- Email: [email protected]
High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study
Clinical Nutrition xxx (2017) 1e6
Contents lists available at ScienceDirect
Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu
Original article
High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study lie Affret a, b, c, Courtney Dow a, b, c, Francesca Romana Mancini a, b, c, Aure Beverley Balkau a, d, Françoise Clavel-Chapelon a, b, c, Fabrice Bonnet a, b, e, Marie-Christine Boutron-Ruault a, b, c, Guy Fagherazzi a, b, c, * a
INSERM U1018, Center for Research in Epidemiology and Population Health, Villejuif, France University Paris-Saclay, University Paris-Sud, Villejuif, France Gustave Roussy, F-94805, Villejuif, France d University Versailles, Saint Quentin, University Paris-Sud, Villejuif, France e CHU Rennes, Universit e de Rennes 1, France b c
a r t i c l e i n f o
s u m m a r y
Article history: Received 10 May 2017 Accepted 25 July 2017
Phosphorus is an essential nutrient; the adult recommended daily intake ranges from 550 to 700 mg/day, with a tolerated upper limit of 4000 mg/day. Phosphorus intake has increased in the general population in recent years, and simultaneously an alarming rise of type 2 diabetes incidences has been observed. No study has investigated the relationship between phosphorus intake and the risk of type 2 diabetes. To evaluate the association between phosphorus intake and incidence of type 2 diabetes. Among 71,270 women from the French E3N-EPIC cohort, 1845 cases of incident type 2 diabetes were validated during follow-up (1993e2011). Adjusted Cox proportional hazards regression models were used to calculate hazard ratios and 95% confidence intervals (95% CI) for the associations between phosphorus intake and type 2 diabetes risk, adjusted on potential confounders. The overall mean (±SD) phosphorus intake was 1477 mg/day (±391 mg/day). High phosphorus intake was associated with risk of type 2 diabetes. In multivariate models, compared with women in the 1st quartile of phosphorus intake (<1203 mg/day), those included in the 2nd (1203e1434.0 mg/day), 3rd (1434e1700 mg/day), and 4th (>1700 mg/day) were at a higher risk of type 2 diabetes, with a hazards ratios (95% CI) of 1.18 (1.00e1.38), 1.41 (1.20e1.66) and 1.54 (1.25e1.90), respectively. Our results may have important public health implications for dietary recommendations in the prevention of type 2 diabetes. More studies are warranted to better understand the biological mechanisms underlying this positive association. © 2017 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
Keywords: Type 2 diabetes Phosphorus Dietary intake Risk E3N cohort
1. Introduction Phosphorus is an essential mineral and plays a crucial role in several biological processes. Indeed phosphorus is structural component of cell membranes and nucleic acids; moreover, phosphorus is also implicated in several biological processes, such as bone mineralization, energy production, cell signaling through phosphorylation reactions, and regulation of acid-base
* Corresponding author. INSERM U1018, Centre for Research in Epidemiology and Population Health (CESP), “Health across Generations” Team, Gustave Roussy, Espace Maurice Tubiana, 114 rue Edouard Vaillant, 94805 Villejuif Cedex, France. E-mail address: [email protected] (G. Fagherazzi).
homeostasis [1]. In adults, 55e80% of phosphorus intake is absorbed in the intestine and the rest is excreted in the feces [2]. The European Food Safety Authority established an Adequate Intake (AI) for adults of 550 mg/day, while the Food and Nutrition Board of the Institute of Medicine of the American National Academy of Sciences set a Recommended Dietary Allowance (RDA) of 700 mg/ d and a tolerable upper limit of 4000 mg/d for adults [3,4]. In European Union countries, the mean phosphorus intake in adults ranges from 1000 to 1767 mg/day. Foods rich in protein content, such as milk and dairy products followed by meat, poultry and fish, grain products, and legumes are the principal dietary contributors to phosphorus intake [3]. Several food additives commonly used in food processing have Phosphorus as main component, moreover soft drinks can contribute to the phosphorus intake due to the use
http://dx.doi.org/10.1016/j.clnu.2017.07.025 0261-5614/© 2017 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
Please cite this article in press as: Mancini FR, et al., High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study, Clinical Nutrition (2017), http://dx.doi.org/10.1016/j.clnu.2017.07.025
2
F.R. Mancini et al. / Clinical Nutrition xxx (2017) 1e6
of phosphoric acid. Food nutrient composition databases do not always include phosphorus derived from food, thus it has been estimated that the total amount of phosphorus intake may be underestimated by more than 20% [5]. Even though it has been suggested that the population dietary intake of phosphorus has increased due to the growing consumption of foods processed with phosphate additives, reports on the long-term health risks associated with high dietary phosphorus intake are scarce and contradictory. Previous studies reported that the intake of phosphorus is inversely associated with hypertension and with bladder cancer incidence [6e8]. Long-term high dietary phosphorus intake, even without leading to hyperphosphatemia, has been described as a risk factor for cardiovascular disease, and osteoporosis and bone fracture [9e11]. Finally, Chang et al. (2014) [12] reported that high phosphorus intake was associated with increased mortality in a healthy US population. The alarming rise of type 2 diabetes (T2D) incidence that has been observed worldwide in the last decades [13], in parallel with the increasing consumption of processed foods which is also associated to an increase of dietary intake of phosphorus [14], led us to investigate the association between dietary intake of phosphorus and T2D risk in the E3N-EPIC study, a large French cohort.
All potential cases were contacted and asked to answer a specific questionnaire. The questionnaire included questions on the circumstances of diagnosis (year of diagnosis, symptoms, biological exams, fasting, or random glucose concentrations at diagnosis), actual diabetes therapy (prescription of diet or physical activity, list of diabetic drugs taken), last measurement of fasting glucose and HbA1c levels. Through this specific questionnaire, diabetes cases were ascertained with fasting glucose at diagnosis 1.26 g/l or random glucose at diagnosis 2.00 g/l and/or current fasting glucose 1.26 g/l and/or current HbA1c 7% and/or reported diabetic drugs. Potential diabetes cases were validated by cross-checking at least two sources: self-reported diabetes in the follow-up questionnaires and/or ascertained diabetes in the specific diabetes questionnaire and/or drugs reimbursed by the insurance. For cases identified through the file listing drugs reimbursed by the insurance, potential cases were classified as non-cases when they declared to be non diabetic and were reimbursed for diabetic drugs only once during the follow-up; potential cases were classified as validated diabetic cases when they did not answer the specific diabetes questionnaire and were reimbursed at least twice. 2.4. Assessment of dietary phosphorus intake
2. Materials and methods 2.1. E3N cohort The E3N cohort is a prospective study involving 98,995 women born in 1925e1950, living in metropolitan France at inclusion and insured by the Mutuelle G en erale de l’Education Nationale, a national teachers' health insurance plan. Details of the cohort have been previously reported [15]. All women enrolled returned a baseline self-administered questionnaire on their lifestyle and medical history from February 1st, 1989, through November 30th, 1991, and, since then, answered follow-up questionnaires every 2e3 years. The average follow-up per each questionnaire cycle has been 83%, and to date, the overall loss to follow-up since 1990 has been 3%. The French National Commission for Computerized Data and Ins, dividual Freedom (Commission Nationale Informatique et Liberte CNIL) provided the needed ethical approval for the study, and an informed consent was signed by all women participating to the study. 2.2. Population for analysis Women whom did not complete the food frequency questionnaire sent in 1993 (n ¼ 24,473) were excluded from the present study since the dietary questionnaire was needed to estimate dietary exposure to phosphorus. Moreover women with extreme values for the ratio between energy intake and required energy (i.e. the 1st and 99th percentiles of the distribution in the population) (n ¼ 1491), with pre-existing diabetes (n ¼ 859), or women who did not complete any questionnaire after the dietary questionnaire (n ¼ 887) were also excluded from the study. Finally, the analysis included 71,270 women, of whom 1 845, meaning 2,6% of the study population, had a validated incident T2D diagnosed during followup (June 1993eDecember 2011). 2.3. Ascertainment of type 2 diabetes Potential diabetes cases were identified first in the 9 follow-up questionnaires through self-reporting of diabetes, diabetes diet, diabetes drugs, hospitalization for diabetes and then through the file listing drugs reimbursed by the insurance since 2004.
In June 1993, a validated 208-item food frequency questionnaire was sent to participants to assess the habitual diet of the previous year. Food and drink consumption throughout the day took French meal patterns into account, with questions for each meal occasion, snacks, and aperitifs. The validity and reproducibility of the questionnaire have been previously described [16]. The conversion of foods into macro- and micro-nutrients used a food composition table that was compiled ad hoc for this study and derived from the official French food composition table [17], completed when necessary by the McCance and Widdowson's food composition table and other published sources [18]. Individual average daily dietary intakes of phosphorus were estimated. 2.5. Statistical analysis Missing values were <5% for all variables and were imputed with the median of the study population (quantitative variables) or the mode (qualitative variables). Due to the high correlation between energy intake and micronutrient intakes, phosphorus intake was energy-adjusted using the residual method to allow analyses independent of total energy intake [19,20]. Hazard ratios (HR) and 95% confidence intervals (CI) of T2D risk were estimated by Cox multivariable regression models with age as the timescale. The time at entry was the age at the start of followup and the exit time was the age when participants were diagnosed with diabetes, died, were lost to follow-up or were censored at the end of the follow-up period, whichever occurred first. Women were categorized by quartiles according to the dietary intake of phosphorus and the lowest quartile was used as the reference category in the models. We first performed a univariate analysis (Model 0), followed by Model 1, that was adjusted for daily energy intake (excluding energy from alcohol consumption, kilocalories/day), alcohol consumption (grams of ethanol/day), level of education (6.2 mmol/l or use of cholesterol lowering drugs: yes vs. no), family history of diabetes (yes vs. no), and physical activity (metabolic equivalents (MET)hours/week).
Please cite this article in press as: Mancini FR, et al., High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study, Clinical Nutrition (2017), http://dx.doi.org/10.1016/j.clnu.2017.07.025
F.R. Mancini et al. / Clinical Nutrition xxx (2017) 1e6
Model 2 was additionally adjusted for dietary intakes of calcium (mg/day), magnesium (mg/day), and vitamin D (mg/day), as well as for the use of calcium supplements (yes vs. no), use of vitamin D supplements (yes vs. no), and the total acid load of the diet, evaluated with the potential renal acid load (PRAL) score [21]. Finally, Models 1 and 2 were further adjusted for body mass index (BMI) (<25, 25e30 or >30 kg/m2) leading to Models 3 and 4, respectively. Tests for linear trends were conducted by assigning the median value to each quintile group and modeling this value as a continuous variable. Dairy products and meat are reported to be the main dietary sources of phosphorous, and animal fat and processed meat have been found associated with T2D risk [22,23]. We tested therefore interactions between dietary phosphorus intake and consumptions of high-fat dairy products and processed meat (categorized as high consumers vs low consumers, using the median as cutoff value); we additionally tested interactions between dietary phosphorus intake and adherence to a Western dietary pattern or to a Healthy dietary pattern (high vs low adherence) [24], and between dietary phosphorus intake and BMI. Cubic spline regression was used to better characterize the continuous shape of the association between dietary phosphorus intake and T2D risk. The reference value for estimating the HRs and 95% CIs was chosen as the first percentile of the residual distribution, and three knots were used (respectively the 25th, 50th, and 75th percentiles of the residual distribution of phosphorus intake). All statistical tests were two-sided and considered as statistically significant if P < 0.05. Data were analyzed with the Statistical Analysis Systems software package, version 9.4 (SAS Institute, Cary, NC). 2.6. Sensitivity analyses Model 4 was additionally adjusted for the consumptions of high-fat dairy products (g/day), processed meat (g/day), and Western and Healthy dietary patterns in three separate analyses. We also estimated the associations between dietary phosphorus intake and T2D after having excluded type 2 diabetes cases occurred during the first 5 years of follow-up in order to rollout the reverse causation bias. Finally, because the residual method and the nutrient density method (nutrient intake divided by total energy intake) rely on different hypotheses regarding the relationship between the
3
studied nutrient and overall energy intake, thus potentially providing different results, all analyses were repeated using the nutrient density method. 3. Results 3.1. Phosphorus intake and baseline characteristics The overall mean (±SD) dietary phosphorus intake was 1477 mg (±391). We describe the participants' characteristics at baseline according to the absolute phosphorus intake quartiles in Table 1. Women with higher phosphorus intake were younger than those with low dietary phosphorus intake. They were also characterized by a higher BMI and more physical activity, more frequently reporting a family history of diabetes, with a higher education level; fewer had hypercholesterolemia and they were less frequently smokers. Moreover, women with higher phosphorus intake also had higher daily intakes of energy, alcohol, calcium, magnesium, and vitamin D, and higher PRAL scores. No difference among quartiles was found concerning hypertension and consumption of calcium and vitamin D supplements. The food group that most contributed to the total phosphorus intake was “High-fat dairy products”, 21.5% of the total intake (Table 2). “Cereal products”, “Fresh meat”, and “Low-fat dairy products” contributed overall to 36.5% of the total dietary phosphorus intake. 3.2. Phosphorus intake and type 2 diabetes risk Higher phosphorus intakes were associated with a higher risk of diabetes, even after adjustment for all potential confounders (Table 3). In particular, Model 4 highlighted that, compared with those in the lowest quartile of phosphorus intake (<1203.8 mg/ day), women in the higher quartiles were at higher T2D risk, with hazards ratios (HR (95% CI)) of 1.18 (1.00e1.38), 1.41 (1.20e1.66) and 1.54 (1.25e1.90) in the 2nd (1203.8e1434.0 mg/day), 3rd (1434.1e1700.6 mg/day), and 4th (>1700.6 mg/day) quartiles, respectively. In all models there was a statistically significant linear trend for increasing T2D risk across quartiles of phosphorous intake (Ptrend < 0.0001). There was no interaction between dietary phosphorus intake and BMI (P ¼ 0.49), between dietary phosphorus intake and consumptions of high-fat dairy products (P ¼ 0.72), or of processed meat (P ¼ 0.37), or between dietary phosphorus intake and the
Table 1 Characteristics of the women studied by quartiles of the dietary phosphorus intake (E3N-EPIC cohort data, n ¼ 71 270). Variable
Quartile 1 (n ¼ 17 817)
Quartile 2 (n ¼ 17 818)
Quartile 3 (n ¼ 17 818)
Quartile 4 (n ¼ 17 817)
P valuea
Dietary phosphorus intake (mg/day) Individual characteristics Age (years) BMI (kg/m2) Family history of T2D (%) Hypercholesterolemia (%) Hypertension (%) Physical activity (MET-hours/week) Smoker (%) Secondary education (%) Dietary characteristics Energy intake (kcal/day) Alcohol (g/day) Potential renal acid load (PRAL) Dietary calcium intake (mg/day) Dietary magnesium intake (mg/day) Dietary vitamin D intake (mg/day) Use of calcium supplements (%) Use of vitamin D supplements (%)
1028 (135.4)
1321 (66.3)
1559 (75.8)
2001 (274.8)
<0.0001
53 (6.9) 22.4 (3.0) 8.3 38.2 13.4 44.8 (53.5) 14.4 83.9
53 (6.7) 22.7 (3.1) 9.1 39.1 13.4 45.8 (48.1) 13.4 85.3
52.6 (6.5) 22.9 (3.2) 9.7 38.7 13.4 46.0 (47.6) 12.6 86.2
52.2 (6.4) 23.5 (3.5) 9.5 35.5 13.3 49.3 (51.6) 13.2 84.9
<0.0001 <0.0001 <0.0001 <0.0001 0.974 <0.0001 <0.0001 <0.0001
1651 (292) 9.5 (12.3) 8.5 (13.8) 716 (164) 351 (110) 1.9 (0.9) 6.7 2.9
2033 (296) 11.0 (13.1) 5.6 (15.5) 933 (166) 409 (111) 2.3 (0.9) 6.4 3.1
2329 (333) 11.9 (13.8) 3.1 (17.1) 1120 (189) 456 (118) 2.7 (1.1) 6.1 2.9
2833 (474) 13.8 (15.7) 0.3 (22.5) 1496 (348) 547 (146) 3.4 (1.4) 6.2 3.2
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.0884 0.2076
a
Chi-square tests for qualitative variables and Kruskal-Wallis tests for quantitative variables.
Please cite this article in press as: Mancini FR, et al., High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study, Clinical Nutrition (2017), http://dx.doi.org/10.1016/j.clnu.2017.07.025
4
F.R. Mancini et al. / Clinical Nutrition xxx (2017) 1e6 Table 2 Percentage of contribution of the main food groups to the total dietary phosphorus intake (E3N-EPIC cohort data, n ¼ 71 270). Food groups
% of contribution
High fat dairy products Cereal products Fresh meat Low fat dairy products Vegetables Fish Fruit and juices Hot beverages (coffee, tea…) Eggs Processed meat Composite dishes Desserts Others
21.5 14.6 12.3 9.6 8.7 5.1 3.7 3.4 3.2 3.2 2.9 2.7 12.3
Western or Healthy dietary patterns (P ¼ 0.29 and P ¼ 0.18, respectively). The cubic spline regression analysis emphasized a linear relationship between phosphorus intake and risk of T2D from (Fig. 1). 3.3. Sensitivity analyses After adjusting Model 4 for consumption of high-fat dairy products or processed meat, or for the Western or Healthy dietary patterns, the magnitudes of the associations between phosphorus and T2D risk did not materially change (Supplementary Table 1). After having excluded 564 cases that occurred during the first five years of follow-up, we performed the analysis including 1281 cases in order to test a reverse causation hypothesis between phosphorus intake and T2D risk. The results obtained were equivalent to those of the main analyses in terms of magnitude and statistical significance (data not shown). For example, compared to women in the 1st quartile, women in the 4th quartile of phosphorus intake had a HR of 1.46 (95%CI 1.14e1.86, Ptrend < 0.001), as in the main analyses. Finally, no relevant difference was found when comparing the results obtained by the nutrient density method with those obtained with the residual method (Supplementary Table 2). 4. Discussion In this cohort of middle-aged women, a higher intake of phosphorus was associated with increased T2D risk. To the best of our knowledge, the present study represents the first prospective study to investigate the risk of T2D associated with dietary intake of
phosphorus. The relationship between increasing dietary phosphorus intake and T2D appeared to be linear and the increased risk was observed already from the 2nd quartile of phosphorus intake. The association between high phosphorus intake and increased T2D risk remained statistically significant after adjusting for most established or potential confounders and throughout all sensitivity analyses, increasing our confidence in these results. The study population's phosphorus intake (1477 mg/day) is greatly above the adequate intake proposed by EFSA (550 mg/day). These excessive intake levels are mainly related to the high consumption of dairy products, which alone justify over 30% of the total phosphorus intake. 4.1. Biological pathways Although phosphorus is an essential nutrient, it has been suggested that a high dietary intake could induce tissue damage through a variety of mechanisms involved in the endocrine regulation of extracellular phosphate. Indeed, in healthy individuals, phosphorus is absorbed in the small intestine, and under the regulatory action of the parathyroid hormone (PTH) and the fibroblast growth factor-23 (FGF-23), the excesses are excreted with the urines [25]. Current evidence suggests that, over the long term, high concentrations of PTH and FGF-23 may have deleterious effects, and that high dietary phosphorus intake could have an impact on health even in the absence of increased phosphorus levels in the serum [26,27]. In particular, high circulating levels of FGF-23 have recently been associated with insulin resistance, using the HOMAIR index, fat mass, and dyslipidemia [28]. Thus, it could be hypothesized that the relationship between dietary phosphorus intake and T2D might be mediated by the action of FGF-23. Considering the dramatic increase of T2D prevalence worldwide and the burden it represents for health-care [13], it is crucial to better understand all possible dietary-related risk factors and to provide more specific and up-to-date dietary recommendations. Furthermore our findings may have considerable implications in terms of public health due to the widespread use of inorganic phosphorus as food additives. Indeed the amount of phosphorus to which the general population is exposed through the diet has increased in the past decades, due to the growing consumption of foods processed with phosphate additives [29]. Moreover, studies are warranted to investigate the influence of phosphorus intake on diabetes progression and onset of complications, especially considering that phosphorus dietary intake has been reported as an independent risk factor for kidney and cardiovascular disease [29], which are also among the main complications associated to diabetes.
Table 3 Hazard ratios for the risk of incident type 2 diabetes according to quartiles of dietary phosphorus intake (mg/day) estimated by the residual method (E3N-EPIC cohort data, n ¼ 71 270). Dietary phosphorus intake (mg/day)
Quartile 1 (<1203.8 mg/day)
Quartile 2 (1203.8e1434.0 mg/day)
Quartile 3 (1434.1e1700.6 mg/day)
Quartile 4 (>1700.6 mg/day)
P trend
Model Model Model Model Model
REF REF REF REF REF
1.20 1.26 1.26 1.16 1.18
1.58 1.60 1.63 1.36 1.41
1.84 1.80 1.90 1.43 1.54
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001
0 1 2 3 4
(1.04e1.39) (1.08e1.46) (1.08e1.48) (1.00e1.35) (1.00e1.38)
(1.38e1.83) (1.39e1.85) (1.38e1.92) (1.18e1.57) (1.20e1.66)
(1.60e2.10) (1.57e2.07) (1.54e2.34) (1.24e1.65) (1.25e1.90)
Model 0: univariate analysis. Model 1: adjusted for daily energy intake (excluding kilocalories from alcohol consumption, kilocalories/day), alcohol consumption (grams of ethanol/day), level of education (30 kg/m2). Model 4: Model 2 further adjusted for body mass index (BMI) (<25, 25e30 or >30 kg/m2).
Please cite this article in press as: Mancini FR, et al., High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study, Clinical Nutrition (2017), http://dx.doi.org/10.1016/j.clnu.2017.07.025
F.R. Mancini et al. / Clinical Nutrition xxx (2017) 1e6
5
Fig. 1. Cubic spline regression model between phosphorus intake and risk of type 2 diabetes (E3N-EPIC cohort data, n ¼ 71 270). Spline regression (3 knots, reference value: Phosphorus intake ¼ 1038.2 mg/day). Continuous line: hazard ratio. Dashed lines: 95% confidence intervals. The model was adjusted for daily energy intake (excluding kilocalories from alcohol consumption, kilocalories/day), alcohol consumption (grams of ethanol/day), level of education (30 kg/m2).
4.2. Strengths and limitations Potential limitations of the present study should be considered. The dietary estimates are based on a single questionnaire at baseline, thus misclassification of exposure is possible. However, as the study is prospective, any effects are likely to be non-differential and would lead to an attenuation of the true association. The fact that our study population includes only can be considered as a minor limitation due to the fact that only few studies have reported differences in T2D risk factors between sexes, although biological mechanisms may differ between men and women. Finally, dietary phosphorus intake did not include inorganic phosphorus driven from food additives; consequently the study population's phosphorus intake is probably underestimated; again this leads to nondifferential misclassification of exposure, therefore the true association between dietary phosphorus intake and T2D could be stronger than what eventually has been found by the present study. The two main strengths of the present study are the large sample size and the prospective design of the E3N cohort, allowing performing sensitivity analyses, maintaining sufficient statistical power to detect associations and make the reverse causation hypothesis unlikely. Moreover, we analyzed validated T2D cases only, based on a well-defined validation algorithm, which reduces the risk of false negatives or false positives. Moreover, extensive information on potential confounders was collected, minimizing residual confounding, and the dietary questionnaire used to assess phosphorus intake was validated and specific to this French population. 4.3. Conclusions In our study, dietary phosphorus intake was associated with type 2 diabetes risk with a linear dose-effect relationship. Since T2D
represents a high disease burden worldwide, and dietary phosphorus intake is increasing with increasing consumption of highlyprocessed industrial foods among the general population, our results may have important public health implications. More studies are warranted to confirm our findings and better understand the biological mechanisms that undergo this positive association. Contribution statement FRM and GF conceived and designed the study. FCC collected the data. FRM, MCB and GF performed the statistical analysis. FRM drafted the original manuscript. All authors contributed to the interpretation of data discussed in the manuscript, revised the manuscript and approved its final version. Conflict of interest The authors declare that they have no conflict of interest associated with this manuscript. Acknowledgments The authors are grateful to all participants for providing the data used in the E3N Study. This present work was supported by grant from the World Cancer Research Fund (WCRF) (Grant number: 2015/1389), the French Research Agency (ANR, Agence Nationale de la Recherche) via an “Investissement d’Avenir” grant (investment for the future grant, ANR-10-COHO-0006) that supports the E4N study, and the Nutriperso IDEX Paris Saclay University funding. The validation of potential diabetes cases was supported by the European Union (Integrated Project LSHM-CT-2006-037197 in the 6th European Community Framework Programme) InterAct project.
Please cite this article in press as: Mancini FR, et al., High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study, Clinical Nutrition (2017), http://dx.doi.org/10.1016/j.clnu.2017.07.025
6
F.R. Mancini et al. / Clinical Nutrition xxx (2017) 1e6
Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.clnu.2017.07.025. References [1] Knochel JP. Phosphorus. In: Shils ME, Shike M, Ross AC, Caballero B, Cousins RJ, editors. Modern nutrition in health and disease. 10th ed. Baltimore: Lippincott Williams & Wilkins; 2006. [2] Walker HK, Hall WD, Hurst JW. Chapter 198 Serum inorganic phosphorus. In: Bansal Vinod K, editor. Clinical methods: the history, physical, and laboratory examinations. 3rd ed. Boston: Butterworths; 1990. [3] European Food Safety Authority (EFSA). Scientific opinion on dietary reference values for phosphorus. Panel on dietetic products, nutrition and allergies. EFSA J 2015;13(7):4185. [4] Institute of Medicine (US) Standing Committee on the Scientific Evaluation of Dietary Reference Intakes. Dietary reference intakes for calcium, phosphorus, magnesium, vitamin D, and fluoride. Washington (DC): National Academies Press (US); 1997. [5] Calvo MS, Uribarri J. Public health impact of dietary phosphorus excess on bone and cardiovascular health in the general population. Am J Clin Nutr 2013;98(1):6e15. [6] Michaud DS, Spiegelman D, Clinton SK, Rimm EB, Willett WC, Giovannucci E. Prospective study of dietary supplements, macronutrients, micronutrients, and risk of bladder cancer in US men. Am J Epidemiol 2000 15;152(12):1145e53. [7] Alonso A, Nettleton JA, Ix JH, de Boer IH, Folsom AR, Bidulescu A, et al. Dietary phosphorus, blood pressure and incidence of hypertension in the atherosclerosis risk in communities (ARIC) Study and the multi-ethnic study of atherosclerosis (MESA). Hypertension 2010;55(3):776e84. [8] Elliott P, Kesteloot H, Appel LJ, Dyer AR, Ueshima H, Chan Q, et al., INTERMAP Cooperative Research Group. Dietary phosphorus and blood pressure: international study of macro- and micro-nutrients and blood pressure. Hypertension 2008;51(3):669e75. [9] Takeda E, Yamamoto H, Yamanaka-Okumura H, Taketani Y. Dietary phosphorus in bone health and quality of life. Nutr Rev 2012;70(6):311e21. [10] Calvo MS, Park YK. Changing phosphorus content of the U.S. diet: potential for adverse effects on bone. J Nutr 1996;(Suppl. 126):S1168e11180. [11] Calvo MS. Dietary phosphorus, calcium metabolism and bone. J Nutr 1993;123:1627e33. rrez OM, Grams ME. High dietary phos[12] Chang AR, Lazo M, Appel LJ, Gutie phorus intake is associated with all-cause mortality: results from NHANES III. Am J Clin Nutr 2014;99(2):320e7. [13] World Health Organization (WHO). Global report on diabetes. WHO library cataloguing-in-publication data. Geneva: World Health Organization; 2016. [14] European Food Safety Authority (EFSA). Opinion of the scientific panel on dietetic products, nutrition and allergies on a request from the commission
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22] [23]
[24]
[25] [26]
[27]
[28]
[29]
related to the tolerable upper intake level of phosphorus. EFSA J 2005;233: 1e19. Clavel-Chapelon F, van Liere MJ, Giubout C, Niravong MY, Goulard H, Le Corre C, et al. E3N, a French cohort study on cancer risk factors. E3N group. Etude epidemiologique aupres de femmes de l'Education Nationale. Eur J Cancer Prev 1997;6:473e8. van Liere MJ, Lucas F, Clavel F, Slimani N, Villeminot S. Relative validity and reproducibility of a French dietary history questionnaire. Int J Epidemiol 1997;1(Suppl. 26):S128e36. pertoire ge ne ral des aliFavier J-C, Ireland-Ripert J, Toque C, Feinberg M. Re ments: table de composition (Food Composition tables). IQUAL-REGAL. 1995, (French). INRA; AFSSA; CIQUAL; TEC & DO; INRA 927 pp. ISBN 2-85206-921-0. Paul AA, Southgate DAT. McCance & Widdowson's the food composition of foods. 4th ed. Amsterdam, New York, Oxford: Elsevier; 1976. North-Holland, Biomedia Press; MRC Special Report No. 297. Willett WC, Howe GR, Kushi LH. Adjustment for total energy intake in epidemiologic studies. Am J Clin Nutr 1997;65(Suppl. 4):1220Se8S. discussion 1229S-1231S. Brown CC, Kipnis V, Freedman LS, Hartman AM, Schatzkin A, Wacholder S. Energy adjustment methods for nutritional epidemiology: the effect of categorization. Am J Epidemiol 1994 Feb 1;139(3):323e38. Remer T, Manz F. Estimation of the renal net acid excretion by adults consuming diets containing variable amounts of protein. Am J Clin Nutr 1994;59:1356e61. rus U, Willett WC, Hu FB. Dietary fats and prevention of type 2 diabetes. Rise Prog Lipid Res 2009;48(1):44e51. Micha R, Michas G, Mozaffarian D. Unprocessed red and processed meats and risk of coronary artery disease and type 2 diabetesean updated review of the evidence. Curr Atheroscler Rep 2012;14(6):515e24. Cottet V, Touvier M, Fournier A, Touillaud MS, Lafay L, Clavel-Chapelon F, et al. Postmenopausal breast cancer risk and dietary patterns in the E3N-EPIC prospective cohort study. Am J Epidemiol 2009;170:1257e67. Martin A, David V, Quarles LD. Regulation and function of the FGF23/klotho endocrine pathways. Physiol Rev 2012;92(1):131e55. Hjelmesaeth J, Hofsø D, Aasheim ET, Jenssen T, Moan J, Hager H, et al. Parathyroid hormone, but not vitamin D, is associated with the metabolic syndrome in morbidly obese women and men: a cross-sectional study. Cardiovasc Diabetol 2009; 3;8:7. Reis JP, von Mühlen D, Kritz-Silverstein D, Wingard DL, Barrett-Connor E. Vitamin D, parathyroid hormone levels, and the prevalence of metabolic syndrome in community-dwelling older adults. Diabetes Care 2007;30(6): 1549e55. € J, Hammarstedt A, Erben RG, Michae €lsson K, Tivesten A, et al. Mirza MA, Alsio Circulating fibroblast growth factor-23 is associated with fat mass and dyslipidemia in two independent cohorts of elderly individuals. Arterioscler Thromb Vasc Biol 2011;31(1):219e27. Uribarri J, Calvo MS. Hidden sources of phosphorus in the typical American diet: does it matter in nephrology? Semin Dial 2003;16(3):186e8.
Please cite this article in press as: Mancini FR, et al., High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study, Clinical Nutrition (2017), http://dx.doi.org/10.1016/j.clnu.2017.07.025
Contents lists available at ScienceDirect
Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu
Original article
High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study lie Affret a, b, c, Courtney Dow a, b, c, Francesca Romana Mancini a, b, c, Aure Beverley Balkau a, d, Françoise Clavel-Chapelon a, b, c, Fabrice Bonnet a, b, e, Marie-Christine Boutron-Ruault a, b, c, Guy Fagherazzi a, b, c, * a
INSERM U1018, Center for Research in Epidemiology and Population Health, Villejuif, France University Paris-Saclay, University Paris-Sud, Villejuif, France Gustave Roussy, F-94805, Villejuif, France d University Versailles, Saint Quentin, University Paris-Sud, Villejuif, France e CHU Rennes, Universit e de Rennes 1, France b c
a r t i c l e i n f o
s u m m a r y
Article history: Received 10 May 2017 Accepted 25 July 2017
Phosphorus is an essential nutrient; the adult recommended daily intake ranges from 550 to 700 mg/day, with a tolerated upper limit of 4000 mg/day. Phosphorus intake has increased in the general population in recent years, and simultaneously an alarming rise of type 2 diabetes incidences has been observed. No study has investigated the relationship between phosphorus intake and the risk of type 2 diabetes. To evaluate the association between phosphorus intake and incidence of type 2 diabetes. Among 71,270 women from the French E3N-EPIC cohort, 1845 cases of incident type 2 diabetes were validated during follow-up (1993e2011). Adjusted Cox proportional hazards regression models were used to calculate hazard ratios and 95% confidence intervals (95% CI) for the associations between phosphorus intake and type 2 diabetes risk, adjusted on potential confounders. The overall mean (±SD) phosphorus intake was 1477 mg/day (±391 mg/day). High phosphorus intake was associated with risk of type 2 diabetes. In multivariate models, compared with women in the 1st quartile of phosphorus intake (<1203 mg/day), those included in the 2nd (1203e1434.0 mg/day), 3rd (1434e1700 mg/day), and 4th (>1700 mg/day) were at a higher risk of type 2 diabetes, with a hazards ratios (95% CI) of 1.18 (1.00e1.38), 1.41 (1.20e1.66) and 1.54 (1.25e1.90), respectively. Our results may have important public health implications for dietary recommendations in the prevention of type 2 diabetes. More studies are warranted to better understand the biological mechanisms underlying this positive association. © 2017 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
Keywords: Type 2 diabetes Phosphorus Dietary intake Risk E3N cohort
1. Introduction Phosphorus is an essential mineral and plays a crucial role in several biological processes. Indeed phosphorus is structural component of cell membranes and nucleic acids; moreover, phosphorus is also implicated in several biological processes, such as bone mineralization, energy production, cell signaling through phosphorylation reactions, and regulation of acid-base
* Corresponding author. INSERM U1018, Centre for Research in Epidemiology and Population Health (CESP), “Health across Generations” Team, Gustave Roussy, Espace Maurice Tubiana, 114 rue Edouard Vaillant, 94805 Villejuif Cedex, France. E-mail address: [email protected] (G. Fagherazzi).
homeostasis [1]. In adults, 55e80% of phosphorus intake is absorbed in the intestine and the rest is excreted in the feces [2]. The European Food Safety Authority established an Adequate Intake (AI) for adults of 550 mg/day, while the Food and Nutrition Board of the Institute of Medicine of the American National Academy of Sciences set a Recommended Dietary Allowance (RDA) of 700 mg/ d and a tolerable upper limit of 4000 mg/d for adults [3,4]. In European Union countries, the mean phosphorus intake in adults ranges from 1000 to 1767 mg/day. Foods rich in protein content, such as milk and dairy products followed by meat, poultry and fish, grain products, and legumes are the principal dietary contributors to phosphorus intake [3]. Several food additives commonly used in food processing have Phosphorus as main component, moreover soft drinks can contribute to the phosphorus intake due to the use
http://dx.doi.org/10.1016/j.clnu.2017.07.025 0261-5614/© 2017 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
Please cite this article in press as: Mancini FR, et al., High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study, Clinical Nutrition (2017), http://dx.doi.org/10.1016/j.clnu.2017.07.025
2
F.R. Mancini et al. / Clinical Nutrition xxx (2017) 1e6
of phosphoric acid. Food nutrient composition databases do not always include phosphorus derived from food, thus it has been estimated that the total amount of phosphorus intake may be underestimated by more than 20% [5]. Even though it has been suggested that the population dietary intake of phosphorus has increased due to the growing consumption of foods processed with phosphate additives, reports on the long-term health risks associated with high dietary phosphorus intake are scarce and contradictory. Previous studies reported that the intake of phosphorus is inversely associated with hypertension and with bladder cancer incidence [6e8]. Long-term high dietary phosphorus intake, even without leading to hyperphosphatemia, has been described as a risk factor for cardiovascular disease, and osteoporosis and bone fracture [9e11]. Finally, Chang et al. (2014) [12] reported that high phosphorus intake was associated with increased mortality in a healthy US population. The alarming rise of type 2 diabetes (T2D) incidence that has been observed worldwide in the last decades [13], in parallel with the increasing consumption of processed foods which is also associated to an increase of dietary intake of phosphorus [14], led us to investigate the association between dietary intake of phosphorus and T2D risk in the E3N-EPIC study, a large French cohort.
All potential cases were contacted and asked to answer a specific questionnaire. The questionnaire included questions on the circumstances of diagnosis (year of diagnosis, symptoms, biological exams, fasting, or random glucose concentrations at diagnosis), actual diabetes therapy (prescription of diet or physical activity, list of diabetic drugs taken), last measurement of fasting glucose and HbA1c levels. Through this specific questionnaire, diabetes cases were ascertained with fasting glucose at diagnosis 1.26 g/l or random glucose at diagnosis 2.00 g/l and/or current fasting glucose 1.26 g/l and/or current HbA1c 7% and/or reported diabetic drugs. Potential diabetes cases were validated by cross-checking at least two sources: self-reported diabetes in the follow-up questionnaires and/or ascertained diabetes in the specific diabetes questionnaire and/or drugs reimbursed by the insurance. For cases identified through the file listing drugs reimbursed by the insurance, potential cases were classified as non-cases when they declared to be non diabetic and were reimbursed for diabetic drugs only once during the follow-up; potential cases were classified as validated diabetic cases when they did not answer the specific diabetes questionnaire and were reimbursed at least twice. 2.4. Assessment of dietary phosphorus intake
2. Materials and methods 2.1. E3N cohort The E3N cohort is a prospective study involving 98,995 women born in 1925e1950, living in metropolitan France at inclusion and insured by the Mutuelle G en erale de l’Education Nationale, a national teachers' health insurance plan. Details of the cohort have been previously reported [15]. All women enrolled returned a baseline self-administered questionnaire on their lifestyle and medical history from February 1st, 1989, through November 30th, 1991, and, since then, answered follow-up questionnaires every 2e3 years. The average follow-up per each questionnaire cycle has been 83%, and to date, the overall loss to follow-up since 1990 has been 3%. The French National Commission for Computerized Data and Ins, dividual Freedom (Commission Nationale Informatique et Liberte CNIL) provided the needed ethical approval for the study, and an informed consent was signed by all women participating to the study. 2.2. Population for analysis Women whom did not complete the food frequency questionnaire sent in 1993 (n ¼ 24,473) were excluded from the present study since the dietary questionnaire was needed to estimate dietary exposure to phosphorus. Moreover women with extreme values for the ratio between energy intake and required energy (i.e. the 1st and 99th percentiles of the distribution in the population) (n ¼ 1491), with pre-existing diabetes (n ¼ 859), or women who did not complete any questionnaire after the dietary questionnaire (n ¼ 887) were also excluded from the study. Finally, the analysis included 71,270 women, of whom 1 845, meaning 2,6% of the study population, had a validated incident T2D diagnosed during followup (June 1993eDecember 2011). 2.3. Ascertainment of type 2 diabetes Potential diabetes cases were identified first in the 9 follow-up questionnaires through self-reporting of diabetes, diabetes diet, diabetes drugs, hospitalization for diabetes and then through the file listing drugs reimbursed by the insurance since 2004.
In June 1993, a validated 208-item food frequency questionnaire was sent to participants to assess the habitual diet of the previous year. Food and drink consumption throughout the day took French meal patterns into account, with questions for each meal occasion, snacks, and aperitifs. The validity and reproducibility of the questionnaire have been previously described [16]. The conversion of foods into macro- and micro-nutrients used a food composition table that was compiled ad hoc for this study and derived from the official French food composition table [17], completed when necessary by the McCance and Widdowson's food composition table and other published sources [18]. Individual average daily dietary intakes of phosphorus were estimated. 2.5. Statistical analysis Missing values were <5% for all variables and were imputed with the median of the study population (quantitative variables) or the mode (qualitative variables). Due to the high correlation between energy intake and micronutrient intakes, phosphorus intake was energy-adjusted using the residual method to allow analyses independent of total energy intake [19,20]. Hazard ratios (HR) and 95% confidence intervals (CI) of T2D risk were estimated by Cox multivariable regression models with age as the timescale. The time at entry was the age at the start of followup and the exit time was the age when participants were diagnosed with diabetes, died, were lost to follow-up or were censored at the end of the follow-up period, whichever occurred first. Women were categorized by quartiles according to the dietary intake of phosphorus and the lowest quartile was used as the reference category in the models. We first performed a univariate analysis (Model 0), followed by Model 1, that was adjusted for daily energy intake (excluding energy from alcohol consumption, kilocalories/day), alcohol consumption (grams of ethanol/day), level of education (
Please cite this article in press as: Mancini FR, et al., High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study, Clinical Nutrition (2017), http://dx.doi.org/10.1016/j.clnu.2017.07.025
F.R. Mancini et al. / Clinical Nutrition xxx (2017) 1e6
Model 2 was additionally adjusted for dietary intakes of calcium (mg/day), magnesium (mg/day), and vitamin D (mg/day), as well as for the use of calcium supplements (yes vs. no), use of vitamin D supplements (yes vs. no), and the total acid load of the diet, evaluated with the potential renal acid load (PRAL) score [21]. Finally, Models 1 and 2 were further adjusted for body mass index (BMI) (<25, 25e30 or >30 kg/m2) leading to Models 3 and 4, respectively. Tests for linear trends were conducted by assigning the median value to each quintile group and modeling this value as a continuous variable. Dairy products and meat are reported to be the main dietary sources of phosphorous, and animal fat and processed meat have been found associated with T2D risk [22,23]. We tested therefore interactions between dietary phosphorus intake and consumptions of high-fat dairy products and processed meat (categorized as high consumers vs low consumers, using the median as cutoff value); we additionally tested interactions between dietary phosphorus intake and adherence to a Western dietary pattern or to a Healthy dietary pattern (high vs low adherence) [24], and between dietary phosphorus intake and BMI. Cubic spline regression was used to better characterize the continuous shape of the association between dietary phosphorus intake and T2D risk. The reference value for estimating the HRs and 95% CIs was chosen as the first percentile of the residual distribution, and three knots were used (respectively the 25th, 50th, and 75th percentiles of the residual distribution of phosphorus intake). All statistical tests were two-sided and considered as statistically significant if P < 0.05. Data were analyzed with the Statistical Analysis Systems software package, version 9.4 (SAS Institute, Cary, NC). 2.6. Sensitivity analyses Model 4 was additionally adjusted for the consumptions of high-fat dairy products (g/day), processed meat (g/day), and Western and Healthy dietary patterns in three separate analyses. We also estimated the associations between dietary phosphorus intake and T2D after having excluded type 2 diabetes cases occurred during the first 5 years of follow-up in order to rollout the reverse causation bias. Finally, because the residual method and the nutrient density method (nutrient intake divided by total energy intake) rely on different hypotheses regarding the relationship between the
3
studied nutrient and overall energy intake, thus potentially providing different results, all analyses were repeated using the nutrient density method. 3. Results 3.1. Phosphorus intake and baseline characteristics The overall mean (±SD) dietary phosphorus intake was 1477 mg (±391). We describe the participants' characteristics at baseline according to the absolute phosphorus intake quartiles in Table 1. Women with higher phosphorus intake were younger than those with low dietary phosphorus intake. They were also characterized by a higher BMI and more physical activity, more frequently reporting a family history of diabetes, with a higher education level; fewer had hypercholesterolemia and they were less frequently smokers. Moreover, women with higher phosphorus intake also had higher daily intakes of energy, alcohol, calcium, magnesium, and vitamin D, and higher PRAL scores. No difference among quartiles was found concerning hypertension and consumption of calcium and vitamin D supplements. The food group that most contributed to the total phosphorus intake was “High-fat dairy products”, 21.5% of the total intake (Table 2). “Cereal products”, “Fresh meat”, and “Low-fat dairy products” contributed overall to 36.5% of the total dietary phosphorus intake. 3.2. Phosphorus intake and type 2 diabetes risk Higher phosphorus intakes were associated with a higher risk of diabetes, even after adjustment for all potential confounders (Table 3). In particular, Model 4 highlighted that, compared with those in the lowest quartile of phosphorus intake (<1203.8 mg/ day), women in the higher quartiles were at higher T2D risk, with hazards ratios (HR (95% CI)) of 1.18 (1.00e1.38), 1.41 (1.20e1.66) and 1.54 (1.25e1.90) in the 2nd (1203.8e1434.0 mg/day), 3rd (1434.1e1700.6 mg/day), and 4th (>1700.6 mg/day) quartiles, respectively. In all models there was a statistically significant linear trend for increasing T2D risk across quartiles of phosphorous intake (Ptrend < 0.0001). There was no interaction between dietary phosphorus intake and BMI (P ¼ 0.49), between dietary phosphorus intake and consumptions of high-fat dairy products (P ¼ 0.72), or of processed meat (P ¼ 0.37), or between dietary phosphorus intake and the
Table 1 Characteristics of the women studied by quartiles of the dietary phosphorus intake (E3N-EPIC cohort data, n ¼ 71 270). Variable
Quartile 1 (n ¼ 17 817)
Quartile 2 (n ¼ 17 818)
Quartile 3 (n ¼ 17 818)
Quartile 4 (n ¼ 17 817)
P valuea
Dietary phosphorus intake (mg/day) Individual characteristics Age (years) BMI (kg/m2) Family history of T2D (%) Hypercholesterolemia (%) Hypertension (%) Physical activity (MET-hours/week) Smoker (%) Secondary education (%) Dietary characteristics Energy intake (kcal/day) Alcohol (g/day) Potential renal acid load (PRAL) Dietary calcium intake (mg/day) Dietary magnesium intake (mg/day) Dietary vitamin D intake (mg/day) Use of calcium supplements (%) Use of vitamin D supplements (%)
1028 (135.4)
1321 (66.3)
1559 (75.8)
2001 (274.8)
<0.0001
53 (6.9) 22.4 (3.0) 8.3 38.2 13.4 44.8 (53.5) 14.4 83.9
53 (6.7) 22.7 (3.1) 9.1 39.1 13.4 45.8 (48.1) 13.4 85.3
52.6 (6.5) 22.9 (3.2) 9.7 38.7 13.4 46.0 (47.6) 12.6 86.2
52.2 (6.4) 23.5 (3.5) 9.5 35.5 13.3 49.3 (51.6) 13.2 84.9
<0.0001 <0.0001 <0.0001 <0.0001 0.974 <0.0001 <0.0001 <0.0001
1651 (292) 9.5 (12.3) 8.5 (13.8) 716 (164) 351 (110) 1.9 (0.9) 6.7 2.9
2033 (296) 11.0 (13.1) 5.6 (15.5) 933 (166) 409 (111) 2.3 (0.9) 6.4 3.1
2329 (333) 11.9 (13.8) 3.1 (17.1) 1120 (189) 456 (118) 2.7 (1.1) 6.1 2.9
2833 (474) 13.8 (15.7) 0.3 (22.5) 1496 (348) 547 (146) 3.4 (1.4) 6.2 3.2
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.0884 0.2076
a
Chi-square tests for qualitative variables and Kruskal-Wallis tests for quantitative variables.
Please cite this article in press as: Mancini FR, et al., High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study, Clinical Nutrition (2017), http://dx.doi.org/10.1016/j.clnu.2017.07.025
4
F.R. Mancini et al. / Clinical Nutrition xxx (2017) 1e6 Table 2 Percentage of contribution of the main food groups to the total dietary phosphorus intake (E3N-EPIC cohort data, n ¼ 71 270). Food groups
% of contribution
High fat dairy products Cereal products Fresh meat Low fat dairy products Vegetables Fish Fruit and juices Hot beverages (coffee, tea…) Eggs Processed meat Composite dishes Desserts Others
21.5 14.6 12.3 9.6 8.7 5.1 3.7 3.4 3.2 3.2 2.9 2.7 12.3
Western or Healthy dietary patterns (P ¼ 0.29 and P ¼ 0.18, respectively). The cubic spline regression analysis emphasized a linear relationship between phosphorus intake and risk of T2D from (Fig. 1). 3.3. Sensitivity analyses After adjusting Model 4 for consumption of high-fat dairy products or processed meat, or for the Western or Healthy dietary patterns, the magnitudes of the associations between phosphorus and T2D risk did not materially change (Supplementary Table 1). After having excluded 564 cases that occurred during the first five years of follow-up, we performed the analysis including 1281 cases in order to test a reverse causation hypothesis between phosphorus intake and T2D risk. The results obtained were equivalent to those of the main analyses in terms of magnitude and statistical significance (data not shown). For example, compared to women in the 1st quartile, women in the 4th quartile of phosphorus intake had a HR of 1.46 (95%CI 1.14e1.86, Ptrend < 0.001), as in the main analyses. Finally, no relevant difference was found when comparing the results obtained by the nutrient density method with those obtained with the residual method (Supplementary Table 2). 4. Discussion In this cohort of middle-aged women, a higher intake of phosphorus was associated with increased T2D risk. To the best of our knowledge, the present study represents the first prospective study to investigate the risk of T2D associated with dietary intake of
phosphorus. The relationship between increasing dietary phosphorus intake and T2D appeared to be linear and the increased risk was observed already from the 2nd quartile of phosphorus intake. The association between high phosphorus intake and increased T2D risk remained statistically significant after adjusting for most established or potential confounders and throughout all sensitivity analyses, increasing our confidence in these results. The study population's phosphorus intake (1477 mg/day) is greatly above the adequate intake proposed by EFSA (550 mg/day). These excessive intake levels are mainly related to the high consumption of dairy products, which alone justify over 30% of the total phosphorus intake. 4.1. Biological pathways Although phosphorus is an essential nutrient, it has been suggested that a high dietary intake could induce tissue damage through a variety of mechanisms involved in the endocrine regulation of extracellular phosphate. Indeed, in healthy individuals, phosphorus is absorbed in the small intestine, and under the regulatory action of the parathyroid hormone (PTH) and the fibroblast growth factor-23 (FGF-23), the excesses are excreted with the urines [25]. Current evidence suggests that, over the long term, high concentrations of PTH and FGF-23 may have deleterious effects, and that high dietary phosphorus intake could have an impact on health even in the absence of increased phosphorus levels in the serum [26,27]. In particular, high circulating levels of FGF-23 have recently been associated with insulin resistance, using the HOMAIR index, fat mass, and dyslipidemia [28]. Thus, it could be hypothesized that the relationship between dietary phosphorus intake and T2D might be mediated by the action of FGF-23. Considering the dramatic increase of T2D prevalence worldwide and the burden it represents for health-care [13], it is crucial to better understand all possible dietary-related risk factors and to provide more specific and up-to-date dietary recommendations. Furthermore our findings may have considerable implications in terms of public health due to the widespread use of inorganic phosphorus as food additives. Indeed the amount of phosphorus to which the general population is exposed through the diet has increased in the past decades, due to the growing consumption of foods processed with phosphate additives [29]. Moreover, studies are warranted to investigate the influence of phosphorus intake on diabetes progression and onset of complications, especially considering that phosphorus dietary intake has been reported as an independent risk factor for kidney and cardiovascular disease [29], which are also among the main complications associated to diabetes.
Table 3 Hazard ratios for the risk of incident type 2 diabetes according to quartiles of dietary phosphorus intake (mg/day) estimated by the residual method (E3N-EPIC cohort data, n ¼ 71 270). Dietary phosphorus intake (mg/day)
Quartile 1 (<1203.8 mg/day)
Quartile 2 (1203.8e1434.0 mg/day)
Quartile 3 (1434.1e1700.6 mg/day)
Quartile 4 (>1700.6 mg/day)
P trend
Model Model Model Model Model
REF REF REF REF REF
1.20 1.26 1.26 1.16 1.18
1.58 1.60 1.63 1.36 1.41
1.84 1.80 1.90 1.43 1.54
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001
0 1 2 3 4
(1.04e1.39) (1.08e1.46) (1.08e1.48) (1.00e1.35) (1.00e1.38)
(1.38e1.83) (1.39e1.85) (1.38e1.92) (1.18e1.57) (1.20e1.66)
(1.60e2.10) (1.57e2.07) (1.54e2.34) (1.24e1.65) (1.25e1.90)
Model 0: univariate analysis. Model 1: adjusted for daily energy intake (excluding kilocalories from alcohol consumption, kilocalories/day), alcohol consumption (grams of ethanol/day), level of education (
Please cite this article in press as: Mancini FR, et al., High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study, Clinical Nutrition (2017), http://dx.doi.org/10.1016/j.clnu.2017.07.025
F.R. Mancini et al. / Clinical Nutrition xxx (2017) 1e6
5
Fig. 1. Cubic spline regression model between phosphorus intake and risk of type 2 diabetes (E3N-EPIC cohort data, n ¼ 71 270). Spline regression (3 knots, reference value: Phosphorus intake ¼ 1038.2 mg/day). Continuous line: hazard ratio. Dashed lines: 95% confidence intervals. The model was adjusted for daily energy intake (excluding kilocalories from alcohol consumption, kilocalories/day), alcohol consumption (grams of ethanol/day), level of education (
4.2. Strengths and limitations Potential limitations of the present study should be considered. The dietary estimates are based on a single questionnaire at baseline, thus misclassification of exposure is possible. However, as the study is prospective, any effects are likely to be non-differential and would lead to an attenuation of the true association. The fact that our study population includes only can be considered as a minor limitation due to the fact that only few studies have reported differences in T2D risk factors between sexes, although biological mechanisms may differ between men and women. Finally, dietary phosphorus intake did not include inorganic phosphorus driven from food additives; consequently the study population's phosphorus intake is probably underestimated; again this leads to nondifferential misclassification of exposure, therefore the true association between dietary phosphorus intake and T2D could be stronger than what eventually has been found by the present study. The two main strengths of the present study are the large sample size and the prospective design of the E3N cohort, allowing performing sensitivity analyses, maintaining sufficient statistical power to detect associations and make the reverse causation hypothesis unlikely. Moreover, we analyzed validated T2D cases only, based on a well-defined validation algorithm, which reduces the risk of false negatives or false positives. Moreover, extensive information on potential confounders was collected, minimizing residual confounding, and the dietary questionnaire used to assess phosphorus intake was validated and specific to this French population. 4.3. Conclusions In our study, dietary phosphorus intake was associated with type 2 diabetes risk with a linear dose-effect relationship. Since T2D
represents a high disease burden worldwide, and dietary phosphorus intake is increasing with increasing consumption of highlyprocessed industrial foods among the general population, our results may have important public health implications. More studies are warranted to confirm our findings and better understand the biological mechanisms that undergo this positive association. Contribution statement FRM and GF conceived and designed the study. FCC collected the data. FRM, MCB and GF performed the statistical analysis. FRM drafted the original manuscript. All authors contributed to the interpretation of data discussed in the manuscript, revised the manuscript and approved its final version. Conflict of interest The authors declare that they have no conflict of interest associated with this manuscript. Acknowledgments The authors are grateful to all participants for providing the data used in the E3N Study. This present work was supported by grant from the World Cancer Research Fund (WCRF) (Grant number: 2015/1389), the French Research Agency (ANR, Agence Nationale de la Recherche) via an “Investissement d’Avenir” grant (investment for the future grant, ANR-10-COHO-0006) that supports the E4N study, and the Nutriperso IDEX Paris Saclay University funding. The validation of potential diabetes cases was supported by the European Union (Integrated Project LSHM-CT-2006-037197 in the 6th European Community Framework Programme) InterAct project.
Please cite this article in press as: Mancini FR, et al., High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study, Clinical Nutrition (2017), http://dx.doi.org/10.1016/j.clnu.2017.07.025
6
F.R. Mancini et al. / Clinical Nutrition xxx (2017) 1e6
Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.clnu.2017.07.025. References [1] Knochel JP. Phosphorus. In: Shils ME, Shike M, Ross AC, Caballero B, Cousins RJ, editors. Modern nutrition in health and disease. 10th ed. Baltimore: Lippincott Williams & Wilkins; 2006. [2] Walker HK, Hall WD, Hurst JW. Chapter 198 Serum inorganic phosphorus. In: Bansal Vinod K, editor. Clinical methods: the history, physical, and laboratory examinations. 3rd ed. Boston: Butterworths; 1990. [3] European Food Safety Authority (EFSA). Scientific opinion on dietary reference values for phosphorus. Panel on dietetic products, nutrition and allergies. EFSA J 2015;13(7):4185. [4] Institute of Medicine (US) Standing Committee on the Scientific Evaluation of Dietary Reference Intakes. Dietary reference intakes for calcium, phosphorus, magnesium, vitamin D, and fluoride. Washington (DC): National Academies Press (US); 1997. [5] Calvo MS, Uribarri J. Public health impact of dietary phosphorus excess on bone and cardiovascular health in the general population. Am J Clin Nutr 2013;98(1):6e15. [6] Michaud DS, Spiegelman D, Clinton SK, Rimm EB, Willett WC, Giovannucci E. Prospective study of dietary supplements, macronutrients, micronutrients, and risk of bladder cancer in US men. Am J Epidemiol 2000 15;152(12):1145e53. [7] Alonso A, Nettleton JA, Ix JH, de Boer IH, Folsom AR, Bidulescu A, et al. Dietary phosphorus, blood pressure and incidence of hypertension in the atherosclerosis risk in communities (ARIC) Study and the multi-ethnic study of atherosclerosis (MESA). Hypertension 2010;55(3):776e84. [8] Elliott P, Kesteloot H, Appel LJ, Dyer AR, Ueshima H, Chan Q, et al., INTERMAP Cooperative Research Group. Dietary phosphorus and blood pressure: international study of macro- and micro-nutrients and blood pressure. Hypertension 2008;51(3):669e75. [9] Takeda E, Yamamoto H, Yamanaka-Okumura H, Taketani Y. Dietary phosphorus in bone health and quality of life. Nutr Rev 2012;70(6):311e21. [10] Calvo MS, Park YK. Changing phosphorus content of the U.S. diet: potential for adverse effects on bone. J Nutr 1996;(Suppl. 126):S1168e11180. [11] Calvo MS. Dietary phosphorus, calcium metabolism and bone. J Nutr 1993;123:1627e33. rrez OM, Grams ME. High dietary phos[12] Chang AR, Lazo M, Appel LJ, Gutie phorus intake is associated with all-cause mortality: results from NHANES III. Am J Clin Nutr 2014;99(2):320e7. [13] World Health Organization (WHO). Global report on diabetes. WHO library cataloguing-in-publication data. Geneva: World Health Organization; 2016. [14] European Food Safety Authority (EFSA). Opinion of the scientific panel on dietetic products, nutrition and allergies on a request from the commission
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22] [23]
[24]
[25] [26]
[27]
[28]
[29]
related to the tolerable upper intake level of phosphorus. EFSA J 2005;233: 1e19. Clavel-Chapelon F, van Liere MJ, Giubout C, Niravong MY, Goulard H, Le Corre C, et al. E3N, a French cohort study on cancer risk factors. E3N group. Etude epidemiologique aupres de femmes de l'Education Nationale. Eur J Cancer Prev 1997;6:473e8. van Liere MJ, Lucas F, Clavel F, Slimani N, Villeminot S. Relative validity and reproducibility of a French dietary history questionnaire. Int J Epidemiol 1997;1(Suppl. 26):S128e36. pertoire ge ne ral des aliFavier J-C, Ireland-Ripert J, Toque C, Feinberg M. Re ments: table de composition (Food Composition tables). IQUAL-REGAL. 1995, (French). INRA; AFSSA; CIQUAL; TEC & DO; INRA 927 pp. ISBN 2-85206-921-0. Paul AA, Southgate DAT. McCance & Widdowson's the food composition of foods. 4th ed. Amsterdam, New York, Oxford: Elsevier; 1976. North-Holland, Biomedia Press; MRC Special Report No. 297. Willett WC, Howe GR, Kushi LH. Adjustment for total energy intake in epidemiologic studies. Am J Clin Nutr 1997;65(Suppl. 4):1220Se8S. discussion 1229S-1231S. Brown CC, Kipnis V, Freedman LS, Hartman AM, Schatzkin A, Wacholder S. Energy adjustment methods for nutritional epidemiology: the effect of categorization. Am J Epidemiol 1994 Feb 1;139(3):323e38. Remer T, Manz F. Estimation of the renal net acid excretion by adults consuming diets containing variable amounts of protein. Am J Clin Nutr 1994;59:1356e61. rus U, Willett WC, Hu FB. Dietary fats and prevention of type 2 diabetes. Rise Prog Lipid Res 2009;48(1):44e51. Micha R, Michas G, Mozaffarian D. Unprocessed red and processed meats and risk of coronary artery disease and type 2 diabetesean updated review of the evidence. Curr Atheroscler Rep 2012;14(6):515e24. Cottet V, Touvier M, Fournier A, Touillaud MS, Lafay L, Clavel-Chapelon F, et al. Postmenopausal breast cancer risk and dietary patterns in the E3N-EPIC prospective cohort study. Am J Epidemiol 2009;170:1257e67. Martin A, David V, Quarles LD. Regulation and function of the FGF23/klotho endocrine pathways. Physiol Rev 2012;92(1):131e55. Hjelmesaeth J, Hofsø D, Aasheim ET, Jenssen T, Moan J, Hager H, et al. Parathyroid hormone, but not vitamin D, is associated with the metabolic syndrome in morbidly obese women and men: a cross-sectional study. Cardiovasc Diabetol 2009; 3;8:7. Reis JP, von Mühlen D, Kritz-Silverstein D, Wingard DL, Barrett-Connor E. Vitamin D, parathyroid hormone levels, and the prevalence of metabolic syndrome in community-dwelling older adults. Diabetes Care 2007;30(6): 1549e55. € J, Hammarstedt A, Erben RG, Michae €lsson K, Tivesten A, et al. Mirza MA, Alsio Circulating fibroblast growth factor-23 is associated with fat mass and dyslipidemia in two independent cohorts of elderly individuals. Arterioscler Thromb Vasc Biol 2011;31(1):219e27. Uribarri J, Calvo MS. Hidden sources of phosphorus in the typical American diet: does it matter in nephrology? Semin Dial 2003;16(3):186e8.
Please cite this article in press as: Mancini FR, et al., High dietary phosphorus intake is associated with an increased risk of type 2 diabetes in the large prospective E3N cohort study, Clinical Nutrition (2017), http://dx.doi.org/10.1016/j.clnu.2017.07.025