Table 2:

Vitamin D deficiency is associated with increased rates of incident delirium

Model	Vitamin D at baseline	N	N delirium	Person-years	HR (95% CIs)	p-value
Model 1*	Sufficient (>50 nmol/L)	162,514	1,484	1,728,790
	Insufficient (25–50 nmol/L)	145,890	1,521	1,530,801	1.38 (1.28 to 1.49)	3.9*10–18
	Deficient (<25 nmol/L)	42,916	629	440,778	2.49 (2.24 to 2.76)	3.4*10–68
+ adj. for education + smoking status~		160,120	1,439	1,703,633
	Insufficient (25–50 nmol/L)	143,609	1,472	1,507,279	1.37 (1.27 to 1.48)	1.3*10–16
	Deficient (<25 nmol/L)	42,012	618	431,646	2.38 (2.14 to 2.64)	1.5*10–59
+ excl. vitamin D supplements		140,274	1,246	1,491,439
	Insufficient (25–50 nmol/L)	134,247	1,374	1,408,789	1.40 (1.29 to 1.52)	6.2*10–17
	Deficient (<25 nmol/L)	40,692	598	418,019	2.46 (2.21 to 2.74)	2.1*10–59
+ excl. co-morbidities^		121,070	448	1,290,397
	Insufficient (25–50 nmol/L)	115,022	473	1,209,590	1.32 (1.15 to 1.51)	4.1*10–5
	Deficient (<25 nmol/L)	33,924	211	349,874	2.40 (2.00 to 2.89)	6.0*10–21
+ adj. for calcium~		110,391	406	1,176,061
	Insufficient (25–50 nmol/L)	105,366	447	1,107,064	1.37 (1.19 to 1.58)	8.9*10–6
	Deficient (<25 nmol/L)	31,194	197	321,497	2.45 (2.02 to 2.97)	2.6*10–20
+ adj. Frailty Index~		92,593	327	987,394
	Insufficient (25–50 nmol/L)	86,341	348	907,548	1.31 (1.12 to 1.54)	6.1*10–4
	Deficient (<25 nmol/L)	24,556	153	253,166	2.33 (1.88 to 2.89)	1.3*10–14

Excl.=excluding diagnoses of. Adj.=models adjusted for. N=sample size included in the model. N delirium=the number of incident delirium cases included in the model. HR=Hazard Ratio. CIs=Confidence Intervals.

^*Cox’s proportional hazards regression models adjusted for age, sex, assessment centre, assessment month, and self-reported ethnicity.

^~at baseline assessment, participants with missing data excluded.

^{^}Ever diagnosed with bone fracture, CKD, dementia, liver disease, or Parkinson’s disease in the hospital admissions data to March 2020, prevalent or incident.