Table 2.

Observational studies on association of vitamin and provitamins A with periodontal disease.

Study Type	Sample	Sex; Age; N	Dietary Intake Assessment	Nutritional Status Assessment	Periodontal Status	Analysis Results (p-Value)	Main Results/Conclusions	Ref
CS	Dental clinic patients (USA)	Both; N.A.; N = 80	Dietary intakes of vitamin A by 24-h recall (N = 56)	Serum levels of vitamin A	RPI	N.S. (mult. linear reg with R2 improvement)	No associations	[36]
CS	Ethnic group subjects (Sri Lanka)	Both; ≥13 y; N = 7944	-	Vitamin A deficiency determined by clinical symptoms	RPI	N.S. (mult. linear reg.)	No association	[39]
CS	NHANES III participants (USA)	Both; ≥20 y; N = 11,480	-	Serum levels of α-carotene (quintiles)	Mild periodontitis 3	OR 1 = 0.60 (95%CI: 0.46–0.77); p = 0.0009 for trend (mult. logistic reg.) in non-restricted model	Negative association of β-& α-carotene, & β-Cryptoxanthin with mild periodontitis; & of β-carotene & vitamin A with severe periodontitis	[38]
						N.S. (mult. logistic reg.) in model restricted to never-smokers
					Severe periodontitis 4	N.S. (mult. logistic reg.) in both models
				Serum levels of β-carotene (quintiles)	Mild periodontitis 3	OR 1 = 0.80 (95%CI: 0.65, 0.98); p = 0.0001 for trend (mult. logistic reg.) in non-restricted model
						OR 1 = 0.99 (95%CI: 0.73–1.35); p = 0.044 for trend (mult. logistic reg.) in model restricted to never-smokers
					Severe periodontitis 4	OR 1 = 0.65 (95%CI: 0.46–0.93); p = 0.001 (mult. logistic reg.) in non-restricted model
						N.S. (mult. logistic reg.) in model restricted to never-smokers
				Serum levels of β-cryptoxanthin (quintiles)	Mild periodontitis 3	OR 2 = 0.74 (95%CI: 0.61–0.89) p = 0.03 for trend (mult. logistic reg.) in non-restricted model
						N.S. (mult. logistic reg.) in model restricted to never-smokers
					Severe periodontitis 4	N.S. (mult. logistic reg.) in both models
				Serum levels of vitamin A (quintiles)	Mild periodontitis 3	N.S. (mult. logistic reg.) in both models
					Severe periodontitis 4	OR 1 = 0.77 (95%CI: 0.58–1.03) p = 0.049 for trend (mult. logistic reg.) in non-restricted model
						N.S. (mult. logistic reg.) in model restricted to never-smokers
CS	Participants in PRIME (Northern Ireland)	Male; 60–70 y; N = 1258	-	Serum levels of retinol (quintiles)	Low-threshold periodontitis 5	OR 2 = 1.64 (95%CI: 1.07–2.51); p = 0.008 for trend (mult. logistic reg.)	Negative association of serum levels of β-carotene & β-cryptoxanthin at both tresholds, & also of α-carotene only at low-threshold	[42]
					High-threshold periodontitis 6	N.S. (mult. logistic reg.)
				Serum levels of α-carotene (quintiles)	Low-threshold periodontitis 5	OR 2 = 1.81 (95%CI: 1.17–2.82); p < 0.001 for trend (mult. logistic reg.)
					High-threshold periodontitis 6	N.S. (mult. logistic reg.)
				Serum levels of β-carotene (quintiles)	Low-thresholdperiodontitis 5	OR 2 = 1.83 (95%CI: 1.19–2.80); p < 0.001 for trend (mult. logistic reg.)
					High-threshold periodontitis 6	OR 2 = 1.87 (95%CI: 0.86–4.06); p = 0.01 for trend (mult. logistic reg.)
				Serum levels of β-cryptoxanthin (quintiles)	Low-threshold periodontitis 5	OR 2 = 1.50; 95%CI: 0.99–2.27); p = 0.02 for trend (mult. logistic reg.)
					High-threshold periodontitis 6	OR 2 = 4.02 (95%CI: 1.61–9.99); p = 0.001 for trend (mult. logistic reg.)
CC	Non-smokers adolescents (Italy)	Female; 17–19 y; N = 54	Dietary intake of vitamin A by 3-d record	-	Gingivitis affected 7 vs. non-affected	N.S. differences between groups (Student t test)	No association	[37]
			Dietary intake marginal deficiency (<1/3 RDA) of vitamin A by 3-d record			N.S. (X2 test with Yates correction)
RC (2 y)	Niigata study participants (Japan)	Both; 75 y; N = 334	Dietary intakes of α-carotene (tertiles) by BDHQ	-	No. of teeth with periodontal disease progression 8	N.S. (mult. logistic reg.)	Negative association with β-carotene intake	[41]
			Dietary intakes of β-carotene (tertiles) by BDHQ			IRR 1 = 0.73 (95%CI: 0.56–0.95); p < 0.05 for trend (mult. logistic reg.)

¹ values compared highest vs. lowest percentile, ² values compared lowest t vs. highest percentile, ³ at least 1 site with CAL ≥4 mm & PPD ≥4 mm, ⁴ ≥2 mesiobuccal sites with CAL ≥5 mm & 1 mesiobuccal sites with PPD ≥4 mm, ⁵ at least 2 teeth with non-contiguous inter-proximal sites with CAL ≥6 mm & 1 PPD of ≥5 mm, ⁶ ≥15% of all sites measured CAL ≥6 mm & there was at least one site with PPD ≥6 mm, ⁷ ≥1 site with BOP, ⁸ CAL ≥ 3 mm in ≥1 site. Abbreviations: 95%CI: 95% confidence interval, BDHQ: brief diet-history questionnaire, BOP: Bleeding on probing; C: Cohort study, CAL: Clinical attachment loss, CPI: Community periodontal index; CS: Cross-sectional study; d: days, h: hours, IRR: incidence rate ratio, mult.: multiple, N: sample size, N.S.: not significant, NHANES III: Third National Health and Nutrition Examination Survey, OR: Odds ratio, PRIME: Prospective Epidemiological Study of Myocardial Infarction, PPD: periodontal pocket depth, RC: retrospective cohort study, RDA: recommended dietary amount, Ref: reference, reg.: regression, RPI: Russell’s Periodontal Index, SD: Standard deviation; US: United states, vs.: versus, y: years.