Table 1 Summary of studies investigating the relationship between lipid levels and ICH

[15] Sturgeon et al. 2007

pooled prospective cohort study (including 2 studies: ARIC and CHS)

21,680

without a history of stroke, American

54.2 (ARIC),72.8 (CHS); 44.8 % (ARIC), 42.4 % (CHS)

3 (ARIC); 5 (CHS)

TG: 1.49 (ARIC), 1.58 (CHS); TC: 5.56 (ARIC), 5.46 (CHS); LDL-C: 3.56 (ARIC), 3.36 (CHS); HDL-C: 1.33 (ARIC), 1.40 (CHS)

135

TG RR 0.56 (95 %CI 0.37–0.84); LDL-C (top 1/4 vs. lower 3/4): RR 0.52 (95 %CI 0.31–0.88)

[11] Bonaventure et al. 2009

population-based prospective cohort study

8393

without a history of stroke, French

≥65; 36.8 %

5

TC: 5.56 (TG ≤0.94), 5.81 (TG 0.95–1.3), 6.08 (TG ≥1.34); LDL-C: 1.82 (TG ≤0.94), 1.62 (TG 0.95–1.33), 1.39(TG ≥1.34); HDL-C: 1.82 (TG ≤0.94), 1.62 (TG 0.95–1.33), 1.39 (TG ≥1.34)

36

TG ≤0.94 mmol/L: adjusted HR 2.35 (95 %CI 1.18–4.70)

[16] Wieberdink et al. 2011

prospective cohort study

9068

without a history of stroke, Dutch

≥55; 57.1 %

9.7

TG, median (IQR): 1.3 (1.0–1.8); TC, median (IQR): 5.8 (5.2–6.5); LDL-C, median (IQR): 3.7 (3.2–4.3); HDL-C, median (IQR): 1.3 (1.1–1.6)

85; 162 CMB in 789 healthy participants

TG: HR 0.20 (95 %CI 0.06–0.69); TG (CMB): HR 0.37 (95 %CI 0.14–0.96)

[17] Zia et al. 2006

population-based nested case–control study

33,346 (1029 stroke-free controls, matched for age, sex and screening-year)

without a history of myocardial infarction or stroke, Swedish

47; 67.3 %

14

/

147

TG (among the lobar ICH): OR 1.7 (95 %CI 0.9–3.2); TG (among the nonlobar ICH):OR 1.4 (95 %CI 0.9–2.3)

[18] Zhou et al. 2003

cross-sectional study

700

without a history of stroke, Chinese

≥57

/

/

500

In ICH patients, TG and LDL-C were significantly increased (P < 0.0001), HDL-C was significantly decreased (P < 0.0001)

[13] O’Donnell et al. 2010

case–control study

4317

be admitted to hospital with first acute stroke whose causes were vascular, 22 countries

66.1; 63.0 %

3

/

663

TC OR 0.62 (99 %CI 0.42–0.92); HDL-C OR 1.91 (99 %CI 1.29–2.83); non-HDL-C: 0.50 (99 %CI 0.34–0.72)

[19] Wang et al. 2013

meta-analysis (19 prospective cohort studies, 4 nested case–control studies)

1,430,141

/

/

/

/

7960

high versus low analysis: TC RR 0.69 (95 %CI 0.59–0.81), LDL-C RR 0.62 (95 %CI 0.41–0.92), HDL-C RR 0.98 (95 %CI 0.80–1.19); dose–response analysis^a: TC RR 0.85 (95 %CI 0.80–0.91), LDL-C RR 0.90 (95 %CI 0.77–1.05), HDL-C RR 1.11 (95 %CI 0.99–1.25)

[20] Suzuki et al. 2010

prospective cohort study

156,892

without a history of stroke, Japanese

53.3; 48.7 %

3

/

361

low TC (<160 mg/dl) was critical risks of ICH

[2] Romero et al. 2014

prospective cohort study

1965

without a history of stroke, American

66.5; 66.0 %

12

TC, mean (SD): 4.95 (0.96); LDL-C, mean (SD): 2.84 (0.83)

173

TC <10th percentile: 1.91 (95 %CI 1.20–3.03); LDL-C <10th percentile: 1.28 (95 %CI 0.75–2.19)

[12] Zhang et al. 2012

prospective cohort study

58,235

without a history of coronary heart disease or stroke, Finnish

25–74; 47.6 %

20.1

/

497

TC in women: <5 mmol/L HR 1.00, 5–5.9 mmol/L HR 0.58 (95 %CI 0.38–0.90), 6–6.9 mmol/L HR 0.40–0.94 (95 %CI 0.40–0.94), ≥7 mmol/L HR 0.50 (95 %CI 0.32–0.78); the relationship of TC and ICH in men was not significant

[21] Mustanoja et al. 2013

observational registry

964 (187 patients used statin before ICH)

Be admitted to hospital with ICH, Finnish

66; 57.1 %

/

TG, median (IQR): 1.0 (0.7–1.3); TC, median (IQR): 4.4 (3.8–5.1); LDL-C, median (IQR): 2.4 (1.8–3.0); HDL-C, median (IQR):1.4 (1.1.–1.8)

964

After adjusting for known ICH prognostic factors, lower LDL-C was independently associated with in-hospital mortality (OR 0.54, 95 %CI 0.31–0.93)