Skip to main content

Table 2 Characteristics of included case–control studies

From: Evidence of perceived psychosocial stress as a risk factor for stroke in adults: a meta-analysis

Author YearCountry

Quality rating

Cases: controls Type of controls

% Male

Cases: controls with stress

Stress exposure and measure

Stroke outcomes

Previous stroke excluded

Number of confounders controlled for in fully adjusted model

Risk estimates Odds Ratio (95 % CI)

Age range (y)

Jood, 2009 [15] Sweden

7

600 : 600 Population 18–69

64

126:46

Permanent self-perceived psychological stress General stress

Non-fatal ischaemic Stroke

No

10

Single question

3.49(2.06-5.91)

O’Donnell 2010 [16] 22 countries

5

3000 : 3000

63

589:440

Psycho-social stress General stress

Non-fatal stroke (excludes SAH)

Yes

13

Hospital or community

  

Single question

1.30(1.11-1.52)

Mean 61.1 (SD 12.7)

Abel, 1999 [32] USA

5

655:1087

44.6 % cases, 39.9 % controls

Categorical boundary for stress risk factor not used so raw numbers ‘with stress RF’ not reported

20 point increase on GSRRS 35 item Geriatric Social Readjustment Rating Scale [43] General stress

Fatal & non-fatal ischaemic stroke

Yes

7

Community

Mean age

Cases 69.8

1.01(0.99-1.03)

Controls 70.2

Egido, 2012 [29] Spain

6

150:300

77.3 % cases, 36.3 % controls

62:50 (41.4 %:16.7 %)

Score > =150on Holmes & Rahe 40-item questionnaire of life events [41]

Non-fatal stroke (90 % were ischaemic)

Yes

9

Population

8 item ERCTA (Recall Scale of Type A Behaviour [44]

Mean age

Cases 53.8

3.84(1.91-7.72)

(SD 9.3)

Controls 53.6

(SD 9.6)