3 ppm F−, the threshold for preclinical SF would be exceeded if 4 qt were consumed per day for more than a decade. This F− exposure seems possible for many individuals who like instant or bottled teas. In fact, when a 36-yr-old coworker learned of our index case, she confided drinking 3–4 qt daily over the past year of what she described as a triple-strength preparation of Nestea dissolved in unfiltered, municipal tap water. She felt well, serum ALP activity was 63 U/liter (reference range, 20–125 U/liter), and DXA (Hologic) BMD was unremarkable (L1–L4 spine Z-score = +0.6, total hip Z-score = +0.5), but a 24-h urine specimen of 2.9 liters contained 8 mg of F−/g crt (reference value <3 mg of F−/g crt). Furthermore, while we were preparing our manuscript, four postmenopausal women were reported with axial osteosclerosis and elevated levels of serum F− from chronic consumption of excessive amounts of various preparations of tea. Unlike our two patients, however, they all had renal compromise expected to impair urinary excretion of F−.
F− and bone health
In adults, daily consumption of 3 mg F− for women and 4 mg F− for men is considered adequate for reducing tooth decay. F− in brewed tea seems to help if F− levels are low in drinking water, because this beverage contains 1–6 ppm F−, depending partly on the water source and brewing time. There is also epidemiological evidence that BMD is enhanced in habitual tea drinkers. In 2001, investigation of hip and overall fracture rates among rural Chinese indicated a skeletal benefit from at least some F− in well water. However, ≥4.32 ppm F− was associated with increased risk. Indeed, in the United States, water fluoridation remains controversial, and new F− sources (e.g., pediatric supplements, fluoridated toothpastes and mouthwashes, and dental treatments) together with diminishing per capita consumption of tap water may complicate future recommendations.
The catechins (antioxidant flavonoids), phytoestrogens, and other components found in tea are said to benefit health. We worry that clinically important SF may result from drinking large volumes of some instant or bottled teas. Although our two cases may seem exceptional, prolonged consumption of lesser amounts of regular-strength preparations could cause SF, especially using water containing F−. Postmenopausal women with osteoporosis treated for 4 yr using NaF increased BMD in their lumbar spine and femoral neck by 35% and 12%, respectively, but had nonspine fracture rates that seemed to increase. Our two patients are from Missouri and Illinois and not from southern states in the United States. Although data from the U.S. Department of Agriculture and the National Health and Nutrition Examination Survey indicate little difference in fluid intake throughout the United States, we are concerned that SF from instant tea may prove more prevalent in hot climates, particularly where F−-containing water is used.
With increasing use of DXA, additional instances of SF from instant tea will likely be seen (see below) because of high axial BMD (i.e., Z-scores > +2.5). For individuals with high axial BMD, when tea is consumed, both the volume and strength lifelong should be considered, with assay of F− in a 24-h urine collection (corrected for creatinine) if SF is suspected. Even if F− exposure abruptly ceases, opportunity for diagnosis is not lost because urine F− levels can remain high for 1 yr in severe cases and for at least several months in more mild cases. Measuring F− in nail clippings remains an interesting investigational approach for diagnosing SF. F− concentrations in nascent fingernail and toenail substance, like those in urine, remain proportional to levels in plasma.
Treatment of skeletal fluorosis
There is no established treatment for SF. Special distillation or reverse osmosis water filters, low-F− bottled waters, avoidance of tea prepared from Camellia sinensis, and especially careful use of F−-containing dental products would seem prudent. Bone antiresorptive pharmaceuticals, such as bisphosphonates, and estrogen replacement therapy for postmenopausal women would likely prolong F− retention in the skeleton. It is not known whether treatment with PTH (e.g., teriparatide: recombinant amino acid fragment 1–34) would safely increase bone turnover in SF, leading to an unloading of skeletal F− and reduction in bone mass without risk of kidney stones or whether it would exacerbate the osteosclerosis of SF. Risk and severity of SF seem inversely related to calcium intake, and results from a long-term study of rats showed than an increase in dietary calcium caused a negative F− balance from enhanced fecal excretion of F−. Assessment and, if necessary, adjustment of dietary calcium intake or use of calcium supplements should be considered for SF, but hypercalciuria must be screened for, especially if F− exposure stops.
Our encounter with a second patient with SF from habitual consumption of large volumes of extra strength instant tea calls for better understanding of the threshold and systemic effects of F− contained in this modern preparation of the world's most popular beverage.
“Be not too zealous; moderation is best in all things.”
Theognis (570–490 BC)
Ruth Craddock, MD, helped to diagnose and treat the patient. Joseph E Zerwekh, PhD, provided advice concerning the F− studies. Angelia English and Cindy Webster at Shriners Hospital for Children, St Louis, MO, USA, contributed expert secretarial help. This work was supported by Shriners Hospitals for Children, The Clark and Mildred Cox Inherited Metabolic Bone Disease Research Fund, and The Barnes-Jewish Hospital Foundation.