Impact of BMI and BMI change on future drug expenditures in adults: results from the MONICA/KORA cohort study

Table 2 Drug utilisation and expenditures by baseline BMI group (unadjusted)

		Total	Normal weight	Overweight	Moderate obesity	Severe obesity	p-value^a
	n	2,946	1,064	1,327	434	121
Number of drugs	Mean	2.14	1.42	2.23	3.09	4.12	<.0001
Number of drugs	SD	2.44	1.80	2.49	2.75	3.23	<.0001
Costs/week (base)	Mean	9.48	6.94	9.69	13.63	14.51	<.0001
Costs/week (base)	95% CI ^b	[8.67-10.35]	[5.63–8.57]	[8.62–10.85]	[10.99–16.76]	[11.65–17.78]	<.0001
Costs/week (prescription drugs) ^c	Mean	8.35	5.67	8.58	12.80	13.55	<.0001
Costs/week (prescription drugs) ^c	95% CI ^b	[7.60-9.11]	[4.51–7.11]	[7.67–9.63]	[10.23–15.78]	[10.84–16.56]	<.0001
Costs/week (7/week) ^d	Mean	12.25	9.67	12.24	17.29	17.09	<.0001
Costs/week (7/week) ^d	95% CI ^b	[11.34-13.20]	[8.09–11.64]	[11.03–13.52]	[14.36–20.71]	[13.96–20.61]	<.0001
Costs/week (discounts) ^e	Mean	9.08	6.60	9.31	13.13	13.93	<.0001
Costs/week (discounts) ^e	95% CI ^b	[8.30-9.93]	[5.35–8.12]	[8.29–10.43]	[10.55–16.15]	[11.26–16.92]	<.0001
Costs per year (base)	Mean	493	361	504	709	755	<.0001
	95% CI ^b	[451–538]	[293–446]	[448–564]	[546–872]	[607–925]

^ap-values were based on Kruskal–Wallis tests.
^b Confidence intervals (CIs) were estimated based on 1,000 bootstrap replications.
^cOnly prescription drugs.
^dIn case the pharmaceuticals were not taken regularly, we assumed 7 days of intake instead of 1 (base analysis) to show up the upper limit of estimated costs as sensitivity analysis.
^ePharmacy discounts, as regularised in the Social Security Code (SGB) V §130 and §130a, were incorporated as sensitivity analysis.

ISSN: 1472-6963